Chapter 5 (CN): Research Design
September 4, 2014
Research Design (master plan that specifies the methods and procedures for collecting and analyzing the
information needed for addressing the marketing research problem).
Types of Research Design
• Exploratory: to confirm research hypotheses, to gain background information, to define terms, to
clarify problems and hypotheses, to establish research priorities.
o Most commonly unstructured, informal research.
▪ No formal set of objectives, sample plan, or questionnaire.
▪ Conducted when the researcher does not know much about the problems.
▪ Conducted at the outset of research problems.
o Secondary Data Analysis, Experience Surveys, Case Analysis, Focus Groups, Projective
• Descriptive: to describe and measure marketing phenomena at a point in time.
o To describe answers to questions of who, what, where, when, and how.
o Used when we wish to project a study’s findings to a larger population through the use of
o Two basic classifications:
▪ Cross-Sectional Studies: multiple samples are asked the same set of questions for
one point in time.
• Snapshots of the population at a point in time.
▪ Longitudinal Studies: one sample is asked the same set of questions over time.
• Longitudinal Study:
o Wendy’s has gone from 100 (20%) to 75 (15%) customers.
o McDonald’s remained at 200 customers (40%).
o Taco Bell increased from 200 (40%) to 225 (45%) customers.
o *Wendy’s is losing market share to Taco Bell. McDonald’s has not lost or gained share from
Wendy’s and Taco Bell.
o Wendy’s lost 50 customers to McDonald’s BUT gained 25 customers from McDonald’s at the
o McDonald’s lost 25 customers to Wendy’s and 25 customers to Taco Bell BUT gained customers
o Taco Bell did not lose any customers BUT gained 25 customers from McDonald’s.
o *For Wendy’s: McDonald’s is the competition. McDonald’s lost and gained share from others.
Taco Bell has the most stable market share. • Causal: to determine cause-effect relationship.
o Conducted through the use of experiments (manipulating an independent variable to see
how it affects an independent variable to see how it affects a dependent variable, while
controlling the effects of extraneous variables.
Group 1 Group 2
Did not see a Pepsi advertisement. Saw a Pepsi advertisement.
“Do you like Pepsi?” “Do you like Pepsi?”
*Lower rates of approval. *Higher rates of approval for Pepsi. The ad worked
• IV: viewing of the advertisement.
• DV: if they liked or didn’t like Pepsi.
• Independent Variable: cause.
o Researcher has control over this variable; can manipulate.
o Called a “treatment.”
• Dependent Variable: effect.
o Influenced by the IV.
• Extraneous Variables: aka external or controlled variables; have some influence or effect on a
DV, yet they are not the IV.
o Must be controlled through proper experimental design.
▪ Randomization: random assignment of test units to experimental and control
groups so as to minimize the influence of extraneous variables.
▪ Matching: compare test units on a set of key extraneous variables before
assigning them to the treatment conditions.
o Control Group: group without treatment.
Group 1 Group 2 Group 3
Takes the MK201 class and the Takes the MK370 class and the Doesn’t take a class. Takes the
MK370 exam. MK370 exam. MK 370 exam.
• IV: the level of class.
• DV: the performance on the exam.
• Control: group 3.
• Criteria for Causality:
o Temporal Precedence: evidence that the cause precedes the effect.
▪ Treatment comes first. Then you measure the effect. o Covariation of the Cause and Effect: must be change in cause and effect.
▪ Treatment should be relevant to IV, so that it’s reasonable the cause infleunces
o No plausible extraneous variables.
• EX: Design an experiment. Store X wants to find out if its new merchandise display system
would increase sales.
Group 1 Group 2
Store with new display. Store with old display.
o IV: presence of the display.
o DV: increase or decrease in sales.
o Control: group 2.
• The choice of the most appropriate design depends on the objectives of the research and how
much is known about the problem and research objectives. Our research is more predictive in
Symbols of Experimental Design
• O: measurement of a DV.
o DV: O 1 O 2 etc.
• X: manipulation/change of an IV.
o X will never have a subset.
• Number of lines = number of groups.
o XO 1
O2= two groups.
• XO 1 treatment
• O 2 control
• EX: XO 1
o 1 experimental group: treatment.
o You give the treatme