This action might not be possible to undo. Are you sure you want to continue?
I. THE SCOPE OF BUSINESS RESEARCH Business research fulfills the manager’s need for knowledge of the organization, the market, the economy, or other area of uncertainty. It helps the manager predict how individuals, markets, organizational units, or other entities will respond to his business decisions. The emphasis of business research is to shift decision-makers from risky intuitive decisions to decisions based on systematic and objective investigations. A business researcher conducting research may be referred to as a “marketing researcher,” an “organizational researcher,” a “director of financial research,” or one of many other titles. II. BUSINESS RESEARCH DEFINED Business research is defined as the systematic and objective process of generating information for aid in business decisions. This research information should be (1) scientific; not intuitive or haphazardly gathered (2) objective (3) impersonal. Business research can be used for any aspect of the enterprise. By providing appropriate information, research should be an aid to managerial judgment although it should not be a substitute for it. Applying the research is a managerial art in itself. All types of organizations that engage in some kind of business activity can use business research. III. BASIC RESEARCH AND APPLIED RESEARCH There are two types of business research: A. Basic or pure research attempts to expand the limits of knowledge. B. Applied research is conducted when a decision must be made about a specific real-life problem. Both types of research employ the scientific method, the analysis and interpretation of empirical evidence (facts from observation or experimentation), to confirm or disprove prior conceptions. IV.MANAGERIAL VALUE OF BUSINESS RESEARCH The prime managerial value of business research is that it reduces uncertainty by providing information that facilitates decision making about strategies and the tactics used to achieve an organization’s strategic goals. The decision-making process involves three interrelated stages. A. Identifying the existence of problems and opportunities. Before any strategy can be developed, an organization must determine where it wants to go and how it will get there. Business research can help managers plan strategies by determining the nature of situations by identifying the existence of problems or opportunities present in the organization. Diagnosis and Assessment . After an organization recognizes a problem or identifies a potential opportunity, an important aspect of business research is often the provision of diagnostic information that clarifies the situation. Managers need to gains insight about the underlying factors causing the situation. If there is a problem they need to specify what happened and why. If an opportunity exists they may need to explore, clarify, and refine the nature of the opportunity. Selecting and implementing a course of action. Business research is often conducted to obtain specific information to help evaluate the various alternatives, and to select the best course of action based on certain performance criteria. Evaluation of the course of action. Evaluation research is conducted to inform managers whether planned activities were properly executed and whether they accomplished what they were expected to do. It serves an evaluation and control function. Evaluation research is a formal, objective appraisal that
provides information about objectives and whether the planned activities accomplished what they were expected to accomplish. This can be done through performance-monitoring research, which is a form of research that regularly provides feedback for evaluation and control of business activity. If this research indicates things are not going as planned, further research may be required to explain why something “went wrong.” V. WHEN IS BUSINESS RESEARCH NEEDED? A manager considering whether or not to conduct research must consider four things: A. Time constraints: Is there sufficient time available before a managerial decision must be made? B. Availability of data: Is the information already on hand adequate for making the decision? If not, can appropriate information be made available? C. Nature of the decision: Is the decision of considerable strategic or tactical importance? D. Benefits vs. costs: Does the value of the research information exceed the cost of conducting the research? If the answer to any of these questions is “no,” then business research should not be conducted. VI.MAJOR TOPICS IN BUSINESS RESEARCH There are many topics that benefit from business research. Some major topics are: general business, economic, and corporate research; financial and accounting research; management and organizational research; sales and marketing research; information systems research; and corporate responsibility research. VII. TOTAL QUALITY MANAGEMENT AND BUSINESS RESEARCH Business research is an essential aspect of total quality management. Total quality management is a business philosophy, which has much in common with the idea that the consumer is king. It embodies the belief that the management process must focus on integrating the idea of customer-driven quality throughout the organization. Total quality management stresses continuous improvement of product quality and service. Managers improve durability and enhance products with additional features as the product ages. They strive to improve delivery and other services to keep their brands competitive. Companies that have adopted the total quality philosophy view employees as internal customers. Implementing a total quality management program requires considerable measurement. It involves routinely asking customers to rate a company against its competitors. It involves measuring employee attitudes. It involves monitoring company performance against benchmark standards. In short, it uses business research. Thus, business research with external customers and with employees in the organization (internal customers) is important under a total quality management program. VII. BUSINESS RESEARCH IN THE 21st CENTURY Business research, like all business activity today, has been strongly influenced by two major trends in business: A. Increased globalization B. Rapid growth of the Internet and other information technologies Business today is global. Business research, like all business activity, has increasingly become more global. “The Internet is transforming society. Time is collapsing. Distance is no longer an obstacle. Crossing oceans is only a mouse click away. People are connected 24 hours a day, seven days a week. “Instantaneous” has a new meaning.” Internet and other information technologies are dramatically changing the face of business research. Getting information from Web sites and conducting Web surveys are examples.
Chapter 2 Information Systems and Knowledge Management
DATA, INFORMATION, AND KNOWLEDGE MANAGEMENT
There is a difference between data and information. Data—the raw facts—record measures of certain phenomena which are necessary to provide information—facts in a form suitable for managers to base decisions on. Knowledge is a blend of information, experience, and insights that provide a framework that can be thoughtfully applied when assessing new information or evaluating relevant situations. Knowledge management is a process to create an inclusive, comprehensive, easily accessible organizational memory, which is often called the organization's intellectual capital. The purpose of knowledge management is to organize the 'intellectual capital' of an organization in a formally structured way for easy use. II. GLOBAL INFORMATION SYSTEMS
The well being of multinational corporations, indeed the health of any business organization planning to prosper in the 21st century, depends on information about the world economy and global competition. Contemporary marketers find that much information can be made instantaneously available. This has changed the way business is conducted. A Global Information System is an organized collection of computer hardware, software, data, and personnel designed to capture, store, update, manipulate, analyze, and immediately display information about worldwide business activity III. DECISION SUPPORT SYSTEMS
A decision support system is a computer-based system that helps decision-makers confront problems through direct interaction with databases and analytical models. The purpose of a decision support system is to store data and transform them into organized information that is easily accessible to marketing managers. Decision support systems serve specific business units within a company. The decision support system operates within the context of the global information system. A decision support system is not independent from the global information system. Most major corporations have decision support systems to provide customer relationship management (CRM). A CRM system might bring together lots of pieces of information about customers, sales, marketing effectiveness, responsiveness and market trends. The goal of the CRM decision support system is to describe customer relationships in sufficient detail so that management, salespeople, people providing customer service, and perhaps the customer directly can access information, match customer needs with satisfying product offerings, remind customers of service requirements, know what other products a customer had purchased, and so forth. A. Database: This is a collection of information that is arranged in a logical manner and organized in a form that can be stored and processed by a computer.
A mailing list of customer names is one type of database. Population characteristics recorded by state, county, and city may be contained in another database. Databases often exist in computer storage devices such as hard disk drives. But, other types of databases may exist in a vendor company's computers. The concept of a data warehouse is a term that managers of information technology use to discuss the multi-tiered, computer storage of current and historical data and the mechanics of selecting and using information, which is relevant to decision-making tasks. Data warehouse management requires that the detailed data from operational systems be extracted and transformation so that layers of summarized data "tables" can be stored (warehoused) so the various data bases are consistent. Organizations with data warehouse may integrate databases from both inside and outside the company. B. Software: The software portion of a decision support system consists of various types of programs that tell computers, printers, and other hardware what to do. Software involves business intelligence systems, statistical software, spreadsheet software, and decision model banks that combine and restructure databases, diagnose relationships, estimate variables, and otherwise analyze the data within the system.
Optical scanners in supermarkets and other retail outlets provide a wealth of product and brand sales information collectively known as scanner data or single-source data. and certain portions are available on the Internet. provides a characteristic example. Computerized database searches offer the most efficient and exhaustive way to find published information. and many others within the organization are involved in data collection and providing input for the decision support system. Systematic accumulation of pertinent. timely. Printed copies of this book can be found in most public and corporate libraries. even more data will be stored in digitized form in computerized data archives. As the twentyfirst century progresses. Geographic Databases. INPUT MANAGEMENT Input include all the numerical data. . systems analysts. and images entering the decision support system. proprietary business research. such as Dow-Jones News Retrieval and Bloomberg Financial Markets without leaving their offices. Business researchers. on CD-ROM. production managers. businesspeople can use personal computers to access online information search and retrieval services. collections of organized and readily retrievable data were available in published form at libraries. Statistical Databases. Most online information services use electronic indexing systems that function much like the periodic indexes found in book form in libraries.A decision support system’s business intelligence software allows managers to combine and restructure databases. some information services can be accessed from remote locations via digital wireless devices. Numerous computerized search and retrieval systems and electronic databases are available as subscription services or in libraries. diagnose relationships. The managers of the decision support system. IV. and computer programmers are responsible for the system as a whole. sales personnel. However. and other relevant marketing variables. sales. Certain databases can be accessed on the Internet. In fact. A researcher can query the library computer to learn whether the library owns a particular book and whether that book is on the shelf or checked out. and otherwise analyze the various databases. Input data can come from sources internal to the organization or from sources outside the company. text. Statistical databases are exemplified by geographical information systems and scanner databases. the task of providing input data is shared by many functions within the organization. COMPUTERIZED DATA ARCHIVES Historically. Scanner Databases. discover patterns. Statistical databases contain numerical data for market analysis and forecasting. The Statistical Abstract of the United States. the Statistical Abstract has also become available in a digital format on CD-ROM. In recent years. voice. and outside vendors/external distributors of data. marketing intelligence. Today. Decision analysts apply complex statistical procedures and computerized decision models to such data. Each source can provide valuable input. which is filled with tables of statistical facts. and accurate data is essential to the success of a decision support system. corporate librarians. estimate variables. A geographic information system is a decision support system that maps geographic areas by demographic. V. accountants. Most online information services also have software that lets the researcher do keyword searches for information in its database Several types of databases from outside vendors and external distributors are so fundamental to decision support systems that they deserve further explanation. The entire 2000 census is also available in print. The four major sources of input for marketing data: internal records.
com). or organizational name associated with the host computer. It has changed the way millions of people think about getting and distributing information. Noncommercial and commercial organizations make a wealth of data and other resources available on the Internet. the manufacturer.com) that allows college professors to access information about the company and its textbooks. names have two or more fields. so does e-mail. An email address consists of two parts. South-Western has an online directory (http://www. and so on. . These are easy to access in financial databases. The Internet consists of host computers that access servers.org). WHAT IS THE INTERNET? The Internet is a worldwide network of computers that allows users access to information and documents from distant sources.The Universal Product Code (UPC) contains information on the category of goods.adcritic. For example. institutional name. Networking is the linking of two or more computers to share data and software programs.jp) or by type of activity (.com. movie studios provide clips of upcoming films and advertising agencies put television commercials on the Internet. CompuStat publishes a financial database of extensive financial data about thousands of companies broken down by industry and other criteria. color. NETWORKS AND ELECTRONIC DATA INTERCHANGES Although personal computers work as independent units.uk. A host is a computer that one or more people use directly by logging onto a personal computer connected to it in order to access network services. Electronic data interchange (EDI) is the term used when one company's computer system is integrated directly with another company's proprietary computer system. A list server permits subscribers to a mailing list to communicate with one another around the globe. The name of the system on which the mailbox resides is on the right-hand side. Computer communication and resource discovery are two central functions of the Internet." The domain. Users send messages on the Internet by e-mail (electronic mail). A user discussion server permits multiple users to communicate in real-time with one another. S. Video Databases Video databases and streaming media are relatively new but they are have a major impact on the marketing of many goods and services. It is a combination of a worldwide communication system and the world's largest public library for a seemingly endless range of information. such as income statements and balance sheets. Just as a letter delivered by a postal worker requires an address.swcollege. NAVIGATING THE WORLD WIDE WEB . VII. Library of Congress provides full text of all versions of House and Senate legislation and full text of the Congressional Record and the Internal Revenue Service makes it possible to download an income tax form. they can be networked to other computers. or domain. For example. The name of the user's mailbox is on the left-hand side. separated by dots and can follow many different naming schemes. . the U. There are several types of servers. is typically a company name. Financial Databases. VIII. may interest marketing managers. Competitors’ and customers’ financial data. VI. A server is a computer that provides services on the Internet. A file server contains documents and programs that can be accessed and downloaded via the host to a user's own personal computer. such as by country (. separated by the “at” symbol "@". (See www. System. flavor. The Internet allows users to access a great deal of information instantaneously and effortlessly. and product identification based on size.
These menu-based software systems present the user with a menu of the files on a particular server and make it easy to move from server to server on the Internet (often called navigating the net. newspapers. academic associations. A web site is any computer host acting as a location that can be accessed with the browser software. A hypermedia publishing system is a system based on pages of text and images that are stored on thousands of computers around the world. or URL. However. such as Netscape Navigator and Microsoft Explorer. a graphical interface system of thousands of interconnected pages (documents) allows access to all the resources of the Internet from one interface displayed on any computer screen. Popular web browsers. The URL is really just a web site address that web browsers recognize. into the web program. TV networks. or an emulation program that simulates a high-speed phone line. or Mac-based machine). HTTP (Hypertext Transfer Protocol) is a method for transferring and displaying HTML information on the Internet. and a highspeed phone line (available at low costs through service providers around the country). universities. many commercial sites require that the user have a valid account and password before access is granted. The World Wide Web allows users to point and click where they want to go and to call up video. INTERACTIVE MEDIA AND ENVIRONMENTAL SCANNING People who use the Internet interact with their computers.or Mac-based machine). government agencies. which may be on any computer connected to the Internet. Prodigy Browser. are menu-based software systems that present the user with a menu of the files on a particular server and make it easy to move from server to server on the Internet (often called navigating the net. Popular web browsers include Netscape. With search engines such as Yahoo. Environmental scanning entails all information gathering that is designed to detect indications of environmental changes in . and the computer responds in sophisticated ways so the user and equipment can have a continuing conversation. A researcher trying to find a particular site or document on the Internet or just looking for a resource list on a particular subject. and corporations have set up documents consisting of graphical pages of information on servers connected to the Internet. or surfing). To access the World Wide Web. The introductory page or opening screen is called the home page because it provides basic information about the purpose of the document along with a menu of selections or links that lead to other screens with more specific information.The World Wide Web (WWW). and graphics from different participating computer networks around the world. or by having a unique icon. navigation. Windows-. The World Wide Web is a do-it-yourself hypermedia publishing system. each page can have connections or hyperlinks to other pages. A search engine is a computerized directory that allows anyone to search the World Wide Web (WWW) for information in a particular way. a modem. by being underlined. People using the World Wide Web may be viewing information on their host computer or on a machine halfway around the world. With these web browsers even a novice on the Internet can search for information by simply using point-and-click graphics that resemble the familiar Windows or Macintosh interface. Dogpile and InfoSeek all a researcher has to do is type the search in plain English or just enter key words and phrases. and sound clips stored on the host computer or on computers thousands of miles away. Excite. video clips. a typical home user needs a web browser (a software program with a graphical user interface that can run on a computer whether it is a Unix-. Windows. sound bytes. navigation. it will have normal paragraphs of information along with a code known as HTML (hypertext markup language) that makes references to other pages as well as graphic images. If you look at a raw page. The links to other documents are usually highlighted by appearing in another color. Most web browsers also allow the user to enter a Uniform Resource Locator. Over the past few years. It is an interactive medium because a user clicks a command. libraries. To access the World Wide Web. Unix-. IX. JAVA is programming language that also helps display information on the Internet. one needs a web browser (a software program with a graphical user interface that can run on a computer whether it is a DOS-. Thus. can use one of the many available on-line search engines. Many web sites allow any user or visitor to access its web pages without previous approval. or surfing). AltaVista. and Microsoft Explorer.
where data is sent without a request being made. There is a well-organized and purposeful effort to retain. and governmental affiliates in the United States. X. it is said to be pulled. When a Web page isn't delivered until a browser requests it. even video and voice) is available only inside the organization. INTERNET2 Information technology changes rapidly. Web pages and browsers to build a communications and data resource at a specific company. Today's information technology uses as “smart agents” or “intelligent agents” to deliver customized content to the viewer’s desktop. Intranets function to make the knowledge of company "experts" more accessible throughout their organizations.their initial stages of development. a key difference between the Internet and an intranet is that "firewalls" or security software programs are installed to limit access to only those employees authorized to enter the system. Internet2 (http://www. Thus. such as the Yahoo portal. In organizations that practice knowledge management. A company's intranet uses Internet features. The opposite of pull is push. INFORMATION TECHNOLOGY The terms pull technology and push technology describe how data is sent over the Internet. As sophisticated as the Internet and Intranets are today. The goal of Internet2 is to build and operate a research network with greater than 100 times the bandwidth typical of the current Internet. industrial organizations. such as Internet2.edu/) is a collaborative effort involving over 100 universities. analyze and organize employee expertise so that it easily obtainable any time from anywhere. . Because of its vastness. new technologies. graphics. including push phase technology) allows a Web site. XI. will dramatically enhance researchers' ability to answer marketing problems in the future. Smart information delivery (know by a variety of technical names. Smart agent software is capable of learning an Internet user’s preferences and automatically searching out information in selected Web sites and then distributing the information on a user’s computer for later viewing. The information on an intranet (data. MyYahoo and MyExcite are portal services that personalized Web pages. What is an Intranet? An Intranet is a company's private data network that uses Internet standards and technology. such as electronic mail. to becomes a one-on-one medium for each individual user. the Internet is an especially useful source for scanning many types of changes in the environment.internet2.
