Marketing Research


Customer Groups Consumers Employees Shareholders Suppliers Uncontrolla ble Environment factors Economy Technology Competition Regulations Political factors Social & Cultural factors

Controllabl e Environme nt factors Product Price Promotio n Distributi on

Marketing Research
Assessing informatio n needs Providing Informatio n Marketing Decision Making

MarketIngManagers Market Segmentation Target market selection Marketing


(or series of iterative steps), and followed often when management is faced with a “problem” and/or “opportunity”, management needs further information in order to make a decision – the need for market(ing) research is an issue that is likely to need addressing...

Research is a process

The question is “when to conduct market(ing) research?”

Conduct Costs Does the value Market of the research Research exceed the cost? No Do not conduct market research! Example issues: (1) What is our market share? (2) Will people drink tomato soup from a plastic jar? (3) Whose machine tools do our potential customers buy? (4) Which medicine is more preferred for a decease? .When to Conduct Market(ing) Research Yes Time Constraints Is sufficient time available? No Yes Yes Yes Availability of Data Is the information on hand inadequate? No Nature of Decision Is the decision of considerable importance? No Benefits vs.

When Research Should be Done •If it clarifies problems or investigates changes in the marketplace that can directly impact your product responsibility •If it resolves your selection of alternative courses of marketing action to achieve key marketing objectives •If it helps you gain a meaningful competitive advantage •If it allows you to stay abreast of .

Determine research How much is already known? Can hypothesis be formulated ? What types of questions need Descriptive and causal to be answered ? What type of study best address research questions ? Exploratory / conclusive design – 3. Formulate problem additional Typical Questions What is purpose of study .solve a problem? Identify opportunity? Is background info necessary? What info is needed to make decision? How will info be utilized? Should research be conducted? 2.Questions addressing the various stages of the Research Process Stage in the Process 1. Determine data collection Can existing data be used to advantage? .

Questions addressing the various stages of the Research Process Stage in the Process Questions Typical 4. Design sample & collect population Probability What Who is target population? Is list of elements available? Is sample necessary? Is sample desirable? How large should sample be? operational procedures will be followed? What methods will be used to ensure quality of data collected? . Design data collection items used in study be made be used? Should structure or unstructured collecting data? Should purpose of known to respondents? Should rating scale What type of rating scale would be most appropriate? forms 5.

The research process .

.The research process Is a set of iterative steps and relationships....

The Concept of Total Error All research has error and this impacts on the research outcome – its usability and accuracy Poorly Written Research Report Poor Logic Poor problem definition formulation Improper use of Statistical Procedures Total Error Poor data collection methods Inadequate sample size Inadequate sample design .

and that research questions are objectives that fit underneath the research problem. “research problem”.... ......Problem definition steps Management problem definition process Research problem definition process Please note that sometimes this is called Research question or research problem...

The management problem may include: – Focuses management on has – Symptoms of failure to achieve an objective.e. Must decide how to seize opportunity (opportunity identification) Formulate Formulate Management Problem Research Problem .Problem Definition • Management problem: the decision that to make and is action oriented (i. once the information is obtained a course of action will be required)…. Must select course of action to regain it. – Symptoms of likelihood of achieving objective.

g. read information) before defining the research problem – Researchers ask yourself: is the issue that management is seeking answers to merely a symptom of X? – Remember the iceberg principle • • The symptoms are what we can see (e. and unbiased information that manages can use to solve their marketing management problems. ask questions.Problem Definition • How to provide relevant. . accurate.. falling sales) The issues (causes) are generally what we cant see and generally the issue (below the surface) is what needs investigating and therefore forms the research problem ………….. • The research problem is information The research problem: oriented and researchers need to do some investigation (e.g.

designs. current image of the store. alternative package Increase store traffic. so we have a problem.Examples of Management Problem Problem Develop package for new effectiveness of Research Evaluate product. Measure Ok. how do we write the problem definition???? .

search industry info.. This is an iterative and ....So you think you have a problem – how do you write it???? Management Problem Research Problem Information oriented To determine consumer preferences and purchase intentions for the proposed new product To determine the effectiveness of the current advertising campaign To determine the price elasticity of demand and the impact of sales and profits of various levels of price changes Should a new product be Decision / action oriented introduced? Should the advertising campaign be changed? Should the price of the brand be increased? To help you develop and write the research problem and research objectives you should consult other sources of information: ask questions. rely on experience.. academic journals (theory).

The problem definition process How much is this information worth?????? Estimate the value of information .

Marketing Research Problem identification research Market Potential Research Market Share Research Image Research Market Characteristics Research Sales Analysis Research For casting Research Business Trends Research Problem solving research Segmenting Research Product Research Pricing Research Promotion Research Distribution Research .

price elasticity. find out segments. . packaging. selection of Product Research : modification. design . and repositioning Pricing Research : price policy. positioning test .Problem solving research Segmenting Research: response of target segment Basis of segmentation. customer response Promotion Research: other tools. relationship with media decision . effectiveness Promotion budget. line policy. testing.

2nd Session .

” . refine.Marketing Research Defined (AMA) “Marketing research is the function which links consumers and the consumer to the organization through informationInformation used to identify and define marketing problems. and evaluate marketing actions . monitor marketing performance. and improve our understanding of marketing as a process. generate.

The role of marketing research within the marketing system .

THE ROLE OF MARKETING RESEARCH MARKETING RESEARCH A FORMAL COMMUNICATION LINK WITH ENVIRONMENT PROVIDE ACCURATE AND USEFUL a) specifying b) collecting c) analyzing d) interpreting a) planning b) problem-solving c) control FOR .

NATURE OF MARKETING RESEARCH Applied/Problem solving research Often based on cost-benefit analysis Vital for implementation of marketing concept Value of information declines with time Dynamic (ongoing) .

DRIVERS OF MARKETING RESEARCH Shift from production to customerorientation Declining cost of unit information (digital age) Increase intensity of competition Globalization Technology and commercialization .

online panels ‘Value for money’ marketing ‘Strategic’ consultants ‘Respondent’ rewards . e.Factors shaping the Marketing Research Industry Competitor Intelligence Low cost survey providers Surveys to generate sales & PR Customer Analytics The nature and future of Marketing Research Internet.g.

Change: To keep updated with changes in ¢ ¢ ¢ ¢ . Credibility: To increase the believability of promotional messages among customers. investigate new target markets. Competition: To identify primary competitors and pinpoint their strengths and weaknesses.Reasons for Doing Marketing Research: The Five Cs ¢ Customers: To determine how well customer needs are being met. Confidence: To reduce the perceived risk in making marketing decisions. and assess and test new services and facilities.

