Home / Summaries / Marketing Models / techniques-relationship-data

ROI, experiments and ANOVA - Experiments

Q: What are the 3 categories of gathering and analyzing data?

Exploratory techniques - e.g. interviews, focus groups etc. Descriptive techniques - e.g. survey data, online purchase histories etc. Causal techniques - e.g lab experiments, test markets in the fields etc.

Q: What 3 kinds of evidence should you find when looking at a causal relationship?

Concomitant variation - If X causes Y , then X and Y should be correlated . Sequential ordening - If X causes Y , we would expect X to occur first and then Y would follow. Elimination of other possible explanations - If X, Q, or Z may cause Y , it would be desirable to show that Q and Z do not , leaving the X -> Y relationship most plausible .

Q: What is external validity and why is it important?

When the respondents are sampled randomly . Important because it allows for generalization of the results to the broader target population .

Q: What is internal validity and why is it important?

When respondents are randomly assigned to one of the two conditions . It is important because it allows us to attribute the results to our intervention , not to some other confounding factor .

Q: What are the dependent and independent variables in a study like this?

Independent variable / explanatory factor / predictor = the thing we manipulate . Dependent variable / criterion measure / response function = the result we measure .

5 important questions on ROI, experiments and ANOVA - Experiments

What are the 3 categories of gathering and analyzing data?

Exploratory techniques - e.g. interviews, focus groups etc.
Descriptive techniques - e.g. survey data, online purchase histories etc.
Causal techniques - e.g lab experiments, test markets in the fields etc.

What 3 kinds of evidence should you find when looking at a causal relationship?

Concomitant variation - If X causes Y, then X and Y should be correlated.
Sequential ordening - If X causes Y, we would expect X to occur first and then Y would follow.
Elimination of other possible explanations - If X, Q, or Z may cause Y, it would be desirable to show that Q and Z do not, leaving the X -> Y relationship most plausible.

What is external validity and why is it important?

When the respondents are sampled randomly.
Important because it allows for generalization of the results to the broader target population.

What is internal validity and why is it important?

When respondents are randomly assigned to one of the two conditions.
It is important because it allows us to attribute the results to our intervention, not to some other confounding factor.

What are the dependent and independent variables in a study like this?

Independent variable/explanatory factor/predictor = the thing we manipulate.
Dependent variable/criterion measure/response function = the result we measure.