Thus. THE MEANING OF THEORY A theory is a coherent set of general propositions. V. Things are not the essence of theory. Thus. NATURE OF PROPOSITIONS Concepts are basic units of theory development.1 indicates that it is possible to discuss concepts at various levels of abstraction. IV. used as principles of explanation of the apparent relationships of certain observed phenomena. THE SCIENTIFIC METHOD A scientific method is a set of prescribed procedures for establishing and connecting theoretical statements about events and for predicting events yet unknown. theory deals with abstraction. the nature of theoretical concepts must be understood.” and “morale. or purposes. III. but to predict phenomena. theories require that the relationship among concepts be understood. The ability to anticipate future conditions in the environment or in an organization may be extremely valuable. Concepts. Concepts are our building blocks and some examples of organizational theory concepts might be “leadership. yet prediction alone may not satisfy the scientific researcher’s goals. prediction and understanding go hand in hand. Accomplishing the first goal allows the theorist to predict the behavior or characteristics of one phenomenon from the knowledge of another phenomenon’s characteristics. Theories are at the highest level of abstraction. of course. however. Before a proposition can be explained.” “productivity. However. WHAT ARE THE GOALS OF THEORY? Prediction and understanding are the two purposes of theory. Variables may be measured. concepts are expressed in words that refer to various events or objects. That is. A hypothesis is a proposition that is empirically testable.2). The abstraction ladder in Exhibit 3. occurrences. Bonus pay and sales are variables. we must have an explanation of why variables behave as they do. Researchers are concerned with the observable world. Exhibit 3.4 shows that the hypothesis “Bonus pay will be associated with sales volume consistently above quota” is an empirical counterpart of the proposition. reflecting concepts at the empirical level. they are networks of propositions. Concepts are at one level of abstraction (see Exhibit 3. the scientific method has two basic levels: the empirical level which is primarily concerned with the facts of the science as revealed by . Investigating propositions requires that we increase our level of thinking.Chapter 3 Theory Building I. or what we will loosely term “reality.” Theorists translate their conceptualization of reality into abstract ideas. The basic or scientific business researcher operates at two levels: the abstract level of concepts (and propositions) and the empirical level of observation and manipulation of objects and events (see Exhibit 3. Understanding is desired. In most situations. Only when we begin to explain how concepts relate to other concepts do we begin to construct theories. ideas are.” CONCEPT Concepts abstract reality. CONCEPTS Theory development is essentially a process of describing phenomena at increasingly higher levels of abstraction. There is no consensus concerning exact procedures for the scientific method. may vary in degree of abstraction. A (or construct) is a generalized idea about a class of objects. A theory is an abstraction from observed reality. but most discussions of the scientific method include a reference to “empirical testability.” The process of empirical observation cannot be divorced from the process of theory development. II.3). PROPOSITIONS are statements concerned with the relationships among concepts.
therefore. theory may be developed with deductive reasoning. Statement of hypotheses D. For example. Theories provide a framework that can guide managerial strategy by providing insights into general rules of behavior. Inductive reasoning is the logical process of establishing a general proposition on the basis of observation of particular facts. the researcher must understand the difference between facts and theories. then we can deduce that Mary Fisher is a human being.5 is a simplified portrayal of a theory to explain “voluntary job turnover. VI.” Two concepts—(1) the “perceived desirability of movement” to another job and (2) the “perceived ease of movement” from the present job—are expected to be the primary determinants of “intentions to quit. all managers are human beings. The concept “intentions to quit” is expected to be a necessary condition before the actual “voluntary turnover behavior” occurs. conceptual level. However. Acquisition of meaningful empirical data F. Deductive reasoning is the logic process of deriving a conclusion from a known premise or something known to be true. researchers gather empirical data or evidence to verify theories. . and to integrate them into a logical system. we know that all managers are human beings. To determine which is the better theory.observation. Assessment of relevant existing knowledge B. VIII. Theory construction is often the result of a combination of deductive and inductive reasoning. IX. Facts are the world’s data. All managers that have ever been seen are human beings. Analysis and evaluation of data G. Design the research to test the hypotheses E. One task of science is to determine if a given theoretical proposition is false or if there are inconsistencies between competing theories—theories are made to be tested. Theory generation may occur at either level. This is a second proposition that links concepts together in this theory. Theories are structures of ideas that explain and interpret facts. At the empirical level. Business research gathers facts to verify theories. AN EXAMPLE OF A THEORY Exhibit 3. HOW ARE THEORIES GENERATED? In this chapter. and the abstract or theoretical level which consists of a serious attempt to understand the facts of the science. conceptual level and at the empirical level. If we also know that Mary Fisher is a manager. theory has been explained at the abstract.” This is a proposition. Provide explanation and state new problems raised by the research X. A good theory allows us to generalize beyond individual facts so that general patterns may be predicted and understood. Facts do not go away when scientists debate rival theories to explain them. PRACTICAL VALUE OF THEORIES Theories allow us to generalize beyond individual facts or isolated situations. theory may be developed with inductive reasoning. OVERVIEW OF THE SCIENTIFIC METHOD Seven operations may be viewed as the steps involved in the application of the scientific theory: A. Formulation of concepts and propositions C. At the abstract. VII. VERIFYING THEORY In most scientific theory there are alternative theories to explain certain classes of phenomena.
subsequent research is usually required. to identify how much information is available.” “what.2). surveys.” “when. and to recognize what information is needed.g.” It does not tell us “why. Exploratory research merely crystallizes the problem and identifies information needed. It is not intended to provide conclusive evidence from which a particular course of action can be determined. decision making is the process to resolve a problem or to choose from alternative opportunities. 2. Measure the concomitant variation (the occurrence of the two phenomena or events varying together) between the presumed cause and the presumed effect. The six stages in the research process (see Exhibit 4. another will follow (e.” “where. increase advertising).” Although it is impossible to completely eliminate error. Before inferring causality. INFLUENCES OF UNCERTAINTY ON THE TYPE OF RESEARCH Uncertainty of the research problem influences the type of research to be conducted.g. Causal research attempts to identify cause-and-effect relationships between variables. B.all of which are functionally interrelated and overlap chronologically (see Exhibit 4. STAGES IN THE RESEARCH PROCESS Systematic enquiry continues throughout all stages of the research. it can be broken into three categories: A.3). Every business problem or decision making situation can be classified on a continuum ranging from complete certainty to absolute ambiguity.) available and the results one wants. manpower. Exploratory studies are conducted to classify the nature of problems. IV. Recognize the presence or absence of alternative plausible explanations or causal factors. DECISION ALTERNATIVES IN THE RESEARCH PROCESS A number of alternatives are available to the researcher during each stage of the research process. there may be a “plurality of causes” for the observed effect. C. the researcher should: 1. Even after these three criteria have been established. 3. Classifying business research on the basis of its purpose. Both forward linkage — when the earlier stages of research will influence the design of the later stages — and backward linkage — when the later stages have an influence on the earlier stages — exist in a research process. therefore. The key to decision making is to recognize the nature of the problem/opportunity. money. the researcher can never be certain that the causal explanation is adequate. The more certain and defined the problem is. Descriptive research is conducted to discover and determine the characteristics of a population. increased sales). III.4) are as follows: .. IV. TYPES OF BUSINESS RESEARCH Business research can be classified on the basis of technique (e.” and “how.g. the researchers are quite knowledgeable about the subject. but the research can be broken into six stages . It seeks to determine the answers to the questions “who. etc. Establish the appropriate causal order or sequence of events. and the method chosen will depend on the resources (time. II. There is no best way to research. experiments. It usually follows exploratory and descriptive research and. unlike exploratory research.Chapter 4 The Research Process: An Overview I. this type of research should be conducted as accurately as possible because. Causal research attempts to establish that when we do one thing (e.. or observation studies) or purpose. DECISION MAKING Formally defined. the evidence it provides will be used to determine a course of action. the more the researcher will move from exploratory to descriptive to causal research (see Exhibit 4..
Business research must have clear objectives and definite designs. there are many alternative methods for solving a problem. 3. For the inexperienced researcher it can be difficult to administer and its biggest disadvantage is that it cannot observe intangible states of mind. A pilot study may send a manager out to experience what workers. Planning the research design: A research design is a master plan specifying the methods and procedures for collecting and analyzing the needed information. consumers. the design technique (e. Evaluating research designs: There is no one best research design. yet most neglected. Secondary data can often be found inside the company. Secondary data or historical data is data previously collected and assembled for some project other than the one at hand.. urgency. primary data is gathered specifically for the project at hand. that is defining the problem. at public libraries and universities. availability of data. but it can be outdated or may not exactly fit the researcher’s needs. One example of experimentation is test marketing. An experiment controls conditions so that one or more variables can be manipulated in order to test a hypothesis. or person-to-person. Secondary data can usually be gathered faster and at less cost than primary data. survey or experiment). . 2. Problem discovery and definition: Most research programs are initiated because managers are uncertain about some aspect of the firm’s marketing program. Selecting the appropriate research design: There are four basic research designs for causal and descriptive research and certain factors. 4. cost. 6. Problem definition is possibly the most important.g. the sampling methodology. 3.A. Questionnaires can be administered by mail. 2. gives a sense of direction to the investigation. or may take the form of a focus group interview that is a loosely structured group dynamics session. Observation techniques: The objective of research is often merely to record what can be observed.g. but may suggest possible topics for formal investigation. or purchased from a firm specializing in providing information. Exploratory research progressively narrows the scope of the research. but it still has the same advantages and disadvantages attached to it. and all are appropriate in different situations. B. case studies. e. area of business research. Each administration method has its advantages and disadvantages. Experiments: Hold the greatest potential for establishing cause-and-effect relationships because they allow investigation of changes in one variable while manipulating one or two other variables under controlled conditions. Surveys: The most common method of generating primary data. Knowing how to select the most appropriate research design develops with experience. or others experience. The advantage of it is that it is nonreactive and the data can be unobtrusively and passively collected without a respondent’s direct participation. exploratory research is often used because it helps refine the problem into one that can be researched. 1. should determine which design is used. In contrast. The data collection methods are informal and findings may lack precision. and experience surveys. There are four basic exploratory techniques: secondary data. the time schedule for the research. and the budget for the research. pilot studies. Pilot studies (a collective term to group together a number of diverse research techniques) collect data from the ultimate subject of the research project to serve as a guide for the larger study... At this stage in the project. It is a technique in which information is gathered from a sample of people by use of a questionnaire. etc. This technique uses greater quantitative sophistication than is involved in using secondary data at the exploratory level of research. phone. 5. Two of these are discussed in this overview chapter. 1. The research design should consider the research objectives and should determine the sources of information. Identifying the problem. Causal factors can be isolated because outside factors do not come into play. Secondary data studies: Typically this involves using past data to project future figures for a particular event or phenomena.
this process is called coding and it facilitates computer or hand tabulation. and consistency in classification. There are two basic types of samples: 1. it varies from the simple determination of patterns to complex statistical analysis. or not a probability sample.a sample is a subset from a larger population. The program strategy refers to a firm’s overall plan to utilize business research. Data collection: Once the research design. legibility. Different types of research may be going on at the same time. E. or that uses part of the population to make a conclusion regarding the whole population . F. Non-probability sample: In which the sample is selected on the basis of personal judgment. through editing and coding. The results of a statistically sound sample should have the same characteristics as the population.. into a format that will answer the marketing manager’s questions.C. or if the unit is not randomly selected. while an unobtrusive method of collecting data does not. then it is totally useless. or if the sample size is too small. RESEARCH PROJECT VERSUS RESEARCH PROGRAM Business research is not a one-shot project. D. Sampling: Involves any procedure that uses a small number of items. Of course. If a research report is too long. the survey method requires direct participation. Editing involves checking the data forms for omissions. has been formalized. VI. etc. The research report should effectively communicate the research findings. . For example. meaningful categories must be established for groups of responses . Before the edited data can be tabulated.Data processing and data analysis: Data must be converted. Analysis is the application of logic to the understanding of data that has been gathered. for managers to use. the process of collecting information from respondents may begin. errors can be made especially if the target population is not representative of potential customers. Probability sample: In which every member of the population has a known. 2. A pre-testing phase. including the sampling plan. there are many methods of data collection. using a small subsample. Conclusions and report preparation: This stage involves preparing the information and making conclusions which will be useful for decision-making purposes. may help minimize errors in the main study. Depending on the design chosen. it is an ongoing process. complex. non-zero probability of selection.
GENERAL RIGHTS AND OBLIGATIONS OF CONCERNED PARTIES In most research situations three parties are involved: The researcher. Consciously or unconsciously.1 diagrams this relationship. . In general. For instance. certain standards of common courtesy have been set by interviewing firms. to avoid possible biased reactions. The Obligation to be Truthful This issue involves the subject’s freedom to choose whether or not to comply with the investigation request. B. They suggest what a member of a group ought to do under given circumstances. A. Generally. RIGHTS AND OBLIGATIONS OF THE RESPONDENT The ethical issues vary somewhat depending on whether or not the participant has given willing and informed consent. The interaction of each of these parties identifies a series of ethical questions. such deception is justified under two conditions: (1) Investigators assume that no physical danger or psychological harm will be caused by the deception. the poorly educated. In an unobtrusive observation study. the researcher creates a false impression by disguising the purpose of the research. Deception In a number of situations. a field worker is not expected to overstep the boundary society places on consumer privacy. For example. the participant’s rights differ from a survey respondent’s rights because he or she has not willingly consented to be a subject of the research. The issue of deception concerns the means-to-an-end philosophical issue. The major question is: Does a small deception substantially increase the value of the research? Observation methods present a number of problems concerning the deception of subjects. critics have argued that the old. Another aspect of the privacy issue is illustrated by the question “Is the telephone call that interrupts someone’s favorite television show an invasion of privacy?” Generally. “I don’t care to answer your question about my income” believes that he or she has the right to refuse to participate. and (2) the researcher takes personal responsibility for informing the respondent of the concealment or deception after the research project ends. Generally it is believed unobtrusive observation of public behavior in such places as stores. the subject is lied to. III. In an observation study. A respondent who says. II. each party expects certain rights and feels certain obligations toward the other party. Traditionally. ethical problems may arise. recording private behavior with hidden cameras represents a violation of this right. For each of the subject’s rights there is a corresponding obligation on the part of the researcher. ETHICAL QUESTIONS ARE PHILOSOPHICAL QUESTIONS The rights and obligations of individuals are generally dictated by the norms of society. airports etc. If there are conflicting perspectives about behavioral expectations. Societal norms are codes of behavior adopted by a group.Chapter 5 Ethical Issues in Business Research I. the poor. Yet some researchers will persist in trying to get that information. This chapter reflects the author’s perceptions of the norms of our society. the major ethical issues concern whether the observed behavior is public or private. the individual’s right to privacy dictates that the researcher has an obligation to protect the anonymity of the respondent. the sponsoring client (user) and the respondent (subject). and other underprivileged individuals may not be aware of their right to choose or of their right to be left alone. Exhibit 4. researchers have assumed that individuals make an informed choice. is not a serious invasion of privacy. several ethical issues concern the researcher’s expected rights versus those of the respondent/subject. Bluntly stated. However. However.
scheme. An Open Relationship with Interested Parties . If the public understands why survey or experimental information has been collected and that the researchers may be trusted with private information. Dissemination of Faulty Conclusions Most research code of ethics state something like “a user of research shall not knowingly disseminate conclusions from a given research project or service that are inconsistent with or not warranted by the data. Protect the Right to Confidentiality of Both Subjects and Clients It is the researcher’s responsibility to ensure that the privacy and anonymity of the respondents are preserved. it may be easier in the long run to conduct research. A number of professional associations have developed a code of ethics—standards and operating procedures for ethical practice by researchers.” V. debriefing is often performed. office. If the respondent’s name and address are known. Objectivity Researchers should maintain high standards to ensure that the data are accurate. The Right to be Informed It has been argued that subjects have a right to be informed of all aspects of the research. RIGHTS AND OBLIGATIONS OF THE CLIENT SPONSOR (USER) Ethics between Buyer and Seller The general business ethics expected between a purchasing agent and a sales representative should hold in the marketing research situation. or other establishment.C. This sales ploy is considered to be unethical as well as illegal. this information should not under any circumstances be forwarded to the sponsoring organization. Misrepresenting Research Basically it is assumed that the researcher has an obligation to both the client and the subjects to honestly analyze the data and to correctly report the actual data collection methods. Several major issues exist that should be further explored in this book: The Purpose of Research is Research Business people are expected to not misrepresent a sales tactic as marketing research. RIGHTS AND OBLIGATIONS OF THE RESEARCHER General business ethics should be a standard for business research firms and business departments. This includes information about its purpose and sponsorship. or ruse that misrepresents the true status of the person making the call as a door-opener to gain admission to a prospect’s home. The Federal Trade Commission has indicated that it is illegal to use any plan. More has been written about the ethics of researchers than about those of the other parties because this group’s purpose is clearly identifiable. When there is a clear-cut deception or when the researcher perceives that there may be psychological harm caused by participating in an experiment (a rarity in business research). IV. An Open Relationship with Research Suppliers The client sponsor has the obligation to encourage the research supplier to objectively seek out the truth.
VI.Conclusions should be based on the data. attorneys in an advocacy research trial rarely will submit research evidence that does not support the client’s position. However. Thus the slightest variation from technically correct sampling procedures may be magnified by an attorney until a standard market research project no longer appears adequate in the judge’s eye. VII. The results of advocacy research rarely are presented unless they support the advocate’s position(s). A FINAL NOTE ON ETHICS There is no question that there are unethical researchers in the world and that a number of shady dealings occur. . The question of advocacy research is one of objectivity: can the researcher search out the truth when the legal client wishes to support its position at a trial? The ethical question stems from a conflict between legal ethics and research ethics. ADVOCACY RESEARCH Advocacy research is business research undertaken to support a specific claim in a legal action. The research design is often developed based on what a judge or jury will consider appropriate rather than what the company deems appropriate. Researchers doing advocacy research do not necessarily bias results intentionally. researchers are generally ethical people. like most business people. some of which may be of questionable ethics. In advocacy research. Under some circumstances even good researchers take shortcuts. the court’s opinion of the value of the research may be based exclusively on sampling design and validity processes. However. A user of research should not knowingly disseminate conclusions from a given research project or service that are inconsistent with the data or not warranted by them.
