Chapter 7 – Causal Research and Experimental Design
Three Key Marketing Research Approaches:
Exploratory Research: aims to gain basic information about an issue, topic, etc. and uses qualitative techniques such as focus groups, interviews, and secondary data.
Descriptive Research: answers who, what, where, when and how related questions. Its techniques are qualitative such as with certain observations studies, and quantitative such as statistically representative surveys.
Causal Research: reveals the why behind the descriptive research results. Methods include surveys and various experimental designs, both simple and complex
Documents the cause-effect relationship among the independent, dependent, and extraneous variables associated with your research topic.
Proving Causation
Requires rigorous testing
3 factors must be present:
Concomitant Variation: a statistically predictable pattern between responses/variables (when it rains, umbrella sales rise)
Time Order of Occurrence: A must precede B if A is the cause of B (customers are happy after they receive good service).
Systematic Elimination: all other possible causes have been eliminated
Experimental Variable Types:
Independent Variables: variables that the researcher can control directly and thus are independent in that they can be implemented regardless of the surrounding environment.
Dependent Variables: are affected by manipulating or changing the associated independent variables; they depend on a change to the independent variables for their effect.
Extraneous Variables: variables that the researcher has no true control over but whose impact the researcher might be able to mitigate (examples include legal, regulatory, and political issues as well as competition)
Types of Extraneous Variables
Testing Effects: occurs when the effect of taking the tests affects the next time the test is taken (example – taste test – the experience of having taken the test before might have an impact on how you answer the same questions the second time you take the test even though the product in this case has not changed)
History: events that occur which might have an impact on the test subjects
Regression to the mean: occurs when test subjects gravitate towards he middle of a given scale
Instrument: changes in the data collection instrument that affect the test subjects’ responses; the data collection instrument or method is causing a change in respondents’ answers even though there has been no real change in the actual respondents’ opinions.
Selection Bias: occurs when the test subjects are inappropriate for the research topic
Mortality: these test subjects are lost during the experimentation
Maturation: occurs when there are changes to the test subjects themselves
Controlling Extraneous Variables
Matching: involves pairing like test subjects to alert the researcher of unexpected changes in one of the test subjects
Statistical Control: uses statistical adjustments to correct for imbalances in the test subjects
Design Control: involves designing the research in a manner that reduces the impact of the extraneous variable(s) in the first place
Randomization: randomly assigning test subjects to the study
Choosing the best method for handling extraneous variables depends on the level of knowledge researchers have about the impact and source of the extraneous variable(s) coupled with their marketing research expertise in selecting the most appropriate mitigation strategy
Experimental Design Methods
Descriptive research describes; causal research takes the research analysis to a more analytical level
Determining why sales are declining is more analytic than simply detailing that they have fallen.
Experimental Design: a form of causal research that tests the manipulation of independent variables to measure the corresponding change in the dependent variables.
Helps researchers to understand whether the variables of