¢ Cost: The cost of the research is too high.Reasons for Not Doing Marketing Research ¢ Timing: It will take to much time. ¢ Reliability: There is no reliable research method available for doing the research. ¢ Competitive intelligence: There is a fear that competitors will learn about the organization’s .

Five Key Requirements of Marketing Research Information ¢ Utility: Does it us? ¢ Timeliness: available in time? ¢ Cost-effectiveness: outweigh ¢ Accuracy: Can we use it? apply to Will it be Do the benefits the costs? Is it accurate? .

Classification of marketing research .

Examples of problem-solving research .

Problem Definition Process E n v ir o n m e n t a l C o n t e x t o f t h e p r o b le m T a s k s in v o lv e d in p r o b le m d e fin it io n D is c u s s io n w it h d e c is io n m a k e r s I n t e r v ie w s w it h e x p e rts S e c o n d a ry d a ta a n a ly s is Q u a lit a t iv e re s e a rc h Management decision problem Marketing research problem .

Factors to Consider Environmental Context •Past information and forecasts •Resources and constraints •Objectives (organizational & decision maker) •Buyer behavior •Legal environment •Economic environment •Marketing and technological skills .

Developing a marketing strategy for the brand .g.Defining the Research Problem Allow the researcher to obtain all the information needed to address the management decision problem Guide the researcher in formulating the research design A broad definition does not provide clear guidelines for the subsequent steps involved in the project e.

So you think you have a problem – how do you write it???? Management Problem Research Problem Information oriented To determine consumer preferences and purchase intentions for the proposed new product To determine the effectiveness of the current advertising campaign To determine the price elasticity of demand and the impact of sales and profits of various levels of price changes Should a new product be Decision / action oriented introduced? Should the advertising campaign be changed? Should the price of the brand be increased? .

Define Research Design A framework or blueprint for conducting the marketing research project. Details the procedures necessary for obtaining the information needed to structure or solve marketing research problems .

A Classification of Marketing Research Designs Research Design Exploratory Research Design Conclusive Research Design Descriptive Research Causal Research Cross-Sectional Design Longitudinal Design .

Information needed defined loosely. . Research process flexible/unstructured. Analysis of primary data is qualitative. Sample is small and nonrepresentative. Sample is large and representative. Research process is formal and structured.Differences Between Exploratory and Conclusive Research Exploratory Objective: Characteristics: To provide insights. Followed by conclusive research. Data Analysis is quantitative. Findings: Outcome: Tentative. Conclusive. Conclusive Test hypothesis/examine relationships. Information needed is clearly defined. Findings input into decision making. understandings.

Establishing priorities for further research Methods Used : case studies secondary data focus groups qualitative research When done? Generally initial research conducted to clarify and define the nature of a problem Does not provide conclusive evidence : Subsequent research expected . versatile.Exploratory Research: Overview Characteristics : flexible. Formulating problems more precisely. but not conclusive Useful for : discovery of ideas and insights. Identifying alternative courses of action.

Descriptive Research: Overview Characteristics :
Describes characteristics of a population or phenomenon Some understanding of the nature of the problem preplanned, structured, conclusive Useful for : describing market characteristics or functions Methods Used : Surveys (primary data) panels scanner data (secondary data) When Used: Often a follow-up to exploratory research Examples include: Market segmentation studies, i.e., describe characteristics of various groups Determining perceptions of product characteristics Price and promotion elasticity studies

Examples of Descriptive Studies
•Market studies that describe the size of the market, buying power of the consumers, availability of distributors, and consumer profiles •Market share studies that determine the proportion of total sales perceived by a company and its competitors •Sales analysis studies that describe sales by geographic region, product line, type of account size of account •Image studies that determine consumer perceptions of the firm and its products •Product usage studies that describe consumption patterns •Distribution studies that determine traffic flow patterns and the number and location of distributors •Pricing studies that describe the range and frequency of price changes and probable response to proposed price changes •Advertising studies that describe media consumption habits and audience profiles for specific television programs and magazines

A Comparison of Basic Research Designs

Objective: Discovery of ideas

Describes market characteristics

Determine cause and effect


Flexible, versatile. Front end research.

Prior formulation of hypothesis. Planned, structured design

Manipulate independent variables. Control of other variables.


Secondary data

Experiments Surveys

Classification of Marketing Research Data Marketing Research Data Secondary Data Primary Data Qualitative Data Quantitative Data Descriptive Causal Survey Data Observational & Other Data Experimental Data .

Relationship among Exploratory. Descriptive and causal Research .

3rd Session .

Sampling Design Management information systems Recom mendations Problem definition Exploratory Data collection analysis & Research design Descriptive Causal Sampling Non-probability Probability .

.Sample or Census A population is the aggregate of all the elements that share some common set of characteristics. The population parameters are typically numbers. and that comprise the universe for the purpose of the marketing research problem. such as the proportion of consumers who are loyal to a particular brand of toothpaste.

called statistics. Sample characteristics. The population parameters can be calculated directly in a straightforward way after the census is enumerated (specify individually). The inferences that link sample characteristics and . are then used to make inferences about the population parameters. A sample is a subgroup of the population selected for participation in the study.Sample or Census A census involves a complete enumeration of the elements of a population.

Sample Versus Census Condition favoring the use of Sample Census Budget Large Time Available Long Population Large Variance in Characteristics Large Small Short Small Small .

Sampling is the process of selecting a sufficient number of elements from the population so that by studying the sample. more representative the sample is of the population. and understanding the properties or characteristics of the sample subjects. it would be possible to generalise the properties or characteristics to the population elements. the more generalisable are the findings of the research .

events or things of interest that the researcher wishes to investigate N Population element – single member of the population Sampling frame – list of all elements or the population from which the sample is drawn Sample (ing) – subset of the population selected for the specific research study .n Sample unit (subject) – single element selected in the sample.Sampling design – key terms Population – entire group of people. could be a group ( could be a two stage process) Census – an investigation of all individual elements that make up the population .

Why sample? time cost accuracy population may be difficult to access greater depth of information .

Managerial objectives of sampling Representative Reliable efficient as time permits .

response error – the statistical difference between a survey that includes only those who responded and a perfect survey that would also include those who failed to respond (occurs between the planned sample and the respondents .an error that occurs when certain sample elements are not listed or are not accurately represented in a sampling frame (occurs between the population and sampling frame) Random sampling error – occurs between the sampling frame and the planned sample for study Non .Errors associated with sampling Sampling frame error .

individuals or other units Step 3: Choose sampling technique/method method of selecting the sampling units Probability (random) vs. institutions. analysis required if probability sampling – based on statistical determination of sample size Step 5: Identify and select sample unit (subject) follow procedures based on sampling technique selected .Sampling design process Step 1: Define Population Entire group under study as defined by research objectives Step 2: Establish Sampling Frame list of sampling units from which a sample will be drawn. the list could consist of geographic areas. cost. non probability (non-random) Step 4: Determine sample size if non-probability sampling method –involves some judgement based on time.