Understand the background of the problem 3. Determine the relevant variables 6. If the submerged portions of the problem are omitted from the problem definition. THE IMPORTANCE OF PROPER PROBLEM DEFINITION Business research is conducted to help solve managerial problems. exploratory research—by illuminating the nature of the business opportunity or problem—helps managers clarify their objectives and decisions. then false conclusions may be drawn from the investigation. They are: 1. Or.Chapter 6 Problem Definition and the Research Proposal I. V. Formal qualitative research should not begin until the problem has been clearly defined. at the other extreme. Problem definition indicates a specific business decision area that will be clarified by answering some research questions. the research design. THE NATURE OF BUSINESS PROBLEMS A decision maker’s degree of uncertainty influences decisions about the type of research that will be conducted. . and subsequently from the research design. The background of the problem is vital. Most business decision face situations falling in-between these two extremes. Isolate and identify the problem not the symptoms 4. However. If quantitative research is conducted before the researchers understand exactly what is important. A situation analysis is the logical first step in defining the problem. Both the research investigator and the manager requesting the research should attempt to have a clear understanding of the purpose of undertaking the research. is neither visible to nor understood by the business managers. A. managers may have only vague insights about a complex situation. a manager or researcher may describe a decision-making situation as absolute ambiguity. when a problem or opportunity is discovered. This analysis involves the informal gathering of background information to familiarize researchers or managers with the decision area. The objectives are vague and the alternatives are difficult to define. The nature of the problem to be solved is unclear. 2. This is by far the most difficult decision situation. It is extremely important to define the business problem carefully because such definition will determine the purpose of the research and. then the decision based on such research may be less than optimal. THE PROCESS OF DEFINING THE PROBLEM The process of defining the problem involves several interrelated steps. State the research questions (Hypotheses) and research objectives IV. UNDERSTAND THE BACKGROUND OF THE PROBLEM. Determine the unit of analysis 5. ASCERTAIN THE DECISION MAKER’S OBJECTIVES The research investigation must attempt to satisfy the decision maker’s objectives. Ascertain the decision maker’s objectives. Often. Exploratory research techniques have been developed to help formulate clear definitions of the problem (see Chapter 7). II. The iceberg principle. III. ultimately. The dangerous part of any business problem. A business manager may be completely certain about the situation s/he is facing. decision makers are not able to articulate precise research objectives. like the submerged part of an iceberg. Sometimes.
VI. ISOLATE AND IDENTIFY THE PROBLEM, NOT THE SYMPTOMS. Anticipating the many influences and dimensions of a problem is impossible for any researcher or executive. Certain occurrences that appear to be the problem may only be symptoms of a deeper problem. Executive judgment and creativity must be exercised in identifying a problem. VII. WHAT IS THE UNIT OF ANALYSIS? The researcher must specify the unit of analysis. Will the individual consumer be the source of information or will it be the parent-child dyad? Industries, organizations, departments, or individuals, may be the focus for data collection and analysis. Many problems can be investigated at more than one level of analysis. VIII. WHAT ARE THE RELEVANT VARIABLES? One aspect of problem definition is identification of the key variables. A variable is a quality that can exhibit differences in value, usually magnitude or strength. In statistical analysis, a variable is identified by a symbol such as X. A category or classifactory variable has a limited number of distinct variables (e.g., sex—male or female). A continuous variable may encompass an infinite range of numbers (e.g., sales volume). Managers and researchers must be careful to include all relevant variables that must be studied in order to be able to answer the managerial problem. Irrelevant variables should not be included. In causal research, a dependent variable is a criterion or variable that is expected to be predicted or explained. An independent variable is a variable that is expected to influence the dependent variable. IX.STATE THE RESEARCH QUESTIONS AND RESEARCH OBJECTIVES The research question is the researcher’s translation of the business problem into a specific need for inquiry.
A. Clarity in Research Questions and Hypotheses Research questions should be specific, clear, and accompanied by a well-formulated hypothesis. A hypothesis is an unproven proposition or possible solution to a problem. In its simplest form, a hypothesis is a guess. Problems and hypotheses are similar; both state relationships, but, whereas problems are interrogative, hypotheses are declarative and more specifically related to the research operations and testing. Hypotheses are statements that can be empirically tested. A formal statement of hypothesis can force researchers to be clear about what they expect to find through their study. The hypothesis can raise critical questions about the data that will be required in the analysis stage. When evaluating a hypothesis, researchers should make sure that the information collected will be useful in decision making.
B. Decision-oriented research objectives The research objective is the researcher’s version of the business problem. The research objective is derived from the problem definition and it explains the purpose of the research in measurable terms, as well as defining what standards the research should accomplish. Such objectives help ensure that the research projects will be manageable in size. In some instances the business problems and the project’s research objectives are identical. The objectives must, however, specify the information needed to make a decision. Statements about the required precision may be necessary to clearly communicate exactly what information is required.
It is useful if the research objective is a managerial action standard. That is, if the criterion being measured turns out to be X, then management will do A; if it is Y, then management will do B. This leaves no uncertainty concerning the decision to be made once the research is finished. The number of research objectives should be limited to a manageable number so that each one can be addressed fully. Exhibit 6.5 shows how statements of the business problem influence the research objective. The specific objectives, in turn, are the basis for the research design. X. HOW MUCH TIME SHOULD BE SPENT DEFINING THE PROBLEM? It is impractical to search for every conceivable cause and minor influence of a problem. The importance of the recognized problem will usually dictate what is a reasonable amount of time and money for determining which possible explanations are most likely. XI.THE RESEARCH PROPOSAL The research proposal is a written statement of the research design—it explains the purpose of the study, defines the problem, outlines the research methodology, details the procedures to be followed, and states all costs and deadlines. The proposal should be precise, specific, and concrete. All ambiguities about why and how the research will be conducted must be “ironed out” before the proposal is complete. The research proposal can act as a communication tool. It allows managers to evaluate the proposed research design and determine if alterations are necessary. The proposal should be detailed enough that managers are clear about exactly how the information will be obtained. Misstatements and faulty communication may occur if the two parties rely on each other’s memory of what occurred at a planning meeting; therefore, it is wise to write down all proposals. Such a written proposal eliminates many problems that may arise and acts as a record of the researcher’s obligation. In the case of an outside consultant, the written proposal serves as a bid to offer a specific service; a company can then judge the relative quality of alternative research suppliers. XII. ANTICIPATING OUTCOMES By anticipating the outcomes of a research study, possibly through the use of a dummy table (a table filled by the researcher with fictitious data), managers may gain a better understanding of what the actual outcome is liable to be. These tables help clarify what the findings of the research will be, and if these findings will meet the needs of the researcher.
Chapter 7 Exploratory Research
I. EXPLORATORY RESEARCH: WHAT IT IS AND WHAT IT IS NOT Exploratory research helps ensure that a rigorous and conclusive study will not begin with an inadequate understanding of the nature of the business problem. Most exploratory research designs provide qualitative data which provides greater understanding of a concept. In contrast, quantitative data provides precise measurement. Exploratory research may be a single research investigation or it may be a series of informal studies; both methods provide background information. Researchers must be creative in the choice of information sources. They should explore all appropriate inexpensive sources before embarking on expensive research of their own. However, they should still be systematic and careful at all times. II. WHY CONDUCT EXPLORATORY RESEARCH? There are three purposes for conducting exploratory research; all three are interrelated: A. Diagnosing a situation: Exploratory research helps diagnose the dimensions of problems so that successive research projects will be on target. B. Screening alternatives: When several opportunities arise and budgets restrict the use of all possible options, exploratory research may be utilized to determine the best alternatives. Certain evaluative information can be obtained through exploratory research. Concept testing is a frequent reason for conducting exploratory research. Concept testing refers to those research procedures that test some sort of stimulus as a proxy for a new, revised, or remarketed product or service. Generally, consumers are presented with an idea and asked if they like it, would use it, etc. Concept testing is a means of evaluating ideas by providing a feel for the merits of the idea prior to the commitment of any research and development, marketing, etc. Concept testing portrays the functions, uses, and possible situations for the proposed product. C. Discovering new ideas: Uncovering consumer needs is a great potential source of ideas. Exploratory research is often used to generate new product ideas, ideas for advertising copy, etc. III. CATEGORIES OF EXPLORATORY RESEARCH The purpose, rather than the technique, of the research determines whether a study is exploratory, descriptive, or causal. A manager may choose from three general categories of exploratory research: A. Experience surveys: Concepts may be discussed with top executives and knowledgeable managers who have had personal experience in the field being researched. This constitutes an informal experience survey. Such a study may be conducted by the business manager rather than the research department. On the other hand, an experience survey may be a small number of interviews with experienced people who have been carefully selected from outside the organization. The purpose of such a study is to help formulate the problem and clarify concepts rather than to develop conclusive evidence. B. Secondary data analysis: A quick and economical source of background information is trade literature in the public library. Searching through such material is exploratory research with secondary data; research rarely begins without such an analysis. An informal situation analysis using secondary data and experience surveys can be conducted by business managers. Should the project need further clarification, a research specialist can conduct a pilot study. C. Case study method: The purpose of a case study is to obtain information from one, or a few, situations similar to the researcher's situation. A case study has no set procedures, but often requires the cooperation of the party whose history is being studied. However, this freedom to research makes the success of the case study highly dependent on the ability of the researcher. As with all exploratory research, the results of a case study should be seen as tentative. D. Pilot studies: The term “pilot studies” is used as a collective to group together a number of diverse research techniques all of which are conducted on a small scale. Thus, a pilot study is a research project which generates primary data from consumers, or other subjects of ultimate concern. There are four major categories of pilot studies:
The discussion may start out general. One individual often triggers a chain of responses from the other participants. no matter how carefully they are selected. and taxis than they do watching the group session. An effective focus group moderator prepares a discussion guide to help ensure that the focus group will cover all topics of interest. being conducted at one time. This avoids one or two members intimidating the others. and yet. This affords closer scrutiny and allows the researchers to check for consistency in the interpretations. If the focus group is being conducted “out of town. Streaming media is a method of making audio. With traditional focus groups managers and creative personnel often watch the moderator lead the group from behind one-way mirrors. i) Structure: the moderator. A number of focus groups use videoconferencing. The discussion guide consists of written prefatory remarks to inform the group about the nature of the focus group and an outline of topics/questions that will be addressed in the group session. The moderator's job is to develop a rapport with the group and to promote interaction among its members. is a small enough group that adequate participation is allowed. Researchers who wish to collect information from different groups should conduct several different focus groups. b) Serendipity: an idea may drop out of the blue. (4) it is inexpensive. The sessions should be as relaxed and natural as possible. the number of companies using these systems to conduct focus groups has increased. The moderator's role is to introduce a topic and to encourage the group to discuss it among themselves. video and other multimedia available in real-time over the Internet or corporate intranets. The offsite manger uses RealPlayer or Microsoft Media Player to view focus groups on their own computer rather than at a remote location. As the ability to communicate via telecommunications and videoconferencing links have improved. not because a question requires an answer. but the moderator should be able to focus it on specific topics. The videoconferencing industry has improved in quality and grown dramatically in recent years. The group typically consists of six to ten participants and a moderator. and affords the group the opportunity to develop such an idea to its full significance.” the executive personnel usually have to spend more time in airplanes. (3) it is easy to execute and very flexible. insights and ideas than will the cumulation of separately secured responses. in effect. (2) it is relatively fast. Focus group interviews: These interviews are free-flowing interviews with a small group of people. With videoconferenced focus groups. they can stay home. One disadvantage is that a small group of people. will not be representative. can control the topics the group discusses. The ideal size for a focus group is six to ten relatively homogeneous people. being one of the group. This new technology for digital media delivery now allows researchers to “broadcast” focus groups that can be viewed online. Homogeneous groups avoid confusion which might occur if there were too many differing viewpoints. g) Specialization: the group interview allows the use of a more highly trained moderator because there are certain economies of scale when a large number of people are “interviewed” simultaneously. They have a flexible format and can discuss anything from brand to a product itself. c) Snowballing: a bandwagon effect occurs. . There are four primary advantages of the focus group: (1) it allows people to discuss their true feelings and convictions. f) Spontaneity: people speak only when they have definite feelings about a subject. h) Scientific scrutiny: the group interview can be taped or even videoed for observation. hotels. d) Stimulation: respondents want to express their ideas and expose their opinions as the general level of excitement over the topic increases. j) Speed: a number of interviews are. Specific advantages of focus group interviews have to be categorized as follows: a) Synergism: the combined effort of the group will produce a wider range of information. with no download wait and no file to take up space on a viewers hard disk. Except for the quality of the video when there are bandwidth problems the experience is similar to videoconferencing. e) Security: the participants are more likely to be candid because they soon realize that the things said are not being identified with any one individual.1.
b) Sentence completion method: This technique is also based on the assumption of free association. It is an interactive media because the user clicks a command and the computer responds. they are more likely to express their true feelings. Projective techniques: Individuals may be more likely to give a true answer if the question is disguised. 3. or what a third person thinks of a product. such as a new easy-opening packaging design. Word association should reveal each individual's true feelings about the subject. transcripts of verbatim responses are available immediately afterward the group session.) to interpret how people are reacting. Online groups can be quick and cost efficiency. interest. and asked to respond with the first word that comes to mind. Role playing is a dynamic reenactment of the third-person technique in a . the respondents are allowed to express emotions and opinions that would normally be hidden from others and even hidden from themselves. Participants in these chat room feel their anonymity is very secure. the moderator’s ability to probe and ask additional questions on the spot is reduced in online focus groups. The group participants either keyboard their remarks during a chat room format or when they are alone at their computers. Also. Because participants do not have to be together in the same room at a research facility. c) Third-person technique and role playing: Providing a “mask” is the basic idea behind the third-person technique. This can be a major advantage for a company investigating sensitive or embarrassing issues. Common techniques are as follows: a) Word association: The subject is presented with a list of words. Research companies often set up a private chat room on their company Web sites for focus group interviews. one at a time. The focus group technique has two shortcomings: (1) Without an experienced moderator. Respondents are asked why a third person does what he or she does. etc. or taste experiences cannot be performed online. such as sexual harassment. the intention of the study is more apparent. A projective technique is an indirect means of questioning that enables respondents to “project their beliefs onto a third party.” Thus. Many online focus groups using the chat room format arrange for a sample of participants to be online at the same time for about typically 60 to 90 minutes. however. the number of participants in online focus groups can be much larger than traditional focus groups. This can be a major advantage for company investigating sensitive or embarrassing issues. especially those in which participants are not simultaneously involved. Interactive Media and online Focus Group: When a person uses the Internet. Respondents are required to complete a number of partial sentences with the first word or phrase that comes to mind. A disadvantage of online focus groups is that the researcher cannot see how people are reacting. he or she interacts with a computer. Often they will make statements or ask questions they would never address under other circumstances. they will make statements or ask questions they would never address under other circumstances. Answers tend to be more complete than in word association. The term online focus group refers to qualitative research where a group of individuals provide unstructured comments by keyboarding their remarks into a computer connected to the Internet. Because respondents enter their comments into the computer. a self-appointed leader will dominate the session resulting in an abnormal “halo effect” on the interview (2) there may be sampling problems. Often. group synergy and snowballing of ideas can suffer. A problem with online focus groups is that the moderator cannot see body language and facial expressions (bewilderment. 2. If respondents are presented with unstructured and ambiguous stimuli and are allowed considerable freedom to respond. excitement. Participants in these chat rooms feel their anonymity is very secure. The use of the Internet for qualitative exploratory research is growing rapidly. Interpreting the results is difficult. Research companies often set up a private chat room on their company web sites for focus group interviews. The respondent can transfer his attitudes onto the third person. because there is less interaction between participants. hesitations). the researcher should avoid subjective interpretations and should consider both what the subject said and did not say (e. Research that requires tactile touch.Research using Internet for on-line qualitative exploratory research is growing rapidly.. However. Both verbal and non-verbal responses are recorded.g.
when everything looks positive in the exploratory stage. However. then no further research should probably be conducted. the subject matter is usually disguised. Construction techniques request that the consumer draw a picture. or picture frustration version of TAT. uses a cartoon drawing in which the respondent suggests dialogue that the cartoon characters might say. When a major commitment of resources is involved. and well-defined problem.given situation. Because of certain problems. A WARNING Exploratory research techniques have their limitations. The researcher asks many questions and probes for additional elaboration after the subject answers. Cartoon tests. conclusive research. Depth interviews have lost their popularity recently because they are time-consuming and expensive as they require the services of a skilled interviewer. business managers should determine the benefits of further information versus the cost of additional research. interesting. the researcher should proceed with caution because there is a possibility that a potentially good idea could be rejected because of unfavorable results at the exploratory stage. . d) Thematic apperception test (TAT): This test consists of a series of pictures in which consumers and products are the center of attention. Depth interviews: Depth interviews are similar to the client interviews of a clinical psychiatrist. In other situations. there is a temptation to market the product without further research. it is often well worth conducting a quantitative study. exploratory findings should be treated as preliminary. they cannot take the place of quantitative. IV. construct a collage. If the findings of exploratory research are very negative. Theses (“thematic”) are elicited on the basis of the perceptual-interpretive (“apperception”) use of the pictures. This technique requires the subject to act out someone else's behavior in a particular setting. The picture should present a familiar. and the interpretation of their results is judgmental—thus. The researcher then analyses the content of the stories that the subjects relate. Most of them are qualitative. but the solution should be ambiguous. 4. The major benefit of exploratory research is that it generates insights and clarifies the business problems for testing in future research. such as interpreter bias or sample size. or write a short story to express their perceptions or feelings. In this situation. The investigator asks the subject what is happening in the picture and what the people might do next.
