Question

Pretend that a retail store wants to do testing regarding atmospheric elements (e.g., displays, lighting, etc.) to test what will keep people in the store the longest. What test market approach is preferred: laboratory or field experiment? 1. You want a rough, directional idea of the best combination of atmospheric elements (as opposed to trying to statistically estimate every possible effect with precision) 2. You want to simulate the conditions that a shopper would typically face in a store 3. You want to determine the effect of each element on how long customers spend in the store For the following four scenarios where you want to conduct a study on a group of respondents, would you recommend a cross-sectional or longitudinal design? 4. You want to do a study that tracks changes in consumer behavior over a two-year period 5. You want to ensure that your results are based off of a representative sample 6. You want to estimate the percentage of households that view TV shows through non-traditional channels (i.e., tablets, phones, website, Hulu, etc.) 7. You want to hand-select (as opposed to randomly selecting) participants for the study You want to do a study for a local gym where email offers will be sent out; you will determine how successful the email campaign is based on the response rate. Please read the following statement and determine which conditions for an experiment are being referenced: control, randomization, or replication. 8. You have to assign participants into different test groups that are roughly equal 9. You need some point of comparison/reference point that you can compare your test group(s) to 10. The more participants involved in the study, the more reliable the experiment results are

Answer 1

Q1 : 1. You want a rough, directional idea of the best combination of atmospheric elements (as opposed to trying to statistically estimate every possible effect with precision) - Field experiment as the requirement is to get a rough, directional idea that indicates further course of analysis and not precise measurements.

2. You want to simulate the conditions that a shopper would typically face in a store - Field experiment as exact simulation of conditions that a shopper would face in a store is required.

3. You want to determine the effect of each element on how long customers spend in the store - Laboratory as one can simulate and tweak the required number of variables while holding the others constant so that effect of each variable or variables can be accurately studied.

4. You want to do a study that tracks changes in consumer behavior over a two-year period - Longitudinal the changes in consumer behavior can only be studied if the same consumers are selected over the length of the survey.

5. You want to ensure that your results are based off of a representative sample - Cross-sectional the study needs to be carried out on a representative sample and hence you need not select the same participants.

6. You want to estimate the percentage of households that view TV shows through non-traditional channels (i.e., tablets, phones, website, Hulu, etc.) - Cross-sectional as the participants must correctly portray a sample of the population

7. You want to hand-select (as opposed to randomly selecting) participants for the study - Longitudinal you do not want random selection and want people based on a certain characteristics

8. You have to assign participants into different test groups that are roughly equal - Randomization as with this we will assign random participants to random groups and thus the groups will be roughly equal.

9. You need some point of comparison/reference point that you can compare your test group(s) to - Control as this will provide a reference point against whom the results of the test can be compared

10. The more participants involved in the study, the more reliable the experiment results are - Replication as more participants means that we will have higher confidence in the results and thus we can replicate the results with a higher degree of confidence