Classification of Sampling Techniques Sampling Techniques Probability Sampling Techniques Nonprobability Sampling Techniques Convenience Sampling Judgmental Sampling Quota Sampling Snowball Sampling Simple Random Sampling Systematic Sampling Stratified Sampling Cluster Sampling Other Sampling Techniques .

then use non-probability .Non Probability Sampling each sampling unit of the population being studied does not have an equal chance of being included in the study (due to the way the sample is selected) non-random (selection process is subjective) researchers rely heavily on personal judgement projecting the findings beyond the sample is statistically inappropriate is less concerned about generalisability.time . other factors are more important . preliminary information .

Non Probability Sampling Common sampling approaches convenience judgement quota snowball .

pedestrian areas Acceptable only in pre .Convenience Sample Also known as haphazard or accidental sampling based on convenient availability of sampling units sample units happen to be in a certain place at certain time – high traffic locations – shopping malls.test/exploration phase when further research will use probability sampling Representativeness highly uncertain Quota sampling can reduce some of the sample selection error .

Judgement Sampling An experienced individual (could be the researchers) selects the sample based on personal judgement about some appropriate characteristics suited to the study Focus group studies use this method .

Quota Samples Various subgroups in a population are represented based on pertinent characteristics Haphazard selection of respondents may introduce bias Similar to stratified random sampling .

Snowball Sampling Judgement sample that relies on researchers ability to locate an initial set of respondents with the desired characteristics. these individuals are then used as informants to identify others with desired characteristic Acceptable when sample units are difficult to locate Advantages reduced sample size and costs .

Probability Sampling In a probability sample each element in the population has some known chance or probability of being included in the sample Used when the representativeness of the sample is important for generalisability of results Random selection eliminating bias of sample thus .

statistical efficiency same sample size and smaller standard error of the mean is obtained economic efficiency precision refers to the level of uncertainty about the characteristics being measured precision is inversely related to sampling error precision is positively related to cost .Probability Sampling cont.

Types of probability sampling Simple random sample Systematic sampling Stratified sampling proportionate disproportionate Cluster sampling Area sampling .


putting all name in a hat and drawing out a sample of 100 (size has been statistically calculated) Random numbers Need to begin with a complete list of the population – sometimes difficult to obtain .Simple Random Sampling Assures each element in the population of an equal chance of being included in the sample Blind draw .

Systematic Sampling A starting point is selected by a random process and then every nth number on the list is selected Calculate skip interval = population list size/ sample size (size has been statistically calculated) Danger of periodicity – if list has a systematic pattern Can be more representative than a simple random sample .

Stratified Sampling Simple random sub samples are drawn from within each stratum in the population that are more or less equal on some characteristic Greater degree of representativeness Two types proportionate .sample size of each stratum is relative to the size of each stratum in the population disproportionate –sample size of each stratum does not reflect their relative proportions in the population .

Cluster Sampling divides the population into groups (clusters). any one of which can be considered a representative sample an economically efficient technique in which the primary sampling unit is not the individual element but a large cluster of elements clusters are selected randomly random sample from within each cluster .

easier to implement than SRS. results  rojectable p Selection bias. no assurance of representativeness Time­consuming Difficult to construct sampling frame.Nonprobability Sampling   onvenience sampling C   udgmental sampling J   uota sampling Q   nowball sampling S Probability sampling   imple random sampling S (SRS)   ystematic sampling S Technique Strengths Least expensive. convenient. most convenient Low cost. sampling frame not necessary Include all important subpopulations. cost effective Difficult to select relevant stratification variables. not time­consuming Sample can be controlled for certain characteristics Can estimate rare characteristics Easily understood. expensive. expensive Imprecise. not recommended for descriptive or causal research Does not allow generalization. lower precision. difficult to compute and interpret results . r Can decrease  epresentativeness r Weaknesses Stratified sampling   luster sampling C Can increase representativeness. subjective Selection bias. least time­consuming. sample not representative. no assurance of  epresentativeness. not feasible to stratify on many variables. precision Easy to implement.

nonprobability sampling Probability sampling Conclusive Larger sampling errors Evaluation Criteria Nature of research Non-probability sampling Exploratory Relative magnitude Larger non-sampling of sampling and error non-sampling error Population variability Statistical Considerations Sophistication Needed Time Budget Needed Low [Homogeneous] Unfavorable Low Relatively shorter Low High [Heterogeneous] Favorable High Relatively Longer High .Choosing probability vs.

Selecting an Appropriate Design degree of accuracy resources time advance knowledge of the population national versus local projects need for statistical analysis .

4 .Session .

The rules for assigning numbers should be standardized and applied uniformly. One-to-one correspondence between the numbers and the characteristics being measured. Rules must not change over objects or .Measurement and Scaling Measurement means assigning numbers or other symbols to characteristics of objects according to certain pre-specified rules.

Consider an attitude scale from 1 to 100. Scaling is the process of placing the respondents on a continuum with respect to their attitude toward Measurement and Scaling . and 100 = Extremely Favorable. Measurement is the actual assignment of a number from 1 to 100 to each respondent. Each respondent is assigned a number from 1 to 100. with 1 = Extremely Unfavorable.Scaling involves creating a continuum upon which measured objects are located.

2 14.1 13.Scale Nominal Primary Scales of Measurement Numbers Assigned to Runners Rank Order of Winners Third place Second place 9.1 First place 9. in 15.2 Ratio Time to Finish.4 .6 Finish 7 8 3 Ordinal Finish Interval Performance Rating on a 0 to 10 Scale 8.

The numbers do not reflect the amount of the characteristic possessed by the objects. . Only a limited number of statistics. are permissible. Primary Scales of Measurement Nominal as labels or Scale numbers serve only tags for When used for identification. e. all of which are based on frequency counts. there is a strict one-toone correspondence between the numbers and the objects.g..The identifying and classifying objects. The only permissible operation on the numbers in a nominal scale is counting.

Kmart 4. Macy’s 3. Store Ordinal Scale Preference Rankings Interval Scale Preference Ratings 1-7 11-17 Ratio Scale $ spent last 3 months 1. Rich’s 5.Wal-Mart 7 2 8 3 1 5 9 6 4 10 79 25 82 30 10 53 95 61 45 115 5 7 4 6 7 5 4 5 6 2 15 17 14 16 17 15 14 15 16 12 0 200 0 100 250 35 0 100 0 10 . Target 8. J.Illustration of Primary Scales of Measurement Nominal Scale No. Penney 6. Lord & Taylor 2.C. Neiman Marcus 7. Sears 10. Saks Fifth Avenue 9.