1 illustrates a series of questions that should be asked to evaluate secondary data. III. Certain sources of data. For example. Model building involves specifying relationships between two or more variables. DISADVANTAGES OF SECONDARY DATA Secondary data may not always be available . decision-makers using simple models.S. Managers often estimate market potential using secondary data. Many of the activities normally associated with primary data collection are eliminated. and do not require access to respondents or subjects. Data are simply facts. it may not be adequate to meet the needs of the researcher. In this case data conversion may be necessary. ADVANTAGES OF SECONDARY DATA Acquiring secondary data is always faster and less expensive than acquiring primary data.TYPICAL OBJECTIVES FOR SECONDARY DATA RESEARCH DESIGNS There are two general categories of research objectives: fact finding and model building. Another shortcoming of secondary data is that it may be inaccurate or even biased to support the vested interest of the source. The reputation of the organization gathering the data should be considered and the research design should be critically assessed. Outdated information: Secondary data must be timely in order to predict the future. A typical secondary research objective for a study might be to uncover all available information about consumption patterns for a particular product category or to identify demographic trends that affect an industry. if one source of data . as a general objective for secondary research. often the primary study may dictate that the data be summarized. have already been assembled. The simplest form of secondary data research is fact finding. IV.Chapter 8 Secondary Data I. however. The three most common reasons that the information may not be adequate are: A. C. WHAT ARE SECONDARY DATA? Secondary data are gathered and recorded by “someone else” prior to (and for purposes other than) the current needs of the researcher. Even if it is available. Data conversion is the process of changing the original form of the data to a format suitable to achieve the research objective. ones that everyone can readily understand. A common secondary data study designed to find facts might be a market tracking study. Exhibit 8. Cross-checks of data from multiple sources — that is. Model building can involve the development of descriptive or predictive equations. often find these models superior to complex models that are difficult to comprehend. privacy laws may prevent its access. but not directly comparable to. II. model building need not be a complicated mathematical process. Secondary data are usually historical in nature. Variation in definition of terms: Researchers frequently encounter secondary data that reports on a population of interest that is similar to. rounded. is more complicated than simple fact finding. comparison of the data from one organization with the data from another source — should be made. B. Whereas information refers to a body of facts in a format suitable for decision making or in a context that defines relationships between pieces of data. In fact. are more prone to acceptance. for example the U. For example. Model building. The researcher may estimate market potential by converting different types of data that are available from two or more sources. the chapter began by describing the sports market in the United States. Market tracking refers to the observation and analysis of trends in industry volume and brand share over time.for example. Differing units of measurement: Differing units of measurement may cause problems if they are not identical to the researcher’s needs. the population of interest to the researcher. government. or reported in such a way that it is not useful to the secondary research needs.
such as catalog retailers. Most organizations routinely gather. However. accounting departments continually gather data. d) Attitude and public opinion research: Specialized syndicated services report the findings of attitude research and opinion polls. e) Stock market sources: Numerous firms sell information on market and individual stocks. State. Managers need information about the future. guides.000 electrical contractors then it may be estimated that 8. the use of data mining to detect sequence patterns. Libraries stock many bibliographies. Professional journals and commercial business periodicals can be especially valuable sources of data. The media like to show that their vehicles for advertising are viewed or heard by the advertising target market. It is a broad term that applies to many different forms of analysis. examples of which follow: Market share data: Market tracking refers to the observation and analysis of trends in industry volume and brand share over time. The term data mining refers to the use of powerful computers. c) Demographic and census updates. or generated by another entity. Books and periodicals: Books and periodicals provide a wealth of information. b) Scanner data: Market tracking through optical character recognition such as the universal product code and other optical scanners provides a wealth of accurate and rapid product and brand sales information collectively known as scanner data. Aggregating or desegregating internal data is a frequent form of internal research. Such information is generally free of charge and can be useful. Media sources: Information on a broad range of subjects is available from broadcast and print media. neural networks are a data-mining form of artificial intelligence in which a computer is programmed to mimic the way in which human brains process information. record.. Accurate sales forecasts. 2. it should be given careful evaluation as it often covers limited aspects of a topic. county. recorded. the Census of Population) can be counted on for accuracy and quality of investigation. Market basket analysis is a form of data mining that analyze anonymous point-of-sale transaction logs to identify coinciding purchases or relationships between products purchased and other retail shopping information. 4. These firms provide diverse types of data. is a popular application among direct marketers. 3. For example. CLASSIFICATION OF SECONDARY DATA Internal and proprietary sources: Data that are external to the organization refers to data created. directories. They need to know what company sales will be next year and in future time periods. Commercial sources: Numerous firms specialize in selling information.indicates that 10 percent of all electrical contractors intend to buy a drill and another source indicates that there are 80.000 drills will be sold to electrical contractors. For example. V. Sequence discover. and store internal data for solving future problems. External sources of data can be categorized as follows: 1. When the identity of the customer who makes repeated purchases from the same organization is known. By exhausting all sources within the company. especially when products are in mature. a) aggregate .g. Government sources: Government agencies produce a prolific amount of data. frequently are the result of secondary data research that identify trends and extrapolates past performance into the future. to dig through volumes of data to discover patterns about an organization’s customers and products. researchers can avoid duplicating another department’s data collection and research efforts. Federal government data (e. and indexes. Sales forecasting is the process of predicting sales totals over a specific future time period. stable markets. and local government data is often more current and structured to meet local needs than federal data. Business researchers often use internal company sales records to project sales. an analysis can be made of sequences of purchases.
The CIA Factbook and the National Trade Data Bank are especially useful. INVESTIGATING GLOBAL MARKETS: AN EXAMPLE The United States federal government offers a wealth of data about foreign countries. Today however. from a trade association. or generated by another entity. This section describes the National Trade Data Bank (NTDB). Intermediate distributors or vendors may disseminate the information to libraries. IX. This is especially likely when official statistics are adjusted for the political purposes of foreign governments. APPENDIX 8B Database Searching and Retrieving with Computers Recent developments in information technology and computer software have had a major impact on the search and retrieval and use of data. VIII. Finally.EXTERNAL DATA THE DISTRIBUTION SYSTEM Data that are external to the organization refers to data created. international researchers should be watchful for certain pitfalls that frequently are associated with foreign data and cross-cultural research. The process is much faster than using a card catalog and can also be more exhaustive if several bibliographic databases are used. Traditionally this information has been in published form. trade associations.S. there may be the simple problem that data is not available in certain countries. Second. various countries use different definitions and accounting/recording practices for many economic concepts.6 illustrates some examples of the traditional ways of distributing information and some examples of modern ways of distributing information. FOCUS ON GLOBAL RESEARCH As business has become more global so has the secondary data industry. the government puts the information in the census report into a formal government report. Some users may prefer to go directly to the initiator to obtain the information and others will use an intermediary to obtain information. . and other organizations create/produce information. The government. including those that offer information about countries around the world. First. newspapers and journals. However. or from a governmental agency. The ProQuest/ABI INFORM system is one such system. computerized data archives and electronic data interchanges make external data as accessible as internal data once were. VII. The initiators of the information produce information that may later be submitted for publication.VI. Exhibit 8. to illustrate what is available. perhaps available at the public library. APPENDIX 8A SELECTED SECONDARY SOURCES X. For example. the accuracy of some data may be called into question. Secondary data complied outside the United States have the same limitations as secondary data available in the United States. The appendix to chapter 8 lists a large number of secondary data sources. Both can be accessed using the Internet. In a modern library. while standardized terms for economic data may be used. patrons command a computer to do the searching by using CD-ROM system or the Internet. The United States government and other organizations compile of interest to international managers. recorded. government’s most comprehensive source of world trade data. the U.
or face-to-face. A sample bias exists when the results of a sample show a persistent tendency to deviate in one direction from the true value of the population parameter. The more formal term. but even with technically proper random probability samples. If only those who responded are included in the survey then nonresponse error will occur. For example. but may also occur in telephone and personal surveys in the form of no contacts (not-at-homes) or refusals. for information using either written or verbal questioning. Systematic error: Systematic errors result from some imperfect research design. B. Thus. on the telephone. and accurate means of assessing information about the population. b) Response bias: Response bias occurs when respondents tend to answer in a certain direction. only those who feel strongly about the subject matter will respond. or from a mistake in the execution of the research. The number of no contacts has been increasing because of the proliferation of answering machines and growing usage of Caller ID to screen telephone calls. ERRORS IN SURVEY RESEARCH Exhibit 9. statistical errors will occur because of chance variation. call-backs can be made on the not-athomes. Most survey research is descriptive research which attempts to identify and explain a particular business activity. Questionnaires or interviews collect data through the mail. when conducted properly. these statistical problems are unavoidable. Selfselection may also occur in self-administered questionnaires. Advantages of surveys: Surveys are quite flexible and. respondents. A. Random sampling error: Most surveys try to portray a representative cross section of a particular target population. Surveys can also be poorly conducted and certain errors can occur to render such a survey useless. emphasizes that the purpose of contacting respondents is to obtain a representative sample of the target population. causing an over-representation of extreme positions. However. a) Nonresponse error: To utilize the results of a survey. a survey is defined as a method of primary data collection based on communication with a representative sample of individuals. few gather only a single type of factual information. the researcher must be sure that those who did not respond to the questionnaire were representative of those who did not.1 outlines the various forms of survey error that can affect the accuracy of a survey. They provide a quick. These errors are also called non-sampling errors. Additional efforts should be made to obtain data from any underrepresented segments of the population. 1. efficient. For example. in new-product development the qualitative objective of a survey is often to test and refine new-product concepts. Business surveys typically have multiple objectives. Comparing demographics of the sample with the demographics of the target population is one means of inspecting for possible biases. B. . The two general categories of systematic error are respondent error and administrative error. THE NATURE OF SURVEYS Surveys require asking people. inexpensive. sample survey. This bias may be caused by an intentional or inadvertent falsification or by a misrepresentation of the respondent’s answer. Nonrespondents are most common in mail surveys. II. Respondent error: If the respondents do not cooperate or do not give truthful answers then two types of error may occur. in this situation. Survey objectives: Type of Information Gathered: The type of information gathered depends on a survey’s objectives. Although surveys are often conducted to quantify certain factual information. Without increasing sample size. random sampling errors can be estimated (see Chapters 13 and 14). are extremely valuable to managers. certain aspects of surveys may also be qualitative. A.Chapter 9 Survey Research: An Overview I.
to avoid embarrassment. b) Sample selection error: This type of error is a systematic error that results in an unrepresentative sample because of an error in either the sample design or execution of the sampling procedure. neglect. These categories overlap and are by no means mutually exclusive. c) Types of response bias: There are five specific categories of response bias. question ambiguity or content. (3) Interviewer bias: If an interviewer’s presence influences respondents to give untrue or modified answers. For example. It may be that the interviewees preferred to be viewed as average and they will alter their responses accordingly. Many researchers have found it useful to use some standard of comparison in order to understand how much error can be expected. 2. to “please” the interviewer. “yea sayers” who accept all statements they are asked about. Time-lapse may lead to best-guess answers. Administrative error: The results of improper administration or execution of the research task are examples of administrative error. d) Interviewer cheating: To avoid possible cheating. Respondents may wish to appear wealthy or intelligent. omission. IV. it is wise to inform the interviewers that a small sample of respondents will be back to confirm that the interview actually took place. to conceal personal information. the survey will contain interviewer bias. c) Interviewer error: Interviewers may record an answer incorrectly or selective perception may influence them to record data supportive of their own attitudes. (5) Social desirability bias: This may occur consciously or subconsciously. Answers to questions that seek factual information or matters of public knowledge are usually quite accurate.(1) Deliberate falsification: People may misrepresent answers in order to appear intelligent. There are four types of administrative error: a) Data processing error: The accuracy of the data processed by computer depends on correct data entry and programming.WHAT CAN BE DONE TO REDUCE SURVEY ERROR? . (2) Unconscious misrepresentation: Response bias can arise from question format. one cable TV company knocks down the number of people saying that they intend to purchase the service by a “ballpark 10 percent” because previous experience has indicated a 10 percent upward bias on the intention questions. RULE-OF-THUMB ESTIMATES FOR SYSTEMATIC ERROR Sampling error may be estimated using certain statistical tools. For example. (4) Auspices bias: The answers to a survey may be deliberately or unintentionally misrepresented because the respondent is influenced by the organization conducting the survey. but the interviewer’s presence may increase a respondent’s tendency toward an inaccurate response to a sensitive question in an attempt by the respondent to gain prestige in the interviewer’s mind. but ways to estimate systematic error are less precise. (2) Extremity bias: Some individuals tend to use extremes when responding to questions which may cause extremity bias. etc. III. Mistakes can be avoided if verification procedures are employed at each processing stage. or some other blunder. Such errors are inadvertently caused by confusion. or they may try to give the “right” answer or the socially acceptable answer. (1) Acquiescence bias: This is a response bias caused by a respondent’s tendency to concur with a particular position.
CLASSIFYING SURVEY RESEARCH METHODS Surveys can be classified in three ways. special promotions. Method of communication: Surveys can be classified according to the method of communication. The panelists record their purchasing habits in a diary for a set period of time. but the researcher can never be sure that the changes in the variable being measured are not actually due to having different people in the sample. Cross-sectional study: This is the most common type of study in which the data is collected at a single point in time. whereas unstructured questions tend to be open-ended which allows the respondent considerable freedom in responding. behavior habits affected by changes in price. Longitudinal study: In longitudinal studies. telephone. Such studies can also be called tracking studies because they are designed to compare aggregate trends and identify changes. Structured and disguised questions: A structured question limits the number of responses available. A.Chapters to follow discuss various techniques for reducing bias in business research. Panels are generally expensive and. mail. Panel Study: This is a longitudinal study which includes gathering data from the same sample over time. Questions can be categorized according to their degree of structure and disguise. C. various segments of the population are sampled so that relationships among variables may be investigated by cross-tabulation. Having two or three different sample groups avoids response bias which might normally result from prior interview. The researcher can also disguise the questions which is particularly advisable if the subject matter is of a threatening nature. Longitudinal studies which involve two or more samples at different times are called cohort studies because similar people are expected to be in each sample over time. V. it is not always easy to categorize the surveys as the categories are not clear-cut and most surveys are a hybrid of structured and unstructured questions. Classifying surveys on a temporal basis: 1. or other aspects of business strategies. . This helps in the selection of the appropriate communication medium for conducting the survey. However. Other questions do not require disguising as it is assumed that the respondent is willing to reveal the information. 2. respondents are questioned at different points in time so that changes occurring can be observed over time. In such a study. B. are usually managed by contractors which specialize in maintaining consumer panels. Such panels enable the investigator to keep track of repeat purchases. thus. or personal interviews.
PERSONAL INTERVIEWS Personal interviews are a direct communication between business and consumers in which interviewers ask respondents questions in face-to-face situations. and for how long they will view or hear it. from our perspective. It only means this type of survey is not as flexible as surveys using interactive communication media. I. They are actively involved in a two-way communication when electronic interactive media are utilized. touch-tone telephone systems. MEDIA USED TO COMMUNICATE WITH RESPONDENTS Interviewers communicated with respondents face-to-face or over the telephone or respondents filled-out a selfadministered paper questionnaire. and other forms of digital technology. provides a prominent example of the new electronic interactive media. B. However. respondents are not passive audience members. A. then a personal interview may be the only alternative as they can last up to one and a half hours. digital technology is having a profound impact on society in general and business research in particular. Length of interview: If the research objective requires an extremely lengthy questionnaire. Hence. we will classify self-administered questionnaires printed on paper as non-interactive. Consumers determine what information they will be exposed to. electronic interactive media are controlled by the users themselves. and to provide the opportunity for interaction by using digital technology. and mail. That is. a telephone interview should not last longer than 10 to 15 minutes and a mail survey should not be more than six pages long. When we think of the traditional role of interviewing. human interaction takes place. C. or self-administered questionnaires. to personalize individual messages. Electronic interactive media also include CD-ROM and DVD materials. No other human need be present. Each method has its merits and its shortcomings. interactive kiosks in stores.Chapter 10 Survey Research: Basic Methods of Communication with Respondents Survey data are obtained when individuals respond to questions asked by interviewers (interviews) or when the individual responds to questions that he has read (questionnaires). Probing complex answers: An important characteristic of personal interviews is the opportunity to probe that the two-way communication provides. The messages are directed at a particular individual (or a small group) who has the opportunity to interact with another human being. a medium that is radically altering many organizations' research strategies. These media remain popular with business researchers. Electronic interactive media allow marketers to reach a large audience. II. most people envision two people engaged in a face-to-face dialog or a conversation on the telephone. There are three basic survey methods of gathering data and communicating with individuals—personal interviews. telephone interviews. In the context of surveys. In contrast. This versatile and flexible method is a two-way conversation between an interviewer and the respondent. which was typically distributed by mail. Probing allows interviews some flexibility. Human interactive media are a personal form of communication. A. B. clarification or expansion of answers to standardized questions. Its greatest impact is in the creation of a new form of communications media. The opportunities for feedback: One of the main advantages of the personal interview is the opportunity it provides for two-way feedback. The Internet. This does not mean they are without merit. Human Interactive Media and Electronic Interactive Media When two people engage in a conversation. NonInteractive Media The traditional questionnaire received by mail and completed by the respondent does not allow a dialog or an exchange of information providing immediate feedback. . which affords more understanding for both the interviewer and the respondent. To a large extent.
For example. For example. the lack of face-to-face contact can also constitute a limitation as nonverbal cues are not detected. I.D. The mall interviews also have the advantage of enabling the interviewer to use visual aids or conduct taste tests. they can greatly help reduce non-response error. Door-to-door interviews: Door-to-door interviews provide a more representative sample of the population than do mail surveys. However. the nonrespondents may represent a certain demographic group and to exclude such a group from the sample could greatly bias the results. children of bike-riding age. This allows companies to hire a staff of professional interviewers and to supervise and control the quality of interviewing more effectively. G. Personal interviews can be more expensive than other basic methods because of the length of the interviews and the number of respondents. D. (4) interviewer cheating. Callbacks: Callbacks are attempts to recontact individuals selected for the sample who did not respond to the first interview. B. although mail surveys are the most anonymous. For example. TELEPHONE INTERVIEWS Evidence suggests that telephone interviews are now the prime method of survey research. but they do eliminate the possibility of bias results. Questions have to be very structured. However. Intercept interviews in malls and other high-traffic areas: These interviews are conducted in a shopping mall or other high-traffic areas. (3) Differential interview techniques may be a source of bias. more people are likely to participate at home than in a shopping mall. Cost: Travel time and costs of travel associated with personal interviews are eliminated. Consider. (2) characteristics of the interviewer may influence the response. E. F. People are also more willing to discuss more personal topics on the phone than in a face-to-face interview. they may cause an underrepresentation in some groups. Props and visual aids: Interviewing respondents face-to-face allows the investigator to show the respondent certain visual aids. . and (5) cost. A.Absence of face-to-face contact: Embarrassing or confidential questions may be answered more willingly. J. H. Central location interviewing: All the interviews can be conducted from the same location using WATS lines. specific group. Research supports the conclusion that telephone interviews can provide a sample representative of the general population and that the data they provide is quality data. III. K. They are low cost and a large number of interviews can be gathered quickly. the investigator would be able to show the subject a new product sample. The locality and the ability of the interviewer also influence the participation rate. Callbacks are expensive. the refusal rate is high and the samples may not be representative of the total population. For example. Anonymity of respondent: Respondents do not remain anonymous and may be reluctant to provide confidential information. However. Complete questionnaires E. Mall intercept interviews can be very useful if the target group is a small. but it may be the only choice if a lot if the residents do not have phones or have unlisted phone numbers. Disadvantages of personal interviews: (1) Lack of anonymity for the interviewee. for example. These systems can also use an automatic sample selection and an automatic callback schedule can be created to aid the interviewer. High participation: The presence of an interviewer increases the percentage of people willing to complete the interview. Computer-assisted telephone interviewing: This technique allows telephone interviews to be directly entered into the computer via a keyboard operated by the interviewer. Door-to-door interviews can also solve the problem of nonresponse which may occur in telephone interviews. Speed: Hundreds of interviews can be collected literally overnight. how hard it would be to obtain door-to-door interviews in a high-rise flat or in the middle of Harlem. C. making telephone interviews inexpensive by comparison.
