You are on page 1of 23

Topic 2:

Research Design

Prof. Roberto Dopeso Fernández

21/04/2016 1
Questions
• What is the Research Design?
• What are the different types of research
design available?
• What similarities and differences exist
between the research design types?
• What are the main sources of error that
can be committed?
• How to determine the subjects to study?
Research Design Definition
• It is a framework or blueprint for
conducting a marketing research project
• It details the procedures necessary to
obtain the information needed to solve de
Marketing Research Problem
• The Research Proposal present the broad
abroach while the Research Design detail
the practical aspects to implement the
research
Research Design Components

1. Define the Information Needed


2. Decide whether the overall design is to be
exploratory, descriptive or causal
3. Design the sequence of techniques of understanding
and/or measurement
4. Construct and pretest an appropriate form of data
collection
5. Specify the qualitative or quantitative sampling
process and sample size
6. Develop a plan for qualitative and/or quantitative
data analysis
Research Design Challenges
• Connect the Decision-Makers information
needed with the target respondents
• Decision-Maker want information
– Accurate
– Current
– Sufficient
– Available
– Relevant
• These are an ideal because always exist
sources of error, and budget and time
restrictions
Research Design Challenges
• The target respondents
– Are generally humans
– The process of measuring and observing
humans may cause them to change
– The experimental design had limitations
Research Design Classification
• Two general classification
– Exploratory
• Flexible Design (Unstructured)
• Evolving approach
• Marketing phenomena difficult to measure
• Small Samples
– Conclusive
• Specific Measuring (Structured)
• Clearly defined marketing phenomena
• Big Samples
Research Design Classification
When to use Exploratory
Design
1. To have some background when nothing is known
2. To define problems further or formulate hypothesis
for quantification
3. To explore concepts (product design or market
communication)
4. To reduce the number of alternatives for further
research
5. To know attitudes, values, motivations or opinions of
the target audience
6. To interpret the failures of quantitative research to
explain a phenomena
7. To treat questions that are difficult to get answers
(embarrassment, sensitive or unconscious)
When to use Conclusive Design
1. Describe characteristics of well defined groups (salesmen,
consumers, potential customers, organizations)
2. Specify proportions inside of the population
3. To make inferences about the population behavior from a
sample of its individuals
4. To integrate data from different sources in Marketing
Information Systems or Decision Support Systems.
5. To understand and quantify relations between variables
6. To study a marketing phenomena over time
7. To obtain results that can be replicable
8. To make predictions or forecasts
Types of Conclusive Research
• Descriptive
– Its objective is describe the characteristics of
the market or the marketing phenomena
under study
• Causal
– Its objective is understand the relation that
exists between variables
• Dependent Variables (it is caused)
• Independent Variables (it is the cause)
Types of Conclusive Research
• Descriptive (Examples)
– Size of the Market or How it is divided (Market
Share)
– Budget of the Potential Customers or General
Customer Profiles
– Product Usage Rates
– Distribution Channels available
– Media Consumption habits
Types of Conclusive Research
• Descriptive
– Cross-Sectional
• Collecting information from a sample a single time
• Two types
– Single: One sample
– Multiple: Several Samples with the same characteristic of
differences (Geographic, Time or Other characteristics)
» Cohort Analysis (Time Intervals, Similar effects)
– Longitudinal
• Fixed sample (or samples) measured repeatedly
– Longitudinal can study trends and obtain more
information accurately but can generate non
representative samples and response bias
Example
Example
Example
Example
Example
• Cross-Section

• Longitudinal
Types of Conclusive Research
• Causal
– Asses cause-effect relationships
– Examples
• Understand what variables are the cause (independent)
and the effect (dependent) of marketing phenomena
• What it is the relation between the causal variables
• Test hypothesis
– The best approach is the experimental design
(controlled environment). Unfortunately in Social
Science doing experiments it is difficult and/or
costly
Relation between the types of
Research
• The distinction between exploratory, descriptive and
causal it is not absolute
• They can be used at the same time to serve different
purposes.
• Exploratory can serve as a beginning point when we
do not know possible attractive segments or to much
about relevant product features. Then we can use
Descriptive only for the attractive segments or make
emphasis on important points
• Also we can conduct a causal research on the effect of
regular consumption on long-term health, but if our
hypothesis are not conclusive maybe we can conduct
exploratory research to reevaluate our model
Sources of Error
• Total Error: The difference between the true mean value
of the interest variable in the population with respect to
the one obtained in the sample obtained in the
marketing research project
– Random Sampling Error: Caused because the sample
obtained it is not representative of the population.
– The difference between the true mean value of the interest
variable in the population with respect to the one obtained in
the original sample
– Non-Sampling Error: Do not depends on the sampling
procedure, can be random or non-random. The main reasons
are Non-Responses, or errors in the response from the
researcher, the interviewer or the contestant

You might also like