Primary Scales of Measurement Ordinal Scale
• A ranking scale in which numbers are assigned to objects to indicate the relative extent to which the objects possess some characteristic. • Can determine whether an object has more or less of a characteristic than some other object, but not how much more or less. • Any series of numbers can be assigned that preserves the ordered relationships between the objects. • In addition to the counting operation allowable for nominal scale data, ordinal scales permit the use of statistics based on centiles, e.g., percentile, quartile, median.

Primary Scales of Measurement • Numerically equal distances on the scale represent equal values in the characteristic being measured. • It permits between objects. comparison of the differences

Interval Scale

• The location of the zero point is not fixed. Both the zero point and the units of measurement are arbitrary. • Any positive linear transformation of the form y = a + bx will preserve the properties of the scale. • It is not meaningful to take ratios of scale values. • Statistical techniques that may be used include

• Possesses all the properties of the nominal, ordinal, and interval scales. • It has an absolute zero point.

Primary Scales of Measurement Ratio Scale

• It is meaningful to compute ratios of scale values. • Only proportionate transformations of the form y = bx, where b is a positive constant, are allowed.

store Percentages. market median position. standard Geometric mean.. harmonic mean Productmoment Coefficient of variation . mean. numbering of football players Nos. index Age.. relative positions rankings of teams of objects but not in a tournament the magnitude of differences between them Differences Temperature between objects (Fahrenheit) Zero point is fixed. sales. Length. indicate the Quality rankings. income. types mode Preference Percentile.Primary Scales of Measurement Scale Nominal Basic Characteristics Numbers identify & classify objects Common Examples Social Security nos. social class Statistics Inferential Chi-square. weight ratios of scale values can be compared Marketing Permissible Examples Descriptive Brand nos. Friedman ANOVA Ordinal Interval Ratio Attitudes. rankings. costs Range. opinions. binomial test Rank-order correlation.

A Classification of Scaling Techniques Scaling Techniques Comparative Scales Noncomparative Scales Paired Comparison Rank Order Constant Sum Q-Sort and Other Procedures Continuous Itemized Rating Scales Rating Scales Likert Semantic Differential Stapel .

• In non-comparative scales. each object is scaled independently of the others in the stimulus set. The resulting data are generally assumed to be . Comparative scale data must be interpreted in relative terms and have only ordinal or rank order properties.A Comparison of Scaling Techniques • Comparative scales involve the direct comparison of stimulus objects.

• Easily understood and can be applied. • Involve assumptions. • fewer theoretical Tend to reduce halo or carryover .Relative Advantages of Comparative Scales • Small differences between stimulus objects can be detected. • all Same known reference points for respondents.

.Relative Disadvantages of Comparative Scales Ordinal nature of the data Inability to generalize beyond the stimulus objects scaled.

. • Under the assumption of transitivity. • The data obtained are ordinal in nature.• objects and asked to select one according to some criterion. Comparative Scaling Techniques Paired Comparison with two Scaling A respondent is presented • Paired comparison scaling is the most widely-used comparative scaling technique. it is possible to convert paired comparison data to a rank order.

please indicate which one of the two brands of shampoo you would prefer for personal use.Obtaining Shampoo Preferences Using We are going to present you with ten pairs of shampoo Paired Comparisons Instructions: brands. For each pair. A 0 means that the row brand was preferred over the column brand. . bThe number of times a brand was preferred is obtained by summing the 1s in each column. Recording Form:   J hirmack Finesse 0 J hirmack Finesse Vidal Sassoon Head & Shoulders Pert 1a 1 0 1 Vidal Head & Sassoon Shoulders 0 1 0 1 1 0 0 1 Pert  0  0  1  0    1  1 0 1 Number of Times 3 2 0 4 Preferredb a A 1 in a particular box means that the brand in that column was preferred over the brand in the corresponding row.

. The consumer is asked to sample two different products and select the one with the most appealing taste. self-perception and brand reputation are very important factors in the consumer’s purchasing decision. New Coke was heavily favored in blind paired comparison taste tests.000 responses is considered an adequate sample. The test is done in private and a minimum of 1. The introduction of New Coke illustrates this point. but its introduction was less than successful. may not be a good indicator of performance in the marketplace.Paired Comparison Selling The most common method of taste testing is paired comparison. A blind taste test for a soft drink. because image plays a major role in the purchase of Coke. where imagery.

Only (n .Comparative Scaling Techniques Rank Order Scaling Respondents are presented with several objects simultaneously and asked to order or rank them according to some criterion. rank order scaling also results in ordinal data. . Furthermore.1) scaling decisions need be made in rank order scaling. It is possible that the respondent may dislike the brand ranked 1 in an absolute sense.

There is no right or wrong answer.Preference for Toothpaste Brands Using Rank Order Scaling Instructions: Rank the various brands of toothpaste in order of preference. Begin by picking out the one brand that you like most and assign it a number 1. The least preferred brand should be assigned a rank of 10. Then find the second most preferred brand and assign it a number 2. No two brands should receive the same rank number. The criterion of preference is entirely up to you. Just try to be consistent. Continue this procedure until you have ranked all the brands of toothpaste in order of preference. .

Stripe Rank Order _________ _________ _________ _________ _________ _________ _________ _________ _________ _________ . Pepsodent 9. Aim 4. Ultra Brite 7. Close Up 8. Gleem 5. Plus White 10.Preference for Toothpaste Brands Using Rank Order Scaling Form Brand 1. Crest 2. Colgate 3. Sensodyne 6.

Hence. the respondent assigns it zero points. If an attribute is twice as important as some other attribute. the name of the scale. such as 100 points to attributes of a product to reflect their importance.Comparative Scaling Techniques Constant Sum Scaling Respondents allocate a constant sum of units. it receives twice as many points. If an attribute is unimportant. The sum of all the points is 100. .

.Importance of Bathing Soap Attributes Using a Constant Sum Scale Instructions On the next slide. it should receive twice as many points. If an attribute is twice as important as some other attribute. If an attribute is not at all important. assign it zero points. Please allocate 100 points among the attributes so that your allocation reflects the relative importance you attach to each attribute. there are eight attributes of bathing soaps. The more points an attribute receives. the more important the attribute is.