V. No matter how the self-administered questionnaires are distributed to the members of the sample. Standardized questions: In mail surveys the simpler the question. unless the interviewee is highly interested in the subject matter.g.F. Cooperation: People are more likely to answer questions over a phone than to let a stranger into their house. A. thus. they are not cheap as most include a follow-up mailing. G. there is more chance of accurate information and carefully thought out answers. selective perception may occur. However. semantic differential scales) hard. Once the questionnaires are mailed it is difficult to change the format or explain the questions. They have the following characteristics. F. These surveys rely on the efficiency of the written word rather than the interviewer. this may still be quicker than conducting personal interviews in the case of a national study. H. Also. interviewers may be somewhat reluctant to conduct face-to-face interviews in certain neighborhoods. Cost: Relatively low in cost compared to other basic survey methods. to administer. I. Lack of visual medium: The lack of visual aids makes some measurement instruments (e. C. unclear or too complex questionnaires get thrown away. The lapsed time between the first mailing and the cutoff date is usually six to eight weeks. the refusal rate tends to be higher in RDD than when using listed telephone numbers. H. Response rates: The major limitations of mail questionnaires relate to response rates. Approximately 10-15 minutes long is a good rule of thumb. Geographic flexibility: Mail questionnaires can reach a geographically dispersed sample simultaneously because interviewers are not required. J. E. the less mistakes there will be in the responses. Respondent convenience: Because the respondents can complete the interview in their own time. the more money that is spent on the printing job. There is no feedback or probing and answers must be assumed to be complete. However. then mail surveys may not be the best communication medium. Interviewer's absence: The respondent does not have the opportunity to ask the interviewer for explanations concerning the questions and. Limited duration: The length of the interview. An incentive may be required to encourage the respondents to complete the questionnaire. Unfortunately. Callbacks: Telephone callbacks can be made much more easily than personal callbacks.. B. they are different from interviews because the respondent takes responsibility for reading and answering the questions. Representative samples: Individuals without telephones or with unlisted numbers can represent certain demographic groups which can cause non-response bias to the sample. Questionnaires are usually printed on paper. D. Such anonymity can be stressed in mail surveys. This is ideal in the case of isolated interviewees. Researchers distribute questionnaires to consumers through the mail and in many other ways. or unless it can be demonstrated with some verification that the respondents are similar to the nonrespondents. Likewise. It is generally agreed that a mail survey cannot be considered reliable unless it has a minimum 50 percent response. Boring. and in some cases impossible. the higher the response rate is likely to be. Anonymity of respondent: Respondents are more likely to provide confidential or embarrassing information when they can remain anonymous. MAIL SURVEYS Mail surveys rely on the efficiency of the written word rather than the interviewer. SELF-ADMINISTERED QUESTIONNAIRES Many surveys do not require an interviewer’s presence. Random digit dialing (RDD) can partially eliminate the problem of unlisted numbers. I. Length of mail questionnaire: A general rule of thumb is that a mail questionnaire should not exceed six pages. There is no guarantee that the intended subject will fill out the form and there is no guarantee that respondents are similar to nonrespondents. IV. is definitely limited. Time is money: If time is a factor. . but they may be programmed into computers and posted on the Internet or sent via e-mail. G.
J. such as a university or the government. Interesting questions: Certain “interesting” questions. Airlines frequently pass out questionnaires to passengers during flights. There are a number of ways to increase the response rates and they are as follows: (1) prepaid postage return envelopes. and many individuals cannot be accessed by it. Color of 2. Personalize letters. SELF-ADMINISTERED QUESTIONNAIRES THAT USE OTHER FORMS OF DISTRIBUTION Many forms of self-administered questionnaires are very similar to mail questionnaires. questionnaire VI. 7. Facsimile machines or fax machines have been used as a way for respondents to return questionnaires and as a means to deliver questionnaires. Faxing a questionnaire to a long-distance number requires that the respondent pay for the transmission of the fax. Increasing response rates to mail surveys: People who feel strongly about. Extremely long questionnaires may be dropped off by an interviewer and then picked up at a later time. hotels. In both these surveys a follow-up was mailed resulting in an increase in responses around week four. A questionnaire inserted in a magazine may instruct the respondent to clip out the questionnaire and fax it to a certain phone number. E-MAIL SURVEYS Questionnaires can be distributed via e-mail. The speed of e-mail distribution and quick response time can be a major advantage for surveys dealing with time-sensitive issues. It is likely that people with extreme opinions will be more likely to respond. faster turnaround time. cheaper distribution and processing costs. Restaurants. 6. 4. 3. A money incentive may be offered to the respondent or to the respondent's choice of charity. Follow-ups: Exhibit 10. Avoid bulk rate postage. Keying mail questionnaires: An anonymous key number can be used to avoid the expense and inconvenience of mailing follow-up letters to those who have already responded. 5. However. Preliminary notification: Advance notification. . more flexibility. Many manufacturers use their warranty or owner registration cards to collect demographic information and data about where and why products were purchased.3 illustrates response rates for surveys. they may act as a means of attracting attention. VII. may positively influence response rates. Questionnaires can also be distributed via fax machines. a disadvantage of fax surveys is that only respondents with fax machines who are willing to exert the "extra" effort return questionnaires. certain circumstances lend themselves to e-mail surveys. 1. The optimal lead time for advance notification is approximately three days before the mail survey is to arrive. possibly completely unrelated to the survey. (2) questionnaires with attractive design and format. In a mail survey. and less paper chasing. such as internal surveys of employees or satisfaction surveys of retail buyers who regularly deal with the organization via e-mail. E-mail is a relatively new method of communication. These fax surveys replace the sender’s printing and postage costs and are delivered and/or returned faster than traditional mail surveys. has been successful in some situations. or electronic mail. either by letter or telephone. PRINTED. and other service establishments frequently print short questionnaires on cards so customers can evaluate the service. can be added to generate interest. (3) easily understood questions. A follow-up may include a second questionnaire or merely serve as a reminder. a prepaid postage enveloped places little burden on the respondent. or are interested in. Monetary incentives: These appear to be the most effective and least biasing incentive. Thus. The benefits of incorporating a questionnaire into an e-mail include speed of distribution. the subject matter are more likely to respond to mail surveys which results in non-response error. Cover letter: A good cover letter should include: a) A statement of why the survey is important—social usefulness appeal b) Appeals to the individual respondent—“help us out” appeal and the egotistical appeal c) Assurance of confidentiality d) Indication of prepaid postage e) A description of the incentive as a “reward” for participation f) An assurance that the questionnaire does not take long to complete g) A description of how the person was scientifically selected. 8. Other techniques. Survey sponsorship: The sponsorship of a well-known business research firm or another prestigious organization.
Because Internet surveys can be accessed anytime (24/7) from anywhere. samples can be much larger than with interviews or other methods used to distribute self-administered questionnaires. These computer-to-computer self-administered questionnaires eliminate the cost of paper. the questionnaire software may be B. And. Some individuals have minimal computer skills. Many in the survey research community believe Internet surveys are the way of the future. as with other questionnaires that rely on voluntary self-selection. The researcher can use more sophisticated lines of questioning based on the respondents’ prior answers.VIII. Respondent Participation and Cooperation Participation in some Internet surveys occurs when a computer user intentionally navigates to a particular Web site where questions are displayed. However. C. the visitor cannot venture beyond the survey page without providing information for the organization’s “registration” questionnaire. Surveys can be conducted in less than a week. individuals anticipate a survey at the web site and in other cases it is totally unexpected. Quick and Cost Effective Internet surveys allow marketers to reach a large audience (possibly from around the globe). postage. Assigning a unique password code also allows the researchers to track the responses of each respondent. In many other Internet surveys. if the purpose of the research is to evaluate how visitors feel about a Web site. This restricts access to individuals who are not part of the scientifically selected sample. Experienced researchers require a respondent to provide a password or PIN to move from the welcome page to the first question. and other administration costs. data entry cost. which may help to increase participant cooperation and willingness to spend more time answering the questionnaires. and very cost-effective manner. In some cases. they are given passwords/ PIN numbers and a link to a certain web site that contains a welcome screen. A. the incremental cost of additional respondents is marginal. When individuals receive e-mail invitations to participate. many individuals cannot quickly download high-resolution graphic files. Visual and Interactive Surveys conducted on the Internet can be interactive. the sample is not likely to be representative of the entire Unites States population because of self-selection error. response rates are low. The Internet is an excellent medium for the presentation of visual materials. clicking an icon. respondents are initially contacted via email. Many lack powerful computers or compatible software required to interact with advanced features programmed into Internet questionnaires. thereby identifying any respondent who makes an effort to answer the questionnaire more than once. D. sound. . Representative Samples The quality of Internet samples may vary substantially. If the survey is merely a sample of those who visit a Web page and voluntarily fill out a questionnaire. They may not know how to navigate through and provide answers to an Internet questionnaire. INTERNET SURVEYS An Internet survey is a self-administered questionnaire posted on a Web site. to personalize individual messages. they can reach certain hard-to-reach respondents. Like a cover letter in a mail survey. or keying in an answer. such as doctors. Because of low speed Internet connections (low bandwidth). And. Scientifically drawn samples from a consumer panel or samples randomly generated in other ways can yield representative samples. Many of these interactive surveys utilize color. and animation. randomly selecting every 100th visitor may accomplish the study’s purpose. Once an Internet questionnaire has been developed. and to secure confidential answers in a simple. quick. the welcome screen on an Internet survey serves as a means to gain the respondents’cooperation and provides brief instructions. A major disadvantage of Internet surveys is that many individuals in the general population cannot access the Internet. all people with Internet access do not access the Internet with the same technological level. Often they are members of consumer panels who have previously indicated their willingness to cooperate. participants tend to be more deeply involved with the subject of the research than the average person. In some instances. When the computer user does not expect a survey on a Web site and cooperation is voluntary. In addition. E. Respondents provide answers to questions displayed online by highlighting a phrase. Hence. Accurate Real-time Data Capture The computer-to-computer nature of Internet surveys means each respondent’s answers are directly entered into the researcher’s computer as soon as the questionnaire is submitted.
Follow-ups like this. Callbacks When the sample is drawn from a consumer panel. Increasing Response Rates A password system that identifies people who. these surveys often require a field-worker be at the on-site location to explain how to use the computer system. IX. That is. Use of a variety of dialog boxes (windows that prompt the respondent to enter information) allows questionnaire designers to be creative and flexible in the way questions are presented. and variations of most other techniques for increasing response rate to mail questionnaires are recommended for Internet surveys.programmed to reject improper data entry. interesting early questions. . Internet surveys do not offer the opportunity to send a physical incentive. While some researchers have had success with the promised incentives. Real-time data capture allows for real-time data analysis. At temporary locations. or other high traffic location to administer an interactive survey. responses to previous questions can lead to questions that can be personalized for individual respondents. the sponsor of this survey will make a sizable contribution to a national charity. self-selection often is a problem in this type of survey. The anonymity of the Internet encourages respondents to provide honest answers to sensitive questions. recontact to those who have not completed the survey is quick and easy. One major advantage of computer-assisted surveys is the computer’s ability to sequence questions based on previous responses. KIOSK SURVEYS A computer may be installed in a kiosk at a trade show. Sometimes the e-mails explain that additional incentives are offered for those individuals who comply with the request to finish the questionnaire. This is an obvious disadvantage. H. Because the respondent chooses to interact directly with an on-site computer. new advertising campaigns. However. G. preliminary notification. The computer can be programmed to skip from question 6 to question 9 if the answer to question 6 is no. software is available to allow questioning to branch off into two or more different lines depending on each respondent’s answer to filtered questions. Because computer software can identify the passwords of respondents who only completed a portion of the questionnaire. Computer literate individuals are most likely to complete these interactive questionnaires. No system can be 100% secure. and other top-secret ideas. Incentives to respond to survey must be in the form of a promise of a future reward such as: As a token of appreciation for completing this survey. Respondent Anonymity Respondents are more likely to provide sensitive or embarrassing information when they can remain anonymous. have not participated in the survey can send a 'friendly' e-mail reminder asking the nonrespondents to participate in the survey before the study ends. academic research about Internet surveys is sparse and few definitive answers about the most effective ways to increase response rates may be given at this moment in time. Up-to-theminute sample size counts and tabulation data from an Internet survey can be viewed in real-time. customized messages can be sent to individuals who terminated the questionnaire with only a few additional questions to answer. such as a convention. after a pre-determined period of time. many research suppliers specializing in Internet surveying have developed password protected systems that are very secure. Respondents may worry if personal information will remain private. Unlike mail surveys. to the respondent. Furthermore. One important feature of these systems is to restrict individuals from filling out a questionnaire over and over again. such as a dollar bill. Personalized and Flexible Questioning Computer-interactive Internet surveys are programmed in much the same way as computer-assisted telephone interviewing surveys. at professional conference. You can vote for your preferred charity at the end of this survey. I. in an airport. F. Security Concerns Many organizations worry that hackers or competitors may access Web sites researching new product concepts. It is often a simple function of computer software automatically sending an e-mail reminder notification to panel members who did not visit the welcome page.
it is advisable to carry out a pretest or a screening procedure before administering the surveys.4 presents a summary of the major advantages and disadvantages of the various methods. anonymity — may be different for each individual survey. Pretests are trial runs with a group of respondents to iron out fundamental problems in the instructions or design. SELECTING THE APPROPRIATE SURVEY RESEARCH DESIGN There is no best form of design — each has its advantages and disadvantages.PRETESTS To check that questionnaires are understandable and not ambiguous or misleading. The different criteria — cost.X. . XI. The test can also be given to managers to ensure that it provides them with the information they require. speed. Exhibit 10. Screening procedures involve administering the test to a panel of experts to identify errors.
It is presumed that response latency indicates the strength of preference between alternatives.3 gives an indication of the amount of additional information that could be acquired from observation. Hidden observation does minimize respondent error. Visible observation occurs when the subject knows the observer is there. In some cases. finding out the amount of time that each customer spends in the different departments of a large store can be easily done by observing the customers. Verbal behavior C. IV.Chapter 11 Observation Methods I. perhaps inaccurately. and can only be used to observe behavior of a short duration. WHAT CAN BE OBSERVED? Six kinds of behavior of people and objects can be observed: A. Exhibit 11. Certain data may be quickly and easily obtained through observation .for example. The researcher utilizing the observation method of data collection witnesses and records information as events occur. Nonverbal behavior: Behavioral scientists have recognized that nonverbal behavior can be a communication process by which meanings are exchanged between individuals. race or sex. Observer bias may occur. direct observation is not error free. observation of nonverbal communication may hold a considerable amount of promise for business researchers. Human behavior and physical actions B. Direct observation: Direct observation can produce a detailed record of what events occur or what people actually do. Errors associated with direct observation: Despite the fact that there is no interaction with the respondent. Physical objects G. but raises an ethical question concerning privacy. observation is the most straightforward method of data collection. B. objects. Expressive behavior D.OBSERVATION OF HUMAN BEHAVIOR A. and occurrences without questioning or communicating with them. Supplementary evidence: Nonverbal behavior may also be used to supplement information from interviews. For example. Another common use of observation is recording the decision time necessary to make a choice between two alternatives. and hidden observation occurs when the subject is unaware of the researcher’s presence. D. III. NATURE OF OBSERVATION STUDIES The un obtrusive or no reactive nature of the observation method often generates data without the subject’s knowledge. For example. Spatial relations and locations E. In direct observation a respondent is never asked to recall. it may be the most economical. Many types of data can be recorded more accurately through direct observation than through questioning the subjects. Thus. It is a particularly useful technique when a particular respondent cannot express their reaction to a product. Verbal records Observation has certain limitations. a distortion of measurement resulting from the cognitive behavior or actions of the witnessing observer. in other cases. or the response latency. WHEN IS OBSERVATION SCIENTIFIC? Scientific observation is the systematic process of recording behavioral patterns of people. or compiles evidence from records of past events.Temporal patterns F. II. an event. it cannot be used to describe any cognitive phenomena. a child playing with a toy cannot express his opinion of it. that is. The observer plays a passive role and merely records what he sees. Accuracy is affected if the observer does not record . but how many of those customers could actually tell you how long they had spent in each department. C.
advertising copy. There are four major devices used: 1. articles. events. not the responses from individuals. etc. V. OBSERVATION OF PHYSICAL OBJECTS Physical phenomena may be the subject of an observation study. etc. the social behavior. It deals with the study of the message itself. this will result in multiple hits on that page. the wear and tear of a book gives an indication of how often it has been read. VII. A “hit” occurs when a user clicks on a single page of a web site. Organizations with Web sites consisting of multiple pages find it useful to track “page views” or single. Measuring physiological reactions: Business researchers have used a number of mechanical devices to evaluate consumers’ physical and physiological reactions to packaging. and relationships in a social setting. VIII.. or space and time relationships. An example of content analysis is projecting social trends through the use of newspaper content analysis. A variety of information technologies are used to measure web traffic and to maintain access logs. and simplify the activities. CONTENT ANALYSIS Content analysis: Content analysis obtains data by observing and analyzing the content of advertisements. IX. the primary means of observation is mechanical rather than human. . For example. For example. OBSERVING SOCIAL SETTINGS In many situations the purpose of the observation is to summarize. and there may be variations in the observers’ interpretations of the actions. meaning. Specifically. Eye tracking monitors: These record how the subject actually reads or views an advertisement. the Nielsen Television Index system for estimating the nation’s TV audiences utilizes a consumer panel and mechanical observation to obtain ratings for TV programs. The researcher can also measure unconscious eye movements using an oculometer. Participant observation refers to situations where an observer gains firsthand knowledge by being in or around the social setting that is being investigated. this helps advertisers identify and emphasize the selling points of an advertisement. this technique measures the extent of emphasis on omission of any given analytical category. If the visitor clicks on many places to access graphics. Such a study of the content of communication is more sophisticated than simply counting the items. etc. The evidence it provides will be used to determine a course of action. For example. the setting. it requires a system of analysis to secure relevant data.every detail that describes the observation. VI. Through the use of physical traces. however.SCIENTIFICALLY CONTRIVED OBSERVATION The creation of an artificial environment to test a hypothesis is called contrived observation and such a setting can increase the occurrence of certain behavioral patterns. The business research companies that track which sites on the World Wide Web are most popular install a special tracking program onto personal computers of a sample of computer users who have volunteered to participate in the research effort. Some common elements that are observed are: the participants. discrete clicks on individual pages. the content of an advertisement might be investigated with regard to the use of words. For example. Most organizations record how many people visit their Web sites. the purpose that brought the participants together. Page views more conservatively indicate how may users visit each individual page on the Web site. such fieldwork can be expensive and it is not always easy to conduct exploratory research.MECHANICAL OBSERVATION In many situations. characters. systematize. A. a researcher can often record data that a respondent could not recall accurately. This eliminates problems such as inaccuracy or lack of truthfulness. a “pantry audit” counts the packages of groceries in a consumer’s home. and the frequency and duration of the social setting. For example. and other stimuli. Physical-trace evidence is a visible sign of some past event or occurrence. the observer can pretend to be a “shopper” in order to investigate a salesman’s behavior. letters. It may also be useful to track the path or sequence of pages that each visitor follows.
