Question

• Les wants to know whether a funny ad or a serious ad is more effective at getting consumers to like his company, XYZ Toys. He randomly assigns 100 consumers to watch the funny ad, the serious ad, or no ad at all (IV). After they watch the ad, he asks them how much they like XYZ Toys on a scale from 1 (Dislike a great deal) to 7 (Like a great deal) (DV).
  • What form of study design is this?
  • Will the researcher be able to make claims about causality with this study?

Answer 1

(a)

This study requires Likert scale.

We have to divide 100 consumers into three groups in completely randomized way using random number table, random number generator or otherwise. Then funny advertisement, serious advertisement and no advertisement are arranged separately for consumers in different groups i.e. one group is provided with funny advertisement, another with serious advertisement and third one with no advertisement.

Then customers are asked to provide their rating. We have to collect this data and store data corresponding to each group separately.

Based on collected data on these three groups we have to perform one-way ANOVA. Corresponding F-statistic is to be calculated. After calculating p-value we can compare with our desired level of significance.

(b)

Based on our proposed study design, we can find whether ratings are affected by advertisement or not.

• If no effect is observed, then the conclusion is certain.
• However, if effect is observed, we have to perform further analysis to get conclusion about which groups are affected and which are not by performing repeated t-test, using Bonferroni correction or otherwise.