Cleaning Power 4 17 7 9 19 9 20 15 100 . Lather 53 17 3. Fragrance 5 3 6. Mildness 2 4 3 9 2. Packaging 13 60 7. Shrinkage 9 0 4. Moisturizing 100 100 Sum 8.Form Importance of Bathing Soap Attributes Using a Constant Sum Scale Average Responses of Three Segments Attribute Segment I Segment II8 Segment III 2 1. Price 7 5 5.

.Q – Sort Scaling A comparative scaling technique that uses a rank order procedure to sort objects based on similarity with respect to some criterion.

Session .5 .

Non-comparative techniques consist of continuous and itemized rating scales. .Non .comparative Scaling Techniques Respondents evaluate only one object at a time. and for this reason noncomparative scales are often referred to as monadic scales.

....................... The form of the continuous scale may vary considerably...................................-I .   How would you rate Sears as a department store? Version 1 Probably the worst .Probably the best 0 10 20 30 40 50 60 70 80 90 100   Version 3 Very bad Neither good Very good nor bad Probably the worst ....Probably the best   Version 2 Probably the worst ....................Continuous Rating Scale Respondents rate the objects by placing a mark at the appropriate position on a line that runs from one extreme of the criterion variable to the other..-................-Probably the best ...-I ........-......-I ...........

sex.” A group of up to 400 respondents is presented with TV or radio spots or advertising copy. . enabling the researcher to view the respondents' scores immediately. they are superimposed on a video screen. which tabulates second-bysecond response profiles.RATE: Rapid Analysis and Testing Environment A relatively new research tool. the perception analyzer. The response scores can be broken down by categories. Each participant is given a dial and instructed to continuously record his or her reaction to the material being tested. income. The measuring device consists of a dial that contains a 100-point range. The responses are also stored in a permanent data file for use in further analysis. such as age. As the respondents turn the dials. As the results are recorded by the computer. or product usage. the information is fed to a computer. provides continuous measurement of “gut reaction.

semantic differential. and the respondents are required to select the specified category that best describes the object being rated.Itemized Rating Scales The respondents are provided with a scale that has a number or brief description associated with each category. The categories are ordered in terms of scale position. and . The commonly used itemized rating scales are the Likert.

  SD D Neither A SA A or D   1. . or a total (summated) score can be calculated. I like to shop at Sears.Likert Scale The Likert scale requires the respondents to indicate a degree of agreement or disagreement with each of a series of statements about the stimulus objects.   1 1 1 2X 2X 2 3 3 3X 4 4 4 5 5 5 The analysis can be conducted on an item-by-item basis (profile analysis). Sears has poor in-store service.   3. the categories assigned to the negative statements by the respondents should be scored by reversing the scale. Sears sells high quality merchandise. When arriving at a total score.   2.

. This controls the tendency of some respondents.or left-hand sides without reading the labels.   Semantic Differential Scale SEARS IS: Powerful --:--:--:--:-X-:--:--: Weak Unreliable --:--:--:--:--:-X-:--: Reliable Modern --:--:--:--:--:--:-X-: Old-fashioned The negative adjective or phrase sometimes appears at the left side of the scale and sometimes at the right. particularly those with very positive or very negative attitudes.The semantic differential is a seven-point rating scale with end points associated with bipolar labels that have semantic meaning. Individual items on a semantic differential scale may be scored on either a -3 to +3 or a 1 to 7 scale. to mark the right.

Concepts. Person Concepts. and Product Concepts 1) Rugged 2) Excitable 3) Uncomfortable 4) Dominating 5) Thrifty 6) Pleasant 7) Contemporary 8) Organized 9) Rational 10) Youthful :---:---:---:---:---:---:---: Delicate :---:---:---:---:---:---:---: Calm :---:---:---:---:---:---:---: Comfortable :---:---:---:---:---:---:---: Submissive :---:---:---:---:---:---:---: Indulgent :---:---:---:---:---:---:---: Unpleasant :---:---:---:---:---:---:---: Obsolete :---:---:---:---:---:---:---: Unorganized :---:---:---:---:---:---:---: Emotional :---:---:---:---:---:---:---: Mature .A Semantic Differential Scale for Measuring Self.

.Stapel Scale The Stapel scale is a unipolar rating scale with ten categories numbered from -5 to +5. This scale is usually presented vertically.   SEARS   +5 +5 +4 +4 +3 +3 +2 +2X +1 +1 HIGH QUALITY POOR SERVICE -1 -1 -2 -2 -3 -3 -4X -4 -5 -5 The data obtained by using a Stapel scale can be analyzed in the same way as semantic differential data. without a neutral point (zero).

and understand More time .comparative Scales Scale Continuous Rating Scale Itemized Rating Scales Likert Scale Degrees of agreement on a 1 (strongly disagree) to 5 (strongly agree) scale Seven . witho ut a neutral point (zero) Easy to construct. . administer.Basic Non .point scale with bipolar labels Measurement of attitudes Easy to construct. product. administer over telephone Confusing and . and company images Measurement of attitudes and images Versatile to whether the Stapel Scale Unipolar ten .point scale.5 to +5.consuming Basic Characteristics Place a mark on a continuous line Examples Reaction to TV commercials Advantages Easy to construct Disadvantages Scoring can be cumbersome unless computerized Semantic Differential Brand.

Itemized Scale Decisions 1) Number of categories optimal there categories Although there is no single. unbalanced In general. respondents accuracy forced 5) Verbal description non-forced the are expected to have no opinion. the of the data may be improved by a nonscale An argument can be made for In situations where . the scale should be balanced to obtain objective data (Next Slide). an odd number of categories should be used 4) Forced vs. of categories If a neutral or indifferent scale response is possible from at least some of the respondents. 3) Odd/even no. traditional guidelines suggest that should be between five and nine 2) Balanced vs. number.

Balanced and Unbalanced Scales Balanced Scale Jovan Musk for Men is Extremely good Very good Good Bad Very bad Extremely bad Unbalance Scale Jovan Musk for Men is Extremely good Very good Good Somewhat good Bad Very bad .

Rating Scale Configurations A variety of scale configurations may be employed to measure the gentleness of Cheer detergent. Very harsh . . Some examples include: Cheer detergent is: 1) Very harsh 2) Very harsh --1 --2 --3 --4 --5 --6 --7 Very gentle Very gentle 3) . . . Neither harsh nor gentle . . Very gentle 4) ____ Very harsh 5) Very harsh Neither harsh nor gentle Very gentle ____ Harsh ____ ____ Somewhat Neither harsh Harsh nor gentle ____ Somewhat gentle ____ Gentle ____ Very gentle .

Measurement Error – Difference between observed score and true score .

