Everyday Research Methods

Report: Internet hyperconnectivity might change young adults

The Pew Research Center recently released a report on the internet's effect on millenials. The report notes that

Teens and young adults brought up from childhood with a continuous connection to each other and to information will be nimble, quick-acting multitaskers who count on the Internet as their external brain and who approach problems in a different way from their elders, according to a new survey of technology experts.

a) Note that this report isn't about what actually will happen to young adults becuase of their hyperconnectivity; it's about what certain people think will happen. According to the quote above, who was polled for this report?

Pew Research Center is helpfully up-front about its sampling techniques. Read this excerpt about their sampling for this particular survey, taken from this page:

The survey results are based on a non-random, opt-in, online sample of 1,021 internet experts and other internet users, recruited via email invitation.

And in case you weren't sure what the implications of that sampling method are, they also add:

Since the data are based on a non-random sample, a margin of error cannot be computed, and the results are not projectable to any population other than the experts in this sample.

b) Among the sampling strategies described in Chapter 6 and summarized in Figure 6.12, what kind of sampling did Pew use?

c) What would Pew have to do differently in order to get a truly representative sample of internet experts?

d) How might you study the true effect of internet hyperconnectivity on young adults? How could you test the hypothesis that they will "solve problems differently," for example?

Working mothers and mental health

The Wall Street Journal recent­ly published an article asking “Is Working Good for Mom’s Health?"

The article opens with the claim “working mothers tend to be happier and healthier than their stay-at-home counterparts.”

a.) What type of claim is this?

Here is a quote from the article describing the study’s sample and methods:

The researchers…interviewed a national sample of 1,364 mothers before and after the birth of their children, with follow-up interviews lasting more than 10 years.

The study found that part-time and full time working mothers were better at managing stressors and had greater mental health and less depression than mothers who did not work.

b.) What type of design does this study seem to use? Why might the researchers be using this design?

c.) Why don’t they just do an experiment?

d.) Do you think the researchers were in theory testing mode or generalization mode?

 

Suggested answers

a.) This is an association claim. The verb here, "to be," (as in, "working mothers tend to be happier and healthier...") is a simple linking verb, not a causal verb. Because there are two variables, maternal employment and mother’s well being, that means it's probably not a frequency claim. And the claim isn't causal because the claim is not that working increases a mothers health and happiness, just that the two go together.

b.) This is a longitudinal correlational design. The clues are that they state that the interviews occurred for more than 10 years and that the researchers studied the same sample and the same variables. By using this type of design, researchers can often establish temporal precedence in a correlational study.

c.) Mothers can’t be assigned to work or not work--that just isn’t feasible or ethical. So an experiment isn't really possible here.

d.) Because this study used a “national sample," we might expect that the study is in generalization mode. We could ask if the national sample was a probability sample, something that would be important for good external validity.

Guest post by Kate Banford, University of Delaware