List two potential threats to the internal validity of a study when you use a subject pool and how researchers have tried to combat these threats.
Imagine that you are doing a study on the effect of wearing glasses on perceptions of professionalism in the workplace. Come up with a title that reflects the recommendations for labeling your study (to avoid demand characteristics). When would you schedule your study (day of week, time of day) for maximum representativeness?
What kinds of connections do you have that would enable you to conduct research with off-campus populations? What steps would you take to safeguard your physical security and the security of your participants? Beyond the steps that are outlined in this chapter, what other ideas do you have about securing your physical safety?
What are the benefits of conducting research at service learning sites? What are the potential drawbacks? What does it mean to be a “double agent” in the context of research? If you want to collect data from a group where you are a member, how can you avoid the possibility of coercion in requesting their participation?
Which of Dustin’s recommendations seem to focus on permissions and which on approaches to potential participants? Of Dustin’s recommendations, list the five you think are most important and justify your choices. Add a recommendation you think he overlooked.
What are some of the major challenges that Kosinski et al. (2015) identified in doing research on Facebook? Why did they discourage using financial incentives? What did they use instead? What is an advert? A snowball sample? What does the research of Gosling et al. (2004) show about the validity of Web-based surveys? What is the “emotional contagion” study, and what ethical issues did it raise?
Using the concept of representativeness, explain why an undergraduate subject pool is a nonprobability sample.
In general, explain why using probability sampling is preferable to nonprobability sampling. What makes something a random sample? How do stratified random and proportionate sampling differ? In what way is systematic sampling like random sampling? For cluster sampling, what does it mean to say that a cluster is the sampling unit?
Why is nonresponse not necessarily nonresponse bias? For basic demographic characteristics of a college sample, where could you find information about the population?
In terms of responding to surveys, what do we know about the relative effectiveness of (a) incentives versus no incentives? (b) Monetary versus gift incentives? (c) Prepaid versus promised incentives?
What is Amazon Mechanical Turk? What do we know about the validity of data collected on AMT and its comparability to data collected through more traditional approaches? Most workers come from which two countries? What is the primary motivation to participate in MTurk studies? Name a step you can take to assure that you will get workers to complete your task (“HIT”) with care.
What are online panels? What kinds of firms use them? Why are such panels expensive? (Think about the cost of finding people with particular characteristics.)