XO = XT + XS + XR where XO = the observed score or measurement XT = observed time)score.Measurement Accuracy The true score model provides a framework for understanding the accuracy of measurement. the true score of the characteristic XS = systematic error ( they affect the in the same way each .

such as poor printing. 5) Lack of clarity of the scale. such as the presence of other people. emotions. 4) Sampling of items included in the scale: addition. overcrowding items in the questionnaire.Potential Sources of Error on Measurement 1) Other relatively stable characteristics of the individual that influence the test score. deletion. and distractions. noise. such as intelligence. such as differences among . 6) Mechanical factors. or changes in the scale items. and poor design. 3) Situational factors. social desirability. 2) Short-term or transient personal factors. and fatigue. and education. 7) Administration of the scale. including the instructions or the items themselves. such as health.

the measure is perfectly reliable.Reliability Reliability can be defined as the extent to which measures are free from random error. Random error produces inconsistency leading to lower reliability . If XR = 0. XR.

Validity The validity of a scale may be defined as the extent to which differences in observed scale scores reflect true differences among objects on the characteristic being measured. rather than systematic or random error. . XR = 0. XS = 0). Perfect validity requires that there be no measurement error (XO = XT.

and XS = 0.Relationship Between Reliability and Validity If a measure is perfectly valid. unreliability implies invalidity. Thus. In this case XO = XT.e. systematic error may also be present. Furthermore. If a measure is unreliable. If a measure is perfectly reliable. because systematic error may still be present (XO = XT + X ). it may or may not be perfectly valid. XS≠0. but not .. it cannot be perfectly valid. XR = 0. since at a minimum XO = XT + XR. Reliability is a necessary. i. it is also perfectly reliable.

Session .6 Data Collection and Questionnaire .

Collection of Data Data can be obtained : Secondary Source Internal Records Primary source .

Collection of Data Primary Data : Questionnaire : interview) Observation Schedule. Interview form (telephone and personal : .

Questionnaire Definition A questionnaire is a formalized set of questions for obtaining information from respondents. .

.Questionnaire Objectives It must translate the information needed into a set of specific questions that the respondents can and will answer. motivate. A questionnaire should minimize response error. and to complete the interview. A questionnaire must uplift. and encourage the respondent to become involved in the interview. to cooperate.

Questionnaire Design Process Specify the Information Needed Specify the Type of Interviewing Method Determine the Content of Individual Questions Design the Question to Overcome the Respondent’s Inability and Unwillingness to Answer Decide the Question Structure Determine the Question Wording Arrange the Questions in Proper Order Identify the Form and Layout Reproduce the Questionnaire Eliminate Bugs by Pre-testing .

Individual Question Content 1. that question should be eliminated. .Is the Question Necessary? If there is no satisfactory use for the data resulting from a question.

Individual Question Content ─ 2. Consider the question: “Do you think Coca-Cola is a tasty and refreshing soft drink?” (Incorrect) Such a question is called a double-barreled question. two distinct questions should be asked:   “Do you think Coca-Cola is a tasty soft drink?” and “Do you think Coca-Cola is a refreshing soft drink?” (Correct) . because two or more questions are combined into one. several questions are needed to obtain the required information in an unambiguous manner. To obtain the required information. Are Several Questions Needed Instead of One? Sometimes.

filter questions that measure familiarity and past experience should be asked before questions about the topics themselves. Is the Respondent Informed? In situations where not all respondents are likely to be informed about the topic of interest. A “don't know” option appears to reduce uninformed responses without reducing the .Overcoming Inability To Answer – 1.

                  ___ 4 to 6 times per week 4.                  ___ Less than once a week 2.                  ___ 1 to 3 times per week 3. Can the Respondent Remember? How many gallons of soft drinks did you consume during the last four weeks? (Incorrect) How often do you consume soft drinks in a typical week? (Correct) 1.                  ___ 7 or more times per week .Overcoming Inability To Answer – 2.

such as pictures. Can the Respondent Respondents Articulate? unable to may be articulate certain types of responses.g. maps. and descriptions to help them articulate their responses. . e.. Respondents should be given aids. describe the atmosphere of a department store.Overcoming Inability To Answer – 3.

.Overcoming Unwillingness To Answer – Effort Required of the Respondents Most respondents are unwilling to devote a lot of effort to provide information.

sensitive information because this may cause embarrassment or threaten the respondent's prestige or self-image.Overcoming Unwillingness To Answer Context Respondents are unwilling to respond to questions which they consider to be inappropriate for the given context. .   Sensitive Information Respondents are unwilling to disclose.   Legitimate Purpose Explaining why the data are needed can make the request for the information seem legitimate and increase the respondents' willingness to answer. The researcher should manipulate the context so that the request for information seems appropriate. at least accurately.

. Use randomized techniques. Preface the question with a statement that the behavior of interest is common. Ask the question using the third-person technique : phrase the question as if it referred to other people. The entire list of questions can then be asked quickly.Overcoming Unwillingness To Answer – Increasing the Willingness of Respondents Place sensitive topics at the end of the questionnaire. Hide the question in a group of other questions which respondents are willing to answer. Provide response categories rather than asking for specific figures.

Unstructured questions are openended questions that respondents answer in their own words. What is your occupation? Who is your favorite actor? What do you think about people who shop at high-end department stores? Choosing Question Structure – Unstructured Questions .

A structured question may be multiple-choice.Choosing Question Structure – Structured Questions Structured questions specify the set of response alternatives and the response format. . dichotomous. or a scale.

In multiple-choice questions. the researcher provides a choice of answers and respondents are asked to select one or more of the alternatives given. Do you intend to buy a new car within the next six months? ____ Definitely will not buy ____ Probably will not buy ____ Undecided ____ Probably will buy ____ Definitely will buy ____ Other (please specify) Choosing Question Structure – Multiple-Choice Questions .

” “both. and so on. Often. such as “no opinion.” “don't know.Choosing Question Structure – Dichotomous Questions A dichotomous question has only two response alternatives: yes or no. agree or disagree. the two alternatives of interest are supplemented by a neutral alternative.” or “none.” Do you intend to buy a new car within the next six months? _____ Yes _____ No _____ Don't know .

Choosing Question Structure – Scales Do you intend to buy a new car within the next six months? Definitely Probably Definitely will not buy will not buy will buy 1 2 Undecided Probably will buy 3 4 5 .

when. where. Which brand of shampoo do you use? (Incorrect) Which brand or brands of shampoo have you personally used at home during the last month? In case of more than one brand. Who. and where are particularly important. (Correct) Choosing Question Wording – Define the Issue . what. what. why. please list all the brands that apply. when. and way (the six Ws).Define the issue in terms of who.