The establishment of a scanner-based consumer panel to replace the consumer purchase diaries is one of the primary ways of implementing this type of research. The major disadvantage of these devices is that subjects are usually placed in an artificial setting and are aware of the fact that they are being observed. etc. Changes in pupil size may be interpreted as changes in cognitive activity resulting from the stimulus. Advances in bar code technology may lead to at-home screening systems which will use hand-held wands that read UPC symbols. Voice pitch analysis: This measures emotional reactions through physiological changes in a person’s voice. 3. Pupilometer: This device observes and records changes in the diameter of the subject’s pupils. Psychogalvanometer: This measures galvanic skin response (GSR) which is a measure of involuntary changes in the electrical resistance of the skin. 4. Scanner-based research: The use of lasers for character recognition and the bar code technology with UPC has accelerated the use of technological observations in business research. . packages. B. This device utilizes the assumption that physiological changes accompany emotional reactions to advertisements.2.
to find out how much consumers utilize unit pricing. For example. The purpose of unit pricing is to allow shoppers to avoid any confusion due to price calculations. .1 shows how such an experiment could be conducted in chain stores. Exhibit 12. A. Events may be controlled in an experiment in a way not possible in a survey. this type of information may not facilitate price comparisons. the researcher can control for potential sources of error in the experiment and get a better measure of the effect that the independent variable had on the experimental group.3 shows the results of such an experiment. Rather than use a survey method. 1. a restaurant chain might investigate the combined effects on sales of an increase in advertising and a change in prices. However. then additional experimental groups can be added (each with the independent variable at different levels) to get a better idea of the relationship. An illustration: a unit pricing experiment: The concept of experimentation is best illustrated with an extended example concerning unit pricing. The researcher must select the appropriate levels of the variable as experimental treatments. THE NATURE OF EXPERIMENTS The purpose of experimental research is to allow the investigator to control the research situation so that causal relationships among variables may be evaluated. Experimental treatments are the alternative manipulations of the independent variable being investigated. 2. BASIC ISSUES IN EXPERIMENTAL DESIGN Decisions must be made about four basic elements of an experiment. 3. By holding conditions in a control group at zero.Chapter 12 Experimental Research I. For example. However. Its value may be changed or altered independently of any other variable. we know that store image may have affected sales. Several experimental treatment levels: If the use of only two groups does not tell the researchers everything that they wish to know about the relationship between the dependent and independent variable. These elements are described below. the number of dollars that can be spent on advertising may be any number of different values. The results show that a change in the presentation of unit price information (the independent variable) caused a change in the average unit price paid (the dependent variable) when other variables were controlled for possible causal effect. while other variables that may confound the relationship are eliminated or controlled. but there is some evidence that unit pricing has failed to change consumers’ purchasing habits. in some situations. but we controlled for this variable by using only one type of store in one city. In business research. which may not yield true responses (who is going to admit that they do not use unit pricing?). In an experiment one variable (the independent variable) is manipulated and its effect upon another variable (the dependent variable) is measured. Experimental and control groups: In the simplest form of experiment only two values of the independent variable are manipulated. Exhibit 12. or unchanged. II. The current form of unit price display is a separate shelf tag for each item. The independent variable is hypothesized to be the causal influence. For example. More than one independent variable: It is possible to assess the effects of more than one independent variable. the influence of the differences between the stores would be minimized. experimentation could be used. By randomly assigning the different experimental conditions to the stores. the independent variable is often a categorical or classificatory variable representing some classifiable or qualitative aspect of business strategy. the independent variable is a continuous variable. Manipulation of the independent variable: The experimenter has some degree of control over the independent variable. A.
which is comparable to systematic error in total survey error. steps are typically taken to make it difficult for subjects to know what the researcher is trying to find out. Random assignment of subjects allows the researcher to assume that the groups are identical with respect to all variables except the experimental treatment. but they may not rebuy. which occurs when subjects exhibit the behavior that they feel is expected of them rather than their actual behavior. If the . A common effect caused by demand characteristic is the Hawthorne effect. Constant experimental error: Random error has already been discussed. Repeated measure: Exposing the same subjects to all experimental treatments eliminates any problems due to subject differences. C. For example. People are the most common test units in organizational behavior and consumer behavior experiments. Selection of the dependent variable is a crucial decision in the design of an experiment. may occur. It is presumed that changes in the dependent variable are a consequence of changes in the independent variable. Random sampling error: This type of error may occur if repetitions of the basic experiment sometimes favor one experimental condition and sometimes the other on a chance basis. it is generally accepted that random assignment of participants to groups and experimental treatments is still the best procedure. this will result in a constant. Such an experimental technique is said to have repeated measure. if the subjects of the experimental group are always administered the treatment in the morning and the subjects of the control group are always administered the treatment in the afternoon. Thus. thereby distorting the experiment. the chance of unknown nuisance effects piling up in particular experimental groups can be identified. The guinea pig effect. Consumers may try a “loser” once. which is comparable to sampling error in total survey error. but it causes some other problems which will be discussed later. Matching: Matching the respondents on the basis of pertinent background information is another technique for controlling assignment errors. 3. 2. Control over extraneous variables: To understand this issue. Sample selection error: Sample selection error may occur because of the procedure utilized to assign subjects or test units to either the experimental group or the control group. the various types of experimental error should be understood. 4. However. Randomization: This is the random assignment of subjects and treatments to groups. Another common effect occurs when people in an experiment interact. Extraneous variables: A number of extraneous variables may affect the dependent variable. For example. D. For example. Randomization is one device for equally distributing or scattering the effects of extraneous variables. if cars with different engine sizes were used in the test then extraneous variables would affect the results. a savings and loan conducting an experiment may have greater assurance that there are no differences among subjects if subjects in all experimental conditions are matched according to age. Selection and assignment of test units: Test units are the subjects or entities whose responses to the experimental treatment are measured or observed. Constant error (bias) occurs when the extraneous variables or the conditions of administering the experiment are allowed to have an influence on dependent variables every time the experiment is repeated. Demand characteristics: The term demand characteristics refers to experimental design procedures that unintentionally hint to subjects about the experimenter’s hypothesis. this may result in “joint” decisions rather than a desired individual decision.B. 1. If participants recognize the experimenter’s expectation or demand. Selection and measurement of the dependent variable: The dependent variable is the criterion or standard by which the results are judged. The experimenter’s choice of a dependent variable determines what type of answer is given to the research question. 1. The errors in research design can be classified into two basic categories—random error. the time period for the effects to become evident should be carefully considered. and constant error. if age is expected to influence savings behavior. when testing the efficiency of a car fuel. 5. even slight nonverbal cues may influence their reactions. This causes people to perform differently just because they know they are experimental subjects. For example. Often the dependent variable selection process is not as carefully considered by the researcher as it should be. 3. systematic error which must be taken into account. they are likely to act in a manner consistent with the experimental treatment. in a test market. 2. To reduce demand characteristics.
III. confidentiality. In a laboratory experiment. This technique is frequently used in medical research when chemically inert pills (placebos) are given to subjects who do not receive any medication.FUNDAMENTAL QUESTIONS IN EXPERIMENTATION A. B. there may be an error due to order of presentation. Establishing control: The major difference between experimental research and other research is the experimenter’s ability to hold conditions constant and manipulate the treatment. Basic versus factorial experimental designs: In basic experimental designs. As experiments increase in naturalism they begin to approach the pure field experiment. The researchers must concern themselves with issues relating to privacy. It is not always possible to control everything that would be desirable in the perfect experiment. When extraneous variables cannot be eliminated. is recommended that researchers debrief subjects of laboratory experiments to explain the purpose of the research. packaging and so on by controlling the amount of time a visual image is exposed to a subject. If certain extraneous variations cannot be controlled. one contrived for a specific purpose (laboratory experiment). Factorial experimental designs are more sophisticated than basic experimental designs. They allow for an investigation of the interaction of two or more independent variables. and as they become more artificial they approach the laboratory type (see Exhibit 12. It may also involve experimenters. and other ethical concerns. The degree of artificiality in experiments refers to the amount of manipulation and control of the situation that the experiment creates to ensure that the subjects will be exposed to the exact conditions desired. . Counterbalancing attempts to eliminate the confounding effects of order of presentation by requiring that half of the subjects be exposed to treatment A first and then to treatment B while the other half receive treatment B first then treatment A. ETHICAL ISSUES IN EXPERIMENTATION Many ethical issues in experimentation. it must be assumed that the confounding effects will be present in all experimental conditions with approximately the same influence. This procedure strives to have all the subjects in each experimental group exposed to situations that are exactly alike except for the differing conditions on the independent variable. When deception is used. This kind of experiment is known as a double-blind design because neither the subject nor the experimenter knows which are the experimental and which are the controlled conditions.5). single dependent variable. IV. Of course. experimenters strive for conditions of consistency. consumers in a taste test may have a tendency to prefer the first product they taste if they cannot taste the difference between the two products. Blinding is utilized to control subjects’ knowledge of whether or not they have been given a particular experimental treatment. the researcher has almost complete control over the research setting. the use of deception presents an ethical question that must be resolved by the researcher. the participant does not know how to be a “good” subject to “help” confirm the hypothesis. Factorial designs are discussed later in this chapter.purpose of the experiment is disguised. An example of a very controlled experiment is a tachioscope which can be used to experiment with the visual impact of advertising. The random assignment of subjects and experimental treatments to groups is an attempt to control extraneous variations resulting from chance. Field and laboratory experiments: A business experiment can be conducted in a natural setting (field experiment) or in an artificial setting. If the experimental method requires that the same subjects be exposed to two or more experimental treatments. 4. Field experiments are generally utilized to fine-tune business strategies and to determine sales volume. a single independent variable is manipulated to observe its effect on another. deception. For example. Some laboratory experiments are more controlled than others. the researchers do not know which of the pills are real and which are placebos. Debriefing is the process of providing subjects with all the pertinent facts about the experiment after the experiment has been completed.
Student surrogates: One issue of external validity concerns the common use of students as experimental subjects. they often are not representative of the total population. Trade-offs between internal and external validity: Natural experiments tend to have greater external validity than artificial laboratory experiments. but laboratory experiments with many controlled factors are usually higher in internal validity. However. Selection: The selection effect is a sample bias resulting in differential selection of respondents for the comparison groups. This occurs when subjects withdraw from the experiment before it is completed. taking a pretest before the independent variable is manipulated may sensitize respondents when they are taking the test for a second time. Thus. national sales projections cannot be made . which provides a natural testing of consumers’ reactions. 3. the experiment is a basic design. Mortality: If the experiment is conducted over a period of a few weeks or more. Testing effects: These are also called pretesting effects because the initial measurement or test alerts respondents to the nature of the experiment and respondents may act differently than they would have if no pretest measures were taken. Mortality effects may occur if many subjects drop from one experimental group and not from other treatment or control groups. Although this may reduce the effect of testing because of a change in the measuring instrument. A common history effect occurs when competitors change their business strategies during a test marketing experiment. If only one variable is manipulated. A special case of the history effect occurs when a change in the dependent variable occurs because members of one experimental condition experienced historical situations different from those of the other experimental conditions. A trade-off must be made when designing the experiment. 2. is the controlled store test. B. One common hybrid between a laboratory experiment. For example. there may be some sample error due to mortality. and a test market. and sales movement and market share can be measured weekly. 5.Generally. there may be a testing effect. If the experimenter wishes to investigate several levels of the independent variable (for example. which simulates a controlled purchasing environment. if an experiment lacks internal validity. during a day-long experiment subjects may grow tired. bored.only benchmark sales data can be obtained because of the relatively small sample of stores and the limitation on the type of outlet where the product is tested. This is referred to as the cohort effect. 2. 4. Maturation: The effects due to maturation are changes within the respondents that operate as a function of time rather than of a specific event. There are numerous sources of instrument decay or variation. If the identical measuring instrument is used more than once. V. VI. the threats to internal validity may also jeopardize external validity. this may result in an instrumentation effect. The products are put into stores in a number of small cities or into selected supermarket chains. It is helpful to classify several different types of extraneous variables that may jeopardize internal validity. an alternate form of measuring instrument may be given during the posttest or after measurement. In essence. although students are easily accessible. subjects will be aware of their participation in laboratory experiments and it is common to debrief subjects of laboratory experiments to explain the purpose of the research. four price . 1. To avoid the effects of testing. The student population is atypical and.CLASSIFICATION OF EXPERIMENTAL DESIGNS There are various types of experimental designs. or sample attrition. 1. History: History refers to specific events in the environment between the first and second measurement that are beyond the control of the experimenter. ISSUES OF EXPERIMENTAL VALIDITY A. it is a sampling question: To what extent can the results of a simulated shopping experiment be transferred to real-world supermarket shopping? Of course. In a before-and-after study. or hungry. Such tests offer secrecy. projecting the result is not possible. External validity: External validity is the quality of being able to generalize beyond the data of the experiment. Internal validity: Internal validity refers to whether or not the experimental treatment was the sole cause of observed changes in the dependent variable.
the static group design is referred to as an “after-only” design with control group. this design is used frequently in business research. We have no means of controlling extraneous variables. then the experiment requires a complex. We need a measure of what would happen when the test units have not been exposed to X to compare with the measure when the subjects have been exposed to X. If the groups can be determined by the experimenter rather than existing as a function of some other causation. Nevertheless. One-shot design. In this experiment.levels). Three examples of quasi-experimental design: Quasi-experimental designs do not qualify as true experimental designs because they do not adequately control for the problems associated with loss of external or internal validity. and mortality that may jeopardize internal validity. However. One-group pretest-posttest design: O1 X O2 This design offers a comparison on the same individuals before and after training. On many occasions “after-only” designs are the only possible ones. 3. This is particularly true when conducting “use tests” for new products or brands. despite its weaknesses. 1. Static group design: Experimental group: Control group: X O1 O2 The results of a static control group are computed by subtracting the observed results in the experimental group from those in the control group (O1 – O2 ). Although this is an improvement over the one-shot design. or the after-only design: X O1 This one-shot design is a case study fraught with problems. 1. If more than one observation is taken. R = random assignment of test units. Symbolism for Diagramming Experimental Designs The following symbolism facilities the description of the various experimental designs: X = exposure of a group to an experimental treatment. Three good experimental designs: The three basic experimental designs discussed next will have the symbol R to the left of the diagram to indicate that the first step in a true experimental design is the randomization of subject assignment. B. 2. A major weakness of this design is that we have no assurance that the groups were equal on variables of interest before the experimental group received the treatment. experimental design. If the groups were selected arbitrarily by the investigator or if entry into either group was voluntary. Pretest-posttest control group design (before-after with control): Experimental group: R O1 X O2 . A. or if he or she wants to investigate effects of two or more independent variables. subscripts will be given to indicate temporal order. testing effect. under certain circumstances this design is the only viable choice. the reader should assume a time flow from left to right. or statistical. this design still has certain weaknesses such as maturation. O = observation or measurement of the dependent variable. we do not have any kind of comparison. Random assignment of subjects may minimize problems with group differences. C. then there may be systematic differences between the groups that could invalidate the conclusions about the effect of the treatment. As we diagram designs utilizing these symbols.
instrument decay. Time series experimental designs: Business experiments that are investigating long-term structural change may require a time series design. the researcher variable. A. circumstances usually dictate a compromise with true experimentation. it is rarely used in business research because of the time. . The effect of the experimental treatment equals (O2 – O1) – (O4 – O3). all the same assumptions about extraneous variables are made. it is possible to isolate the effects of the experimental treatment and interactive testing in this design. In such cases the following quasi-design is utilized: O1 O2 O3 X O4 O5 O6 Several observations are taken to identify trends before the treatment is administered. VII. and regression effects. this time series design cannot give the researcher complete assurance that the treatment caused the change in the trend. several observations are made to determine if the patterns after the treatment are similar to those before. In some situations no pretest measure is possible about the effect of the treatment (for example. Completely randomized designs: This design uses a random process to assign experimental units to treatments. These compromise designs may fall short of the requirements of assigning subjects or treatments randomly to groups. Of course. but it does enable the researcher to distinguish temporary changes from permanent changes. Further. This assumption is also made for history. It is assumed that the effect of extraneous variables will be the same on both the experimental and the control groups. Business researchers employ various complex designs. An improvement on the basic time series design is to utilize a time series with control group. Randomization of experimental units is the researcher’s attempt to control all extraneous variables while manipulating a single factor. Generally this is considered a greater weakness than utilizing groups that have already been established. Posttest-only control group design (after-only with control): Experimental group: Control group: R R X O2 O1 The effect of the experimental treatment is equal to O2 . testing effects. interactive testing effect is possible when the subjects are sensitized to the subject of the research. COMPLEX EXPERIMENTAL DESIGNS Complex experimental designs are statistical designs that allow for the manipulation of more than one independent variable in the experiment. 2. This weakness in the before-after with control group design can be corrected (see the next two designs). D. The alternative to the compromise design is to conduct the experiment without a control group. an experiment testing the effectiveness of athlete’s foot remedy). Several of the experiments discussed in the previous chapters are completely randomized. Although this design allows for isolation of the various effects. When the experiment involves a longitudinal study. the effects of testing and instrument decay are eliminated. they operate equally on both groups. they are most vulnerable to historical changes. maturation. that is. E. 3. and cost of implementing it. However. effort. With only posttest measurement. When experiments are conducted over long periods of time.O1 . Compromise experimental designs: In many instances of business research true experimentation is not possible—the best the researcher can do is approximate an experimental design. Solomon four-group design: Experimental group 1: Control group 1: Experimental group 2: Control group 2: R R R R O1 O3 X X O2 O4 O5 O6 Although we will not go through the calculations. After the treatment.Control group: R O3 O4 This design has the advantage of the before-after design with the additional advantages gained from having a control group.
