Last fall, some social scientists published a study using data from one of my favorite national surveys–the American Time Use Survey. (I’m a nosy psychologist who is fascinated by how people spend their days.). They concluded that doing almost any daily activity is more enjoyable when we do it with friends. Here’s how the research was covered by the press.

I’ll paste a quote from the journalist’s piece:

The study found that across more than 80 different types of activities, people reported being happier doing almost anything with other people than doing it alone.

The activities ranged from the mundane – getting gas, doing homework, commuting, grocery shopping – to ones that may not seem inherently social, such as reading, doing car repairs or managing finances.

a) Think for a moment: If you used a survey to study this question, how would you ask? Here’s a hint: I think you’d need to ask three separate questions about each activity.

Now here are some details about the methodology of the actual study:

The researchers took advantage of a rich dataset from the American Time Use Survey, conducted by the Census Bureau, which asks participants how they spent the previous day. In all, they analyzed how more than 40,000 participants felt – and whether they interacted with others – during more than 100,000 instances when they were doing various activities.

Every activity was significantly more enjoyable with others than alone in at least one of the years.

The activities we more often do with others seem to be associated with the most benefits, such as eating or drinking, walking, running and playing games.

But the benefits also applied to the activities that participants most frequently did alone, such as reading, arts and crafts, and commuting.

b) Pick any two activities from the ones listed above, and sketch a bar graph (either one bar graph for each activity, or two different colors of bars for the different activities) based on what the result would look like (here’s a hint: Put happiness on the y axis).

c) Now think about how you’d sketch an interval plot of this result. Instead of plotting two bars like you did above (one for alone and one with friends), you’ll make a box representing a difference score of “with friends minus without friends.” There’d also be a 95% CI around each box. What might this interval plot look for the two activities you plotted in question b? What would zero mean for this difference score? What would a positive value mean? What would a negative value mean?

d) Pop over to the published empirical article (It’s open-access) and scroll to the forest plot in Figure 1. What does each blue circle represent there? What do the wings represent? What does the solid yellow line represent? Finally, what is the overall conclusion of Figure 1?

e) Is the American Time Use Survey based on a random sample of Americans? Do a quick google search and see what you can find out. Which of the four big validities are you looking into here?