In a typical month. how often do you shop in department stores? _____ Less than once _____ 1 or 2 times _____ 3 or 4 times _____ More than 4 times (Correct) Choosing Question Wording – Use Unambiguous Words . how often do you shop in department stores? _____ Never _____ Occasionally _____ Sometimes _____ Often _____ Regularly (Incorrect) In a typical month.

as in the following:   Do you think that patriotic Americans should buy imported automobiles when that would put American labor out of work? _____ Yes _____ No _____ Don't know (Incorrect) Do you think that Americans should buy imported automobiles? _____ Yes _____ No _____ Don't know (Correct) .Choosing Question Wording – Avoid Leading or Biasing Questions respondent to what A leading question is one that clues the the answer should be.

Do you like to fly when traveling short distances? (Incorrect) 2.   1. Do you like to fly when traveling short or would you rather drive? (Correct) .Choosing Question Wording – Avoid Implicit Alternatives An alternative that is not explicitly expressed in the options is an implicit alternative. distances.

Choosing Question Wording – Avoidshould not be worded so that the Implicit Assumptions Questions answer is dependent upon implicit assumptions about what will happen as a consequence. Are you in favor of a balanced budget? (Incorrect) Are you in favor of a balanced budget if it would result in an increase in the personal income tax? (Correct) 2. .   1.

basic information should be obtained first. should be placed late in the sequence. identification information. finally.Determining the Order of Questions Opening Questions The opening questions should be interesting. embarrassing. followed by classification. or dull.   Type of Information As a general guideline.   Difficult Questions Difficult questions or questions which are sensitive. . complex. and non-threatening. simple. and.

how important is convenience of location?” (Correct) .Determining the Order of Questions Effect on Subsequent Questions General questions should precede the specific questions (funnel approach).   Q1: “What considerations are important to you in selecting a department store?” Q2: “In selecting a department store.

The questionnaires themselves should be numbered serially. The questions in each part should be numbered. particularly when branching questions are used.Form and Layout Divide a questionnaire into several parts. The questionnaires should preferably be precoded. .

......-6 .... Considering all the times you pick it up..The American Lawyer Example of a Precoded Questionnaire A Confidential Survey of Our Subscribers (Please ignore the numbers alongside the answers........) 1...............-4 2 hours to 2 hours 59 minutes............. about how much time....................... in total.....-1 30 to 59 minutes......................... They are only to help us in data processing...-5 3 hours or more....-2 1 hour to 1 hour 29 minutes.. do you spend reading or looking through a typical issue of THE AMERICAN LAWYER? Less than 30 minutes....-3 11/2 hours to 1 hour 59 minutes.....

Grids are useful when there are a number of related questions they use the same set of response categories. The tendency to crowd questions together to make the questionnaire look shorter should be avoided. Vertical response columns should be used for individual questions.Reproduction of the Questionnaire The questionnaire should be reproduced on good-quality paper and have a professional appearance. Directions or instructions for individual questions should be placed as close to the questions as possible. Questionnaires should take the form of a booklet rather than a number of sheets of paper clipped or stapled together. . Each question should be reproduced on a single page (or double-page spread).

A questionnaire should not be used in the field survey without adequate pretesting. wording. Pretests are best done by personal interviews. sequence. and instructions. All aspects of the questionnaire should be tested. telephone. question difficulty. including question content. form and layout. or .Pretesting Pretesting refers to the testing of the questionnaire on a small sample of respondents to identify and eliminate potential problems. The respondents for the pretest and for the actual survey should be drawn from the same population. even if the actual survey is to be conducted by mail.

telephone. or electronic means if those methods are to be used in the actual survey. . another pretest could be conducted by mail. A variety of interviewers should be used for pretests. Finally. The pretest sample size varies from 15 to 30 respondents for each wave. Protocol analysis and debriefing are two commonly used procedures in pretesting.Pretesting After the necessary changes have been made. the responses obtained from the pretest should be coded and analyzed.

Measurement of Central Tendency Session .7 .

according to some attribute – Male and Female Quantitative i.e.e. Area wise classification – cities .income . on the basis of time – year wise Qualitative i. In terms of magnitude – some characteristics.Classification of Data Geographic i. districts Chronological i.e .e.

Refrigerator sold each day in Oct.Formation of Frequency Distribution e. 2008 Classification according to class intervals Class Limits Class intervals Class frequency .g.

Tabulation Simple Tables or one way table Two way Tables .

. percentages. A frequency distribution for a variable produces a table of frequency counts.Frequency Distribution In a frequency distribution. and cumulative percentages for all the values associated with that variable. one variable is considered at a time.

mode. etc. Quartile Measure of variation Range. median. skewed. coefficient of variation Shape Symmetric. using boxand-whisker plots Coefficient of correlation . variance and standard deviation.Measures of central tendency Mean. interquartile range.

Summary Measures

Central Tendency

Quartile Mode


Mean Median

Range Variance

Coefficient of Variation

Geometric Mean

Standard Deviation

Data:100, 78, 65, 43, 94, 58 Mean: The sum of a collection of data divided by the number of data 43+58+65+78+94+100=438 438÷6=73 Mean is 73

Sample Mean
X1  X 2  L  X n X  n n Population Mean
i 1 i



Sample Size


i 1


Population Size



X1  X 2  L  X N  N

Mean Direct Method : X .

• The most common measure of central tendency • Acts as ‘Balance Point’ • Affected by extreme values (outliers)
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 12 14

Mean = 5

Mean = 6

Robust measure of central tendency Not affected by extreme values
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 12 14

Median = 5

Median = 5

In an ordered array, the median is the “middle” number
If n or N is odd, the median is the middle number If n or N is even, the median is the average of the two middle numbers

A measure of central tendency Value that occurs most often Not affected by extreme values Used for either numerical or categorical data Mode = 9 No Mode There may be no mode or several modes
1 2 34 5 6 7 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

is the value such that 25% of the observations are smaller.Quartiles Q1. is the median. . corresponding to (n+1)/4 ordered observation Q2. is the value such that 75% of the observations are smaller. the third quartile. the second quartile. the first quartile. corresponding to 2(n+1)/4 = (n+1)/2 ordered observation Q3. 50% of the observations are smaller.

Quartiles Split Ordered Data into 4 Quarters 25% 25% 25% 25%  Q3  i  n  1 Position of ith Quartile  Qi    Q2  4 Data in Ordered Array: 11 12 13 16 16 17 17 18 21  Q1  1 9  1  12  13  12.5 Position of Q1   2. Q3 = 17.5 Q1  2 = Median 4 16.5 = .

Measures of Variation Variation Variance Standard Deviation Coefficient of Variation Range Population Variance Sample Variance Population Standard Deviation Sample Standard Deviation Interquartile Range .