each letter occurs only once in each row and in each column. In other words. two-way classification square. The number of treatment levels for the confounding factors must be equal which can present certain problems. The end result of this design is that each treatment will be administered under conditions involving all levels of both confounding factors. . A major advantage of the factorial design is its ability to measure the interaction effect which may be more or less than the total of the main effects.11 shows the result of a believability study conducted using a factorial design. Randomized block design: The randomized block design is an extension of the completely randomized design. Exhibit 12. Factorial designs allow for the testing of two or more treatments (factors) at various levels. Latin square design: This design attempts to block out the effect of two or more confounding extraneous factors. the believability of the advertising copy depends on whether a man or a woman views the advertisement. Subject 1 1 2 3 A B C Order of Usage 2 B C A 3 C A B The letters A. For example. The slopes of the two lines are different which indicates interaction between the two treatment variables. It is not necessary that the treatments have the same number of levels. D. A Latin square design may have any number of treatments. The logic behind the randomized block design is similar to the logic underlying the selection of a stratified sample rather than a simple random sample. suppose a retail group wishes to control for shelf space and city where the product is sold. A 3 x 2 design requires six cells or six experimental designs (3 x 2 = 6). and the interaction effect is the effect of the combination of the independent variables on the dependent variable. one known source of extraneous variation may be separately accounted for. the Latin square design can be very advantageous to use in certain situations. if the researcher has identified a single “extraneous variable” that might affect test units’ response to the treatment. one type of extraneous variation is partitioned out and a more efficient experimental design results.12 portrays a graph of the same believability study that is portrayed in Exhibit 12. they are still limited. but it wishes to experiment with four shelf heights. The first factor (price) is varied in three ways and the second factor (package design) is varied in two ways.B. A form of randomization is utilized to control for most extraneous variables. C. In the following 3 x 3 matrix. and C identify the three treatments. A major assumption of the Latin square design is that interaction effects are expected to be minimal or nonexistent. Factorial design: Even though the single-factor experiments already considered may have one specific variable blocked and others controlled for. By grouping test units into homogeneous blocks of some relevant characteristic. Factorial designs also allow researchers to measure the interaction effects of the independent variables. 1. For example. For example a 3 x 2 x 4 factorial design (with 24 cells) is possible.11. Even with some drawbacks. The main effect is the influence on the dependent variable by each independent variable.10 shows a 3 x 2 factorial design. Exhibit 12. However. By isolating the block effects. the rows and columns of the table identify the confounding factors. The retail group may be limited in its experiment because it markets in only three cities. B. The Latin square is a balanced. Graphic Interaction: Exhibit 12. A factorial design allows for the simultaneous manipulation of two or more independent variables at various levels. then the researcher will attempt to isolate the single variable by blocking out its effects. a 5 x 5 matrix may be used with 25 cells.
a 2 to the place position and a 3 to a show position. This is merely a label for betters and horse racing enthusiasts. Further. The location of the zero point is arbitrary. To have precise measurement in business research requires a careful conceptual definition.” We know that “excellent” is better than “good. The classic example of an interval scale is the Fahrenheit temperature scale. an item’s. A. and so on. they measure order (or distance) in units of equal intervals. present great problems in terms of definition and measurement. Concepts such as brand loyalty. The four types of scale in business research are as follows: A. many measurement scales used in business research are not directly comparable. and a system of consistent rules for assigning numbers or scales. assign a 7. If the individual is an extremely brand loyal individual. If the individual is a total brand switcher with no brand loyalty.” or “poor. An example of a measurement rule might be “assign the numerals 1 through 7 to individuals according to how brand loyal they are. Ordinal scale: This scale arranges objects or alternatives according to their magnitude. we assign a 1 to a win position. both money and weight are ratio scales because they possess an absolute zero and interval properties. A concept (or construct) is a generalized idea about a class of objects. It specifies what the investigator must do to measure the concept under investigation. attributes.3 depicts a horse race in which the win horse was two lengths ahead of the place horse. an operational definition. a scale is a continuous spectrum or series of categories.” “fair. The numbers or letters assigned to objects serve as labels for identification or classification. An operational definition tells the investigator “do such-and-such in so-and-so manner. The first question the researcher must ask is “What is to be measured?” This question is not as simple as it first seems. usually quantitatively. III. The absolute zero represents a point on the . personality. The first drawing in Exhibit 13. B. B. If the temperature is 80 . A precise definition of the concept may require a description of how it will be measured. occurrences. D. or an event’s place in the spectrum. Nominal scale: The simplest type of scale. The purpose of scaling is to represent. WHAT IS TO BE MEASURED? Any researcher has the opportunity to select a measuring system. Unfortunately. Concepts: Before the measurement process can occur. and there is frequently more than one way of measuring a concept. or processes. Interval scale: Exhibit 13.” “good. C. RULES OF MEASUREMENT A rule is a guide instructing us what to do. Interval scales not only indicate order. the researcher has to identify and define the concepts relevant to the problem. Ratio scale: Ratio scales have absolute rather than relative scales. and so on as “excellent.2 presents some operational definitions and measures of job challenge from a study on the quality of life.” Exhibit 13.Chapter 13 Measurement and Scaling Concepts I. For example.3 depicts the number 7 on a horse’s color.” but we don’t know by how much. In other words. a person’s. TYPES OF SCALES A scale may be defined as any series of items that are progressively arranged according to value or magnitude into which an item can be placed according to its quantification. II. companies.” Operational definitions help the researcher specify the rules for assigning numbers. An operational definition gives meaning to a concept by specifying the activities or operations necessary to measure it. but the distance between the horses is known. In our race horse example. it cannot be said that it is twice as hot as a 40 temperature. true measurement of concepts requires a process of precisely assigning scores or numbers to the attributes of people or objects. assign a 1. A typical ordinal scale in business asks respondents to rate brands. Not only is the order of the finish known. Operational definitions: Concepts must be made operational in order to be measured.
It is the degree to which measures are free from random error and. if the duration of the time period between measures is long. or some other form of maturation. the first measure may sensitize the respondents to their participation in a research project and. Assessing the repeatability of a measure is the first aspect of reliability. The second underlying dimension of reliability concerns the homogeneity of the measure. therefore. THREE CRITERIA FOR GOOD MEASUREMENT There are three major criteria for evaluating measurements: A. E. There are two dimensions of reliability. Exhibit 13. interval scales are typically the best measurements. and residence. person. index of social class may be based on three weighted averages: residence. or composite measures. there could be a systematic bias to identify brands “I wish I could afford” rather than the brand usually purchased. there may be attitude change. there will be problems. V. However. the researcher will be unsure as to whether the measure has intrinsically low reliability. when a researcher checks the results of one half of the scale items to the other half. For example. If there is a high correlation between the two scales. measuring more complex concepts may require more than one question because the concept has several attributes. Asking different questions in order to measure the same concept provides a more accurate cumulative measure than does a single-item measure. indicates a high degree of reliability. or whether the particular equivalent-form has failed to be similar to the other form. subsequently.Mathematical and statistical analysis of scales: The type of scale utilized in business research will determine the form of the statistical analysis. The equivalent-form method is utilized when two alternative instruments are designed to be as equivalent as possible.4 shows the appropriate descriptive statistics for each type of scale. If the measure is stable over time. the researcher can conclude that the scales are reliable. for most behavioral business research. An attribute is a single characteristic or fundamental feature pertaining to an object. yield consistent results. An attempt to measure an attitude may require asking several questions or a battery of scale items. if there is a low correspondence. first. is the most basic method for checking internal consistency. However. High stability correlation. There are three basic approaches to dealing with the issue of validity: 1. they measure homogeneity rather than over-time stability. of the subjects which will affect the responses. Both of the above methods assume that the concept is uni-dimensional. Further. influence the results of the second measure. To measure the internal consistency of a multiple-item measure. The split-half method. . in measuring intention to buy. scores on subsets of items within the scale must be correlated. or issue. Validity: The purpose of measurement is to measure what we intend to measure. For example. Face or content validity: This refers to the subjective agreement of professionals that a scale logically appears to be accurately reflecting what it purports to measure. situation.scale where there is an absence of the given attribute. IV.INDEX MEASURES This chapter thus far focused on measuring a concept with a single question or a single observation. Multi-itemed instruments for measuring a single concept with several attributes are called index measures. if it does not. Researchers attempt to provide some evidence of a measure’s degree of validity. However. Validity addresses the problem of whether or not a measure does indeed measure what it purports to measure. the repeated test administered under similar conditions should obtain similar results. The test-retest method involves administering the same scale or measurement to the same respondents at two separate points in time to test for stability. There are two problems with measures of test-retest reliability. B. occupation. or consistency between the two measures at time one and time two. repeatability and internal consistency. Reliability: Reliability applies to a measure when similar results are obtained over time and across situations.
D. they are more sensitive than single item scales.5. Sensitivity refers to the ability of a instrument to accurately measure variability in stimuli or responses. or other hypothetical constructs. if the measure behaves the way it is supposed to in a pattern of intercorrelation with a variety of other variables. If a new measure of length were developed. Criterion validity may be classified as either concurrent validity (when the measure is taken at the same time as the criterion measure) or predictive validity (when the measure predicts a future event). 3. The differences between reliability and validity can be illustrated using the rifle target in Exhibit 13. finding that the new measure correlated with other measures of length would provide some assurance that the measure was valid. because index measures allow for a greater range of possible scores. Criterion validity: Criterion validity is an attempt by researchers to answer the question “Does my measure correlate with other measures of the same construct?” Consider the physical concept of length. This is a complex method of establishing validity and of less concern to the applied researcher than to the basic researcher. In other words. Sensitivity: The sensitivity of a scale is important. particularly when changes in attitude. Construct validity: Construct validity is established by the degree to which the measure confirms a network of related hypotheses generated from a theory based on the concept. then there is evidence for construct validity. C. Reliability. A measure has discriminant validity when it has a low correlation with measures of dissimilar concepts. are under investigation. 4. although necessary for validity. In its simplest form. .2. Reliability versus validity: The concepts of reliability and validity should be compared. Convergent and discriminant validity: Convergent validity is the same as criterion validity because a new measure is expected to predict or converge with similar measures. is not in itself sufficient. The sensitivity of a scale which is based on a single question or a single item can be increased by adding additional questions or items.
Attitudes as Hypothetical Constructs The term hypothetical construct is used to describe a variable that is not directly observable. the measurement of attitudes is a very important task. B. such scales are used if questionnaires are extremely long. ATTITUDE RATING SCALES Rating scales are perhaps the most common attitude measures. but several other scales have been developed that do provide more precise measurement. events. Attitudes are considered to be such variables.Chapter 14 Attitude Measurement I. •Behavioral component: reflects buying intentions and behavioral expectations. physiological to verbal. attitude scaling requires that an individual agree with a statement or respond to a single question. or prospects’. THE ATTITUDE-MEASURING PROCESS There are a remarkable variety of techniques that have been devised to measure attitudes.. A. Simple Attitude Scaling In its most basic form. Physiological measures of attitudes provide a means of measuring attitudes without verbally questioning the respondent. •Rating asks the respondent to estimate the magnitude of a characteristic. A number of simplified scales are merely checklists. etc. However. or making a choice or a comparison. Measuring Attitudes is Important to Managers Most marketing managers hold the intuitive belief that changing consumers’. or for other specific reasons. •Cognitive component: represents one’s awareness of and knowledge about an object. The purpose of an attitude scale is to find out an individual’s position on the continuum. but is measurable by an indirect means such as verbal expression or overt behavior. •Ranking tasks require that the respondent rank order a small number of objects in overall preference on the basis of some characteristic or stimulus. ATTITUDE DEFINED Attitudes are usually viewed as an enduring disposition to consistently respond in a given manner to various aspects of the world. III. attitudes towards a product is a major marketing goal. galvanic skin responses. rating. These techniques range from direct to indirect. measure blood pressure. There are three components of attitude: •Affective component: reflects an individuals general feelings towards an object. Category Scales . or quality. and objects. The respondent indicates the position on a scale(s) where he or she would rate an object. Most attitude theorists believe that attitudes vary along continua. thus. These simple scales do not allow for making fine distinctions in attitudes. II. For example. that an object possesses. it has only the properties of a nominal scale. are physiological measures. sorting. •Choice between two or more alternatives is another type of attitude measurement—it is assumed that the chosen object is preferred over the other(s). etc. Obtaining verbal statements from respondents generally requires that the respondent perform a task such as ranking. •Sorting might present the respondent with several product concepts typed on cards and require that the respondent arrange the cards into a number of piles or otherwise classify the product concepts. including persons. This type of self-rating scale merely classifies respondents into one of two categories. and. thus. B. Some examples of ratings scales follow. A.
Individuals generally choose from approximately five (although alternatives may range from three to nine) response alternatives: “strongly agree. This step in the questionnaire design is too often neglected by business researchers. The subject makes repeated judgments of the concept under investigation on each of the scales. The advantages and disadvantages of a Stapel scale. Items that are poor because they lack clarity or elicit mixed response patterns are eliminated from the final statement list. Numerical Scales Numerical scales have numbers as response options. The total score is the summation of the weights assigned to an individual’s total responses.3 shows a series of scales related to measuring attitudes towards jazz saxophone recordings. Exhibit 14. rather than “semantic space” or verbal descriptions. In the Likert procedure. -2. Stapel Scale Modern versions of the Stapel scale place a single adjective as a substitute for the semantic differential when it is difficult to create pairs of bipolar adjectives.1 shows some common wordings for category scales. C. to the answers. and the weight of five is assigned to this response. If a negative statement toward the object were given. the Stapel scale tends to be easier to conduct and administer. G. such as “good” and “bad. The numerical scale utilizes bipolar adjectives in the same manner as the semantic differential. Bipolar adjectives. +2. Constant Sum Scale An example of a constant-sum scale is as follows: . Exhibit 14. However. scores are 7. as well as the results. brand. store.A category scale is a more sensitive measure than a scale having only two response categories—it provides more information.” “uncertain. 1 or +3. With the Likert scale. which are not printed on the questionnaire. Business researchers have found the semantic differential versatile and have modified the use of the scale for business applications. respondents indicate their own attitudes by checking how strongly they agree or disagree with carefully constructed statements about the attitudinal object. 0. 3. -1. are very similar to those for a semantic differential. to identify categories (response positions).5 illustrates a Stapel scale used in a measurement of a retailer’s store image. followed by a series of seven-point bipolar rating scales. A weight is assigned to each position on the rating scale. Traditionally. If the scale items have five response positions. 2. the weights would be reversed and “strongly disagree” would be assigned the weight of 5. the scale is called a five-point numerical scale. Strong agreement indicates the most favorable attitudes on the statement. Exhibit 14. -3. The purpose of the item analysis is to ensure that final items evoke a wide response and discriminate among those with positive and negative attitudes. or other concept. 6. researchers assign scores or weights. Question wording is an extremely important factor in the usefulness of these scales. Replacing the bipolar adjectives with descriptive phrases is an adaptation in image studies.” and “strongly disagree. This popular attitude measurement technique consists of the identification of a product.” To measure the attitude. Semantic Differential The semantic differential is actually a series of attitude scales. 5. Many business researchers assume that the semantic differential provides interval data. 4. A single scale item on a summated rating scale is an ordinal scale. F. D.” “agree. Method of Summated Ratings: The Likert Scale The Likert scale is an extremely popular means for measuring attitudes.” “disagree. E. a large number of statements is generated and then an item analysis is performed. +1. but some critics argue that the data has only ordinal properties since the weights are arbitrary.” anchor both ends (or poles) of the scale.
•Behavioral differential: The behavioral differential instrument has been developed for measuring the behavioral intentions of subjects towards any object or category of objects. I. . The first stage is a ranking operation performed by judges who assign scale values to attitudinal statements.9 provides a summary of the attitude-rating techniques discussed in this section. This technique is easily understood by the respondents.6 shows a traditional graphic scale. The results will approximate interval measures. The disadvantage of the graphic scale is that there are no standard answers.” “I would write. Category scales to measure the behavioral component of an attitude ask a respondent’s “likelihood” of purchase or intention to perform some future action. The graphic scale has the advantage of allowing the researcher to choose any interval he wishes for purposes of scoring. •Paired comparisons: In paired comparisons the respondents are presented with two objects at a time and asked to pick the one they prefer. For example: A 25-year old woman sales representative Would _ : _ : _ : _ : _ : _ : _ : _ : _ : Would not Ask this person for advice. and the subjects indicate their behavioral intentions toward this object on a series of scales. Exhibit 14. V. RANKING Consumers often rank order their preferences. An ordinal scale may be developed by asking respondents to rank order (from most preferred to least preferred) a set of objects or attributes. The wording of statements used in these cases often includes phrases such as “I would recommend. Graphic Rating Scales A graphic rating scale presents respondents with a graphic continuum. MEASURING BEHAVIORAL INTENTION The behavioral component of an attitude involves the behavioral expectation of an individual toward an attitudinal object. the respondent’s score is determined by measuring the length (in millimeters) from one end of the continuum to the point marked by the respondent. Thurstone Interval Scale The construction of the Thurstone scale is a rather complex process that requires two stages. this and other picture response options enhance communication with respondents. This method is time consuming and costly and is rarely used in applied business research. Typically. The respondents are allowed to choose any point on the continuum to indicate their attitude. H. A frequently used variation on the graphic scale design is the scale ladder. Ranking objects with respect to one attribute is not difficult if only a few products are compared.” to indicate action tendencies. this technique becomes more complex. However. A description of the object to be judged is placed on the top of a sheet.” or “I would buy. as the number of stimuli increase.Divide 100 points among each of the following brands according to your preference for the brand: Brand A __________ Brand B __________ Brand C __________ This constant-sum scale works best with respondents having a higher education level. The second stage consists of asking subjects to respond to the attitudinal statements. Exhibit 14. IV.
