1. Two of the following three examples have internal validity problems. Identify the one example that does not have an internal validity problem. For the remaining 2 examples, identify (i) whether the problem is a design confound, a selection effect, or an order effect. Finally, (ii) explain how each of the problematic studies might be changed so that it does not have the internal validity problem. (10 points) a. An applied psychologist wants to test the effectiveness of an intervention to increase awareness of the environmental impact of disposable water bottles. The initiative involves asking students to use a smartphone to track the number of disposable water bottles they use, categorizing whether they reuse them, throw them away, or recycle them. He asks for volunteers in a large geology course. Students must own a smartphone in order to participate in the intervention group. Fifty volunteers who owned a smartphone were assigned to the tracking condition (they downloaded a free smartphone app for this purpose). Fifty more students who were interested, but who did not own a smartphone, made up the comparison group. This group was simply exposed to a short video on the impact of plastic water bottles. At the end of a two-week period during which the students in the smartphone group tracked their water bottle use, the researcher found that students in the tracking program were more likely to have purchased a reusable water bottle in the past week compared with the students in the comparison group. He concluded that his smartphone tracking program raised awareness, causing students to purchase reusable bottles.
i. Design confound, selection effect, or order effect, or no problem?
ii. How would you change the design so that it does not have an internal validity problem (NA if no problem)?
b. A cognitive psychologist believes that people learn better when they spread out their studying over several days, so she creates a study with three groups of participants. Each group studies the same list of 120 Chinese vocabulary words (none of the participants had studied Chinese before). One group studies the words for 20 minutes on the first day. The second group studies the words for 20 minutes on the first and second days. The last group studies the words for 20 minutes on the first, second, and third days. On the fourth day, all of the participants are tested on how well they have learned the Chinese vocabulary words. The people in the last group scored the best, so the researcher concludes that distributed studying does improve people’s ability to learn.
i. Design confound, selection effect, or order effect, or no problem?
ii. How would you change the design so that it does not have an internal validity problem (NA if no problem)?