7 = 5 Range = 12 .7 = 7 12 8 9 10 11 7 12 8 9 10 11 .Range Measure of variation Difference between the largest and the smallest observations: Range  X Largest  X Smallest Ignore the way in which data are distributed5 Range = 12 .

5  5 Not affected by extreme values .5  12.Interquartile Range Measure of variation Also known as midspread Spread in the middle 50% Difference between the first and third quartiles Data in Ordered Array: 11 12 13 16 16 17 17 18 21 Interquartile Range  Q3  Q1  17.

Variance •Important measure of variation •Shows variation about the mean Sample variance: S  2  X i 1 n i X 2 n 1 i Population variance 2   X i 1 N  2 N .

Standard Deviation Most important measure of variation Shows variation about the mean Has the same units as the original data n 2  Sample standard deviation:  X i  X  i 1 S n 1 Population standard deviation:    2  Xi  i 1 N N .

338 11 12 21 Data B 13 14 15 16 17 18 19 20 Mean = 15.5 s = 3.57 11 21 12 13 14 15 16 17 18 19 20 .Comparing Standard Deviations Data A Mean = 15.5 s = 4.9258 11 12 20 21 13 14 15 16 17 18 19 Data C Mean = 15.5 s = .

Coefficient of Variation Measure of Relative Dispersion Always in % Shows Variation Relative to Mean Used to Compare 2 or More Groups Formula (Sample Coefficient of Variation) S CV = ⋅ 100% X .

Session .8 Skewness and Kurtosis .

Range Review of Previous Lecture The difference between the largest and smallest values Interquartile range The difference between the 25th and 75th percentiles Variance The sum of squares divided by the population size or the sample size Standard deviation The square root of the variance .

•Another Measure of Dispersion •Coefficient of Variation (CV) •Skewness •Kurtosis .

It is the ratio of the sample standard deviation to the sample mean s CV = ×100% x It is sometimes expressed as a percentage .Measures of Dispersion – Coefficient of Variation Coefficient of variation (CV) measures the spread of a set of data as a proportion of its mean.

measures of dispersion) Both measures tell us nothing about the shape of the distribution A further characterization of the .Measures of Skewness and Kurtosis A fundamental task in many statistical analyses is to characterize the location and variability of a data set (Measures of central tendency vs.

Skewness Skewness measures the degree of asymmetry exhibited by the data .

Skewness Positive skewness There are more observations below the mean than above it When the mean is greater than the median Negative skewness There are a small number of low observations and a large number of high ones When the median is greater than the mean .

Shape of a Distribution Describes how data is distributed Measures of shape Mean > median: right-skewness Mean < median: left-skewness Mean = median: symmetric Right-Skewed Left-Skewed Symmetric Mean < Median < Mode Mean = Median =Mode Mode < Median < Mean .

Kurtosis Kurtosis measures how peaked the histogram is n kurtosis = ∑ (x − x) i i 4 ns 4 −3 The kurtosis of a normal distribution is 0 Kurtosis characterizes the relative peakedness or flatness of a .

there are high frequencies in only a small part of the curve (i..Kurtosis Platykurtic– When the kurtosis < 0. the frequencies throughout the curve are closer to be equal (i. positive kurtosis indicates a relatively peaked distribution .e. negative kurtosis indicates a relatively flat distribution Leptokurtic– When the kurtosis > 0. the curve is more flat and wide) Thus. the curve is more peaked) Thus.e.

Kurtosis k>3 Frequency k=3 k<3 Value • • • Kurtosis is based on the size of a distribution's tails. Negative kurtosis (platykurtic) – distributions with short tails Positive kurtosis (leptokurtic) – distributions with relatively long tails .

-It helps in planning future operations -It helps in evaluating current accomplishments . observed or recorded at successive intervals of time – such data are referred as TIME SERIES : -It helps in understanding the past behavior.TIME SERIES ANALYSIS Statistical data which are collected.

TIME SERIES ANALYSIS Components of Time Series: -Secular trends – General movement persisting over long term -Seasonal variations .pattern year after year -Cyclical variations – Fluctuations moving up and down every few years -Irregular variations.Variations in .

Methods of Measurement -Moving Avg. Method -Method of least square .

Correlation Analysis .

these quantities are said to be correlated. The statistical tool for calculating such relationship is known as correlation and is denoted by = r. .If two quantities vary in such a way that movement in one are accompanied by movement in another. partial and multiple.linear .Simple. Types of correlation ship .Linear and Non . .Positive and Negative.

Scatter Plots and Correlation A scatter plot (or scatter diagram) is used to show the relationship between two variables Correlation analysis is used to measure strength of the association (linear relationship) between two variables Only concerned with strength of the relationship No causal effect is implied .

Scatter Plot Examples Linear relationships y y Curvilinear relationships x y y x x x .

Scatter Plot Examples Strong relationships y y Weak relationships x y y x x x .

Scatter Plot Examples No relationship y x y .

Correlation Coefficient The population correlation coefficient ρ (rho) measures the strength of the association between the variables The sample correlation coefficient r is an estimate of ρ and is used to measure the strength of the linear relationship in the sample observations .

the weaker the linear relationship .Features r Range between -1 and 1 The closer to -1. the stronger the negative linear relationship The closer to 1. the stronger the positive linear relationship The closer to 0.

Calculating the Correlation Coefficient r = ( ∑ x −x )( y −y ) [ ∑ x −x ) ][ ∑ y −y ) ( ( 2 2 ] or the algebraic equivalent: r= [n( ∑ x 2 ) − ( ∑ x )2 ][n( ∑ y 2 ) − ( ∑ y )2 ] n∑ xy − ∑ x ∑ y where: r = Sample correlation coefficient n = Sample size x = Value of the independent variable y = Value of the dependent variable .

For Example Tree Height y 35 49 27 33 60 21 45 51 Σ=321 Trunk Diameter x 8 9 7 6 13 7 11 12 Σ=73 xy 280 441 189 198 780 147 495 612 Σ=3142 y2 1225 2401 729 1089 3600 441 2025 2601 Σ=14111 x2 64 81 49 36 169 49 121 144 Σ=713 .

886 r = 0. x .Tree Height. y 70 60 50 40 30 20 10 0 0 2 4 6 8 10 12 14 r= [n( ∑ x 2 ) − ( ∑ x)2 ][n( ∑ y 2 ) − ( ∑ y)2 ] 8(3142) − (73)(321) [8(713) − (73)2 ][8(14111) − (321)2 ] n∑ xy − ∑ x ∑ y = = 0.886 → relatively strong positive linear association between x and y Trunk Diameter.

Calculations of Correlation when deviations are taken from Assumed Mean .

Rank Correlation coefficient .

Sign up to vote on this title
UsefulNot useful