OTHER METHODS OF ATTITUDE MEASUREMENT Attitudes.1)/2). With the growth of computer technology. subject to the imagination of the researcher. measurement of attitude is. IX. A comparative rating scale asks a respondent to rate a concept in comparison with a benchmark—in many cases. even if they are unfamiliar with the attitudinal object. when respondents are asked to provide sensitive or embarrassing information in a survey. the manager’s problem definition.M.g. the researcher must make a decision whether to use a standard of comparison. each question has two possible questions associated with it—one sensitive and one nonsensitive. Therefore. • What type of category labels. as hypothetical constructs. is a commonly used sorting technique. The choice is generally influenced by what is planned for the later stages of the research project. Bruskin’s A. an unbalanced scale may eliminate this type of “end piling. The researcher has many scaling options. it asks a respondent to rate a single concept in isolation. and the backward and forward linkages to other choices that have already been made (e. (Association-Identification-Measurement) technique.I. who confidentially determines which of the two questions will be answered (e. • Should a single measure or an index measure be used? The researcher’s conceptual definition will be helpful in making this choice. sorting. While estimates are subject to error. to an extent. techniques such as multidimensional scaling and conjoint analysis are used more frequently. will be used for the rating scale? We have discussed verbal labels. the number of comparisons increases geometrically (n*(n . • How many scale categories or response positions are required to accurately measure an attitude? The researcher must determine the number of meaningful positions that is best for each specific project. coins or cards) to be divided among items being tested. R. There are several issues that will be helpful to consider: • Is a ranking. VII. • Should a balanced or unbalanced rating scale be chosen? The fixed-alternative format may be balanced—with a neutral or indifferent point at the center of the scale—or unbalanced. numerical labels. rating. VIII. a respondent has not formed an attitude towards a concept.g.. H. if any. The respondent will answer “yes” or “no” to the question asked. Chapter 24 discusses some complex statistical techniques. A formula is used to estimate the proportion of “yes” answers to the sensitive question. A variation of the constant-sum technique uses physical counters (e. The choice of scale will be a function of the nature of the attitudinal object to be measured. by the toss of a coin). VI. other than the ones presented. telephone survey versus mail survey). which measures how well customers associate and identify elements of advertising with a product. The question asked is randomly selected by the respondent. The BBDO photo sort process (mentioned in Chapter 13) is another sorting technique used to measure the affective component of attitudes. There are several other techniques. Unbalanced scales may be used when the responses are expected to be distributed at one end of the scale.g. . SELECTING A MEASUREMENT SCALE: SOME PRACTICAL QUESTIONS There is no best scale that applies to all research projects. If the number of comparisons is too great. SORTING Sorting tasks require that respondents indicate their attitudes or beliefs by arranging items. The maturity and educational levels of the respondents and the required statistical analysis will influence this decision. • Should a monadic or comparative scale be used? If a scale is other than a ratio scale. If many respondents in the sample are expected to be unaware of the attitudinal object under investigation. “the ideal situation” presents a reference for comparison with the actual situation. the researcher may utilize randomized response questions. A monadic rating scale uses no such comparison... In this type of questionnaire. respondents may fatigue and no longer carefully discriminate among them. and simply cannot provide an answer. or choice technique best? The answer to this question is largely determined by the problem definition and especially by the type of statistical analysis that is desired. the respondent remains anonymous. RANDOMIZED RESPONSE QUESTIONS In special cases. and response bias is therefore reduced. are not measured directly. this problem may be eliminated by using a nonforced-choice scale that provides a “no opinion” category. The argument for forced choice is that people really do have attitudes.” • Should respondents be given a forced-choice scale or a nonforced-choice scale? In many situations. that can be utilized when a situation dictates. and unlisted choices.but as the number of items increases.
” . B. Questionnaire relevancy: A questionnaire is relevant if no unnecessary information is collected and if only the information needed to solve the marketing problem is obtained. A SURVEY IS ONLY AS GOOD AS THE QUESTION IT ASKS The importance of question wording is easily overlooked. Single-dichotomy or dichotomous-alternative questions: require that the respondents choose one of two alternatives. IV. the researcher must also be thinking about the types of statistical analysis that will be conducted. Also. II. and all possible omissions should be considered. several decisions must be made. fixed-alternative questions. B. or closed questions. difficult to answer. generally learn that their hasty efforts are inadequate. A. WHAT SHOULD BE ASKED? The specific questions to be asked will be a function of the previous decisions. There are various types of fixed-alternative questions. Open-ended response questions are free response questions.PHRASING QUESTIONS There are many ways to phrase questions and many standard question formats have been developed in previous research. Questionnaire accuracy. The cost of open-ended response questions is substantially higher than that of fixed-alternative questions. since the job of coding.if questions are lengthy. When designing the questionnaire. and nonirritating words. The answer can be a simple “yes” or “no” or a choice between “this” and “that. the researcher must be specific about data needs. QUESTIONNAIRE DESIGN: AN OVERVIEW OF THE MAJOR DECISIONS Relevance and accuracy are the two most basic criteria to be met if the questionnaire is to achieve the researcher’s purpose. III. take less time. Question wording and sequence substantially influence accuracy. Accuracy means that the information is reliable and valid. this chapter deals with these decisions. give the respondents specific. the researcher may find some unanticipated reaction toward the project. Fixed-alternative questions require less interviewer skill. A. no step-by-step procedure can be generalized. The later stages of the research process will have an important impact on the questionnaire wording. In contrast. open-ended response questions allow potential interviewer bias to influence the answer . By gaining free and uninhibited responses. Unfortunately. newcomers. To ensure information relevance. and are easier for the respondent to answer. or ego threatening. While it is generally believed that one should use simple. alternative responses and ask the respondent to choose the response closest to his or her viewpoint. editing. In order to achieve this. unambiguous. there is a high probability of biased answers. who naively believe that common sense and good grammar are all one needs to construct a questionnaire. and analyzing the data is quite extensive. They may also be useful at the beginning of an interview as they allow the respondent to warm up to the questioning process.even the best interviewer can take shortcuts in recording answers. Open-ended response questions pose some problem or topic and ask the respondents to answer in their own words. but questionnaire design is one of the most critical stages in the survey research process. unbiased. 1.Chapter 15 Questionnaire Design I. Respondents tend to be most cooperative when the subject of the research is interesting . limited. understandable. Open-ended response versus fixed-alternative questions: We may categorize two basic types of questions asked on the amount of freedom respondents are given in answering. Open-ended response questions are most beneficial when the researcher is conducting exploratory research. there should be a rationale for each item.
Invoking the status quo is a form of loading that results in bias because the majority of people tend to be resistant to change. an assurance of anonymity may help elicit honest responses to embarrassing questions. that is. which reverses the wording of attitudinal questions for 50 percent of the sample. Most questionnaires include a mixture of open-ended and closed questions. To control for this bias. Phrasing questions for mail. Such questions may result in a “bandwagon effect” which threatens the study’s validity. rather than stating why they chose a given product. split-ballot technique. b) The frequency determination question is a determinant choice question that asks for an answer about the general frequency of occurrence. In many cases.2. V. An introductory counterbiasing statement to a question. and personal interview surveys: In general. A. Or. a) Determinant choice questions ask a respondent to choose one and only one . that reassures respondents that their “embarrassing” behavior is not abnormal. Partial mention of alternatives is a variation of this phenomena. telephone. and the resulting answers may not portray the respondent’s true feelings. they may select a given alternative rather than thinking of the most correct alternative. may help yield truthful responses. Respondents. There should be no overlap among categories in the check list . Some questions invite only positive answers. only one dimension of an issue should be related to that alternative. there are no hard and fast rules. B. conversational language: Words used in a questionnaire should be readily understandable to all respondents. C. A question statement may be leading because it is phrased to reflect either the negative or positive aspects of an issue. Asking respondents “how often” leads them to generalize about their behavior and one is more likely to portray one’s ideal behavior rather than one’s average behavior. d) The check list question allows respondents to provide multiple answers to a single question. as a matter of convenience. Avoid complexity: use simple. Each form has unique benefits.response from among several possible alternatives. Multi-choice alternatives: there are several types. can be used. Avoid leading and loaded questions: Asking leading and loaded questions is a major source of bias in question wording.each alternative should be mutually exclusive. However. respondent boredom and fatigue are eliminated with a change of pace offered by a mixture of question types. the choices are adjectives that describe a particular object. Exhibit 15. The researcher should strive to ensure that there are sufficient response choices to include almost all possible answers. Also. c) Attitude rating scales. Loaded questions suggest social desirability or are emotionally charged. mail and telephone questions must be less complex than those utilized in personal interviews. such as the Likert scale. are discussed in chapter 13. may select an alternative among those presented. Some guidelines have been developed to avoid the most common mistakes. Leading questions suggest or imply certain answers. including a category lower than the answers you expect often helps to negate the potential bias caused by respondents avoiding an extreme category. In other cases. in addition. The technical jargon of top executives should be avoided. THE ART OF ASKING QUESTIONS In developing a questionnaire.1 illustrates how a question may be revised for a different medium. respondents are able to interpret which answers are most socially acceptable. . Questionnaires for telephone and personal interviews should be written in a “conversational” manner.
often. and other visual features displayed on the computer screen. Notice the logical order of questions which can help ensure the respondent’s cooperation.WHAT IS THE BEST QUESTION SEQUENCE? The order of questions may serve several functions for the researcher. LAYOUT OF INTERNET QUESTIONNAIRES A web questionnaire site should be easy to use. often the rate of return can be improved by adding the money. Filter questions minimize the asking of questions that may be inapplicable. VI. The researcher can design the questionnaire to make the interviewee’s job of following interconnected questions much easier by utilizing several forms.. Randomization of these items on a questionnaire helps to minimize this order bias. and pivot questions may be used to obtain information that the respondent may be reluctant to provide. etc. it refers to the researcher’s control over the background. Avoid ambiguity .the questions should be as specific as possible. special instructions. In mail questionnaires. VII. and can help eliminate any confusion or indecision. A more common practice is to pencil X’s or check marks on the printed questionnaires to indicate that the interviewer should start a series of repetitive questions at a certain point. Many of the guidelines for layout of paper questionnaires apply to Internet questionnaires.000?” If under. there also may be an anchoring effect. With respect to Internet surveys. and have a graphic look-and-feel that creates an overall experience that motivates the respondent to cooperate from start to finish. if the respondents’ curiosity is not aroused at the outset. have many different meanings. ready. .2 gives an example of a flow chart plan for a questionnaire.000?” If over. Questionnaires should be designed to appear as short as possible and experienced researchers have found that it pays to carefully phrase the title to be printed on the questionnaire. many important differences. It allows the researcher to design questionnaires that allow respondents to click on what they want rather than having to type answers or codes. have a logical flow. That is. Order bias: Order bias results from an alternative answer’s position in a set of answers or from the sequencing of questions. Order bias tends to distort survey results. Avoid double-barreled items: A question covering several items at once is referred to as a doublebarreled question and should always be avoided. D.C. A. Another mistake that question writers sometimes make is assuming that the respondent has previously thought about an issue -research that induces people to express attitudes on subjects that they do not ordinarily think about is meaningless. VIII. This technique is known as the funnel technique. For example. they can become disinterested and terminate the interview.000?” Exhibit 15. There is no need for the confusion that results in a double-barreled question. fonts. When using attitude scales. to improve the attractiveness and quality of the questionnaire. colors. For example. Thus. it is advisable to ask the general questions before the specific questions to obtain the freest of open-ended responses. WHAT IS THE BEST LAYOUT? The layout and attractiveness of the questionnaire are of crucial importance. ask “Is it over or under $50. E. Graphical User Interface (GUI) refers to software that provide an attractive and easy to use interface between a computer user and an application. There are. a respondent is asked “Is your family income over $30.Avoid making assumptions: The researcher should not place the respondent in a bind by including an implicit assumption in the question. the first concept measured tends to become a comparison point from which subsequent evaluations are made. and it allows researchers to understand the respondent’s frame of reference before asking more specific questions about the respondent’s particular level of information and intensity of opinion. and other tricks of the trade.be as specific as possible: Items on questionnaires are often ambiguous because they are too general. Rarely do marketing researchers print alternative question forms to eliminate problems arising from order bias. Use of such words should be avoided . ask “Is it over or under $10. that might have been used as an incentive. however. Specific questions tend to influence more general ones. Indefinite words such as frequently.
simpler solution. graphics. Initially a general statement. showing individual questions on single screens (Web pages) or on a scrolling basis with the entire questionnaire appearing on a one screen (Web page). is often used to format the questionnaire so the researcher will know how it should appear online. if the respondent needs to go to the next page a large arrow labeled “NEXT” might appear in color at the bottom of a screen. such as FrontPage or Netscape Composer. Even if the questionnaire designer’s computer and the respondents’ computer are compatible. such as the one shown in the Exhibit. it is important to provide a status bar or another visual indicator of questionnaire length. and other computer configuration issues so that questionnaire delivered is as compatible as possible. browser software. One sophisticated remedy is to use the first few questions on an Internet survey to ask about operating system. . Based on respondent’s answers to filter questions. and small graphics can make a questionnaire become more interesting and appealing.” increases the likelihood that the respondent will finish the entire sequence of questions. Because many Internet surveys offer no visual clues about the number of questions to be asked. A drop-down box. the display should advance smoothly. Provide appropriate links to allow users to go to the top and bottom parts of each section. If the entire questionnaire appears on the screen. misaligned text. a respondent’s computer settings for screen configuration (e. However. For example. the researcher is increasing his or her chances that response bias will occur. Clicking on a down-facing arrow makes the full range of choices appear. The paging layout going from screen to screen greatly facilitates skip patterns. When a scrolling questionnaire is long.Web publishing software. For example. there are several other layout issues. An Internet questionnaire uses special windows known as dialog boxes to display questions and record answers. avoiding the problems associated with splitting questions and response categories may be difficult. Scrolling down beyond the first few statements may not allow the respondent to simultaneously see both the statements at the end of the list and the response categories at the top of the grid.g. and strongly disagree at the beginning of the question. Another. One thing to remember is that for people with very powerful computers sophisticated graphics are not a problem. For example. using multiple category or section headings is helpful to respondents. Whether a page-by-page or scrolling format is utilized push buttons with labels should clearly describe the actions to be taken. but they may present problems for respondents with older computers and/or low bandwidth Internet connections. is to limit the horizontal width of the questions to 70 characters or less to decrease the likelihood of wraparound text. enabling the respondent to navigate through the questionnaire more easily than having to scroll through the entire document. The scrolling layout gives the respondent the ability to scroll down to read any portion of the questionnaire at any time and because there are no page boundaries problems can arise. sound. The first layout decision is whether the questionnaire will appear page-by-page. Suppose a Likert scale consists of 15 statements and a grid-format layout places the response categories strongly agree. For example using a partially filled rectangular box as a visual symbol and a statement such as: “The status bar at top right indicates approximately what portion of the survey you have completed. the respondent knows how many questions he or she must answer. such as “strongly agree”. colored headings. using a textured background. If the first choice in a list. is a space saving device that allows the researcher to provide a list of responses that are hidden from view. For example. as if it were a piece of paper being moved up or down. perhaps “please select” or “click here” is shown. if at all. many respondents’ computers are not powerful enough to have complex features operationally delivered at a satisfactory speed. several features of a respondent’s computer may influence the appearance of an Internet questionnaire. or other visual appearance problems. The Exhibit entitled Alternative Ways of Displaying Internet Questions in the textbook portrays four common ways of displaying questions on a computer screen. Drop-down boxes may present a problem for individuals with minimal computer skills who may not know how to reveal hidden responses behind a drop-down menu or how to move from one option to another in a moving-bar menu. However. 640 x 480 pixels versus 800 x 600 pixels) may result in questions that are not fully visible on the respondent’s screen. With a paper questionnaire. and other special features that the Internet makes possible. The fact that the questionnaire image the researcher/designer constructs on his or her computer may be different from the questionnaire image on the respondent’s computer should always be considered when designing Internet surveys. On paper Likert scales are often shown in a multiple-grid format. Many responses to Internet questions require the respondent to activate his or her answer by clicking a radio button on a response. is shown with other responses hidden. disagree. clicking on an alternative response deactivates the first choice and replaces it with the new response. agree. animation. Layout decisions must also be made about the use of color. Like push buttons on automobile radios. the computer automatically inserts relevant questions on subsequent pages.
Most of these programs have hidden skip logic so respondents never see any evidence of skips. Can respondents answer the questions easily? . blocks of questions. the questionnaire is “tried out” on a group selected on a convenience basis and similar in make-up to the one that ultimately will be sampled. Of course. pop-up boxes may be use to provide a privacy statement.Checklist questions may be placed in a check box where several. respondents cannot skip over questions as they do in mail surveys. Questionnaire software with Boolean skip and branching logic is readily available. or can be allowed to stop mid-questionnaire and come back later to finish. Some research suppliers offer interactive help desks to solve problems that might be encountered in completing a questionnaire. With forced answering software. Variable piping software allows variables. Which alternative forms of questions work best? Pretests provide the means for testing the sampling procedure and also provide estimates for the response rates for the mail surveys and completion rates for telephone surveys. HOW MUCH PRETESTING AND REVISING IS NECESSARY Usually. This is called error trapping. If the respondent fails to answer a question. The software may insert a bold-faced error message on the question screen or insert a pop-up box instructing the respondent how to continue. Respondents can be blocked from backing up. it is inappropriate because it is does not solve the marketing problem. ranking. an immediate error message appears. However. Can the questionnaire format be followed by the interviewer? . or rating questions. and response alternative order from respondent to respondent. For example. Pretesting allows the researcher to determine if the respondents have any difficulty understanding the questionnaire or if there are any ambiguous or biased questions. It is best if the questions the respondent sees always flow in numerical sequence. X. Pretests are typically conducted to answer questions about the questionnaire such as: . a respondent may answer questions 1 thru 11 and then next see a question numbered 15 because of the skip logic. This process can save the potential disaster of administering an invalid questionnaire to several hundred individuals. it is not necessary to get a statistical sample for pretesting. The researchers can use software to control of the questionnaire flow. or all responses may be checked. Some open-ended boxes are designed so numbers can be entered for frequency response. Open-ended boxes may be designed as One line Text Boxes or Scrolling Text Boxes depending on the breadth of the expected answer. Internet software allows questioning to branch off into two or more different lines depending on each respondent’s answer to skip or filter questions. Other software can randomly rotate questions. or answers it with an incorrect type of response. none. Openended boxes require respondents to use their keyboards to enter text for open-ended questions. IX. Are the questions clear and easy to understand? . some programs number all potential questions in a numerical order and the respondent sees only the number on the questions he or she answers. Pop-up boxes are message boxes that can be used to provide highlighted information. such as answers from previous questions. Software can systematically or randomly manipulate the question a respondent sees. A preliminary tabulation of the pretest results often illustrates that while a question is easily comprehended and answered by the respondent. open-ended response questions require that respondents have both the skill and the willingness to keyboard lengthy answers into the computer. Does the questionnaire flow naturally and conversationally? . Thus. The program will not let them continue if they fail to answer a question. DESIGNING QUESTIONNAIRES FOR GLOBAL MARKETS . to be inserted in to unfolding questions.
International business researchers must take cultural factors into account when designing questionnaires. The back translator is often a person whose native tongue is the language that will be used on the questionnaire. Back Translation is the process of translating the questionnaire from one language to another and then having it translated back again by an second. . International questionnaires are often back translated. The most widespread problem involves translation into another language. independent translator.