Question

an online retailer wants to study whether more lenient return policies increase purchasing behaviors. they select a random sample of 2000 customers who have made purchases in the past year. they use a random number generator to select 1000 customer account numbers to assign to the treatment group. the remaining 1000 customers are assigned to the control group. what type of experiment design is this? choose 1 answer: a cluster b completely randomized c stratified d randomized block

Explanation:

1. Recall the definitions of each experimental design:

• Cluster design: Divides the population into clusters (groups) and randomly selects entire clusters to be in the sample. Here, we are not selecting clusters but individual customers, so A is incorrect.
• Completely randomized design: Randomly assigns experimental units (here, customers) to different treatment groups (treatment and control) without any blocking or stratifying. In this case, a random number generator is used to assign 1000 customers to treatment and the rest to control, which fits the completely randomized design.
• Stratified design: Divides the population into strata (sub - groups with similar characteristics) and then samples from each stratum. There is no mention of stratifying the customers (e.g., by age, spending level, etc.) here, so C is incorrect.
• Randomized block design: First divides the experimental units into blocks (groups with similar characteristics) and then randomly assigns treatments within each block. There is no blocking done here, so D is incorrect.

Answer:

B. Completely randomized