Research Design and Implementation

1

• Research Design and Implementation

2
Research design

• Detailed blueprint used to guide the research
  study to its objectives
• Defines research approach to be used for
  obtaining information

3
RESEARCH DESIGN PROCESS

Research Approach Exploratory Descriptive Cau
sal
Research Design
Choice of data collection method Secondary,
standardized, qualitative, Surveys,
experiments Role of research supplier Project
design Raw data collection
Research Tactics Develop measures,
Questionnaire, Design experiments, design
sampling plan, anticipate analysis
4
Categories of Research Approaches
Exploratory
Causal
Descriptive
Each element plays a distinct and complementary
role!
5
Exploratory Research

• Used when seeking insights
• Into the general nature of the problem
• The decision alternatives
• The variables to be included
• The questions why and how are important
• The probe

6
Exploratory research examples
7
Limitations of Exploratory Research

• No hypothesis or vague or ill-defined hypothesis
• Not possible to confirm / disconfirm
• No firm preconceptions of what will be found
• Biased relies too much on the interpretation of
  the researcher
• Inaccurate recording and reporting
• Technology overcomes this to a large extent
• Limited generalizability (external invalidity)

8
Descriptive Research

• Most marketing research falls in this category
• Greatest application describe a marketing
  situation
• Hypothesis exist but may be tentative and
  speculative
• Can only show if two variables are associated
• May precede causal research

9
Limitations of Descriptive Research

• Does not focus on the why understanding
  phenomena
• Cannot be used for prediction and control

10
Descriptive Research examples
11
Some causal research hypotheses
12
Causal Research

• Shows that one variable causes or determines the
  value of another variable
• Hence controlling one variable can help control
  the other
• Causality can be inferred only if
• Hypothesis stated very specifically in terms of
  variables and quantity or direction

13
Exploratory / Descriptive / Causal
Exploratory Descriptive Causal
Seek Insights Describe a marketing situation Establish predictive relationships
Broad research questions Specific research questions No research questions
No hypotheses Hypotheses with 1 / 2 variables and direction specified Only hypotheses with 2 variables and direction specified
Hypotheses with 2 variables can only confirm associations Hypotheses with 2 variables can confirm predictive power
Results feed into descriptive / causal Results may feed into causal research
14
Experimental Designs

• Designs to infer causality
• Aid understanding (why) of phenomena
• Allows prediction and control
• The observation and exposure or
  manipulation
• If really does cause then the
  experiment has internal validity
• If there is a question then we may have Plausible
  Rival Explanations (PRE)

15
Example Exp. Designs

• Carnival Cruise Lines known for its party image
  wants to attract the family crowd and creates a
  new campaign showing good, clean fun for
  families. The campaign runs in Rock Hill for six
  months.
• Assume we want to measure the impact of the
  campaign on a subjects likelihood of sailing on
  Carnival Cruise lines
• Basic experimental design

16
Some problems with the X O design
17
An improvement over the X O design

• Which effects have been controlled?
• Continued Problems

18
An improvement over the OXO design

• Which additional effects have been controlled
  here?
• Continued Problems

19
More improvement

• Further improvement
• The design now controls history, maturation and
  selection effects

20
Another case

• Sometime ago IBM sold their PC division to
  Lenovo. At that time Lenovo wanted to know which
  brand name (IBM ThinkPad or Lenovo) should they
  use on their line of laptops. Assuming they
  wanted to ascertain subjects willingness to buy
  based on the logo design only, identify a good
  experimental design that controls for PREs.

21
Some PREs for the Lenovo case
22
Lenovo Case

• What experimental design best controls for the
  PREs?
• Design now controls for History, Maturation and
  selection effects

23
Exploratory/Descriptive/Causal? Choice of data
collection method?
24
Exploratory/Descriptive/Causal? Choice of data
collection method?