By Karen Sternheimer

We’re all amateur would-be social scientists. We make observations of the world around us and draw conclusions. This is not social science: science involves creating hypotheses, or educated guesses, and then testing our hypothesis using a systematic empirical method. We also have to be open to being wrong. Always.

Many of us look around and see people on their mobile devices quite often. We might opine that this creates disconnection, isolation, and even rudeness in people who don’t pay attention to their surroundings in public.

Armchair experts just stop there: they’ve seen it with their own eyes, more than once, thus it is a social pattern as far as they are concerned. But this is not a systematic observation.

How would we make this assumption into a testable hypothesis?  Here are seven steps to consider:

1. First, a hypothesis is an educated guess, so we should get educated on research already conducted on this issue.
   
2. We also have to get more specific about both the independent and dependent variables. We are proposing that mobile device use leads to something else, so that would be the independent variable. An independent variable precedes the dependent variable, or the result or outcome.
   
3. Next, we need to be much more specific. Who is using the device? How are they using it? When are they using it? Where are they using it?
   
   For instance, a student who is texting during class? Scrolling Instagram before class when they could be talking to their classmates? We would have to figure out a way to measure this, perhaps by counting how many texts get sent during class, or measure the minutes spent on Instagram in social situations.
   
   Our outcome variable, or dependent variable, needs to be more specific as well. We should start by picking one of the outcomes listed above (disconnection, isolation, rudeness). Let’s choose isolation, which is similar to disconnection. Rudeness might be harder to measure because people might be reluctant to admit to it compared with feeling isolated.
   
   How to test isolation? The number of close friends someone reports? Their feeling of isolation on a scale of 1 to 10?
   
4. Let’s say we decide to revise our hypothesis to propose that students who scroll Instagram before class will have fewer friends overall.
   
5. Next steps would be deciding on a population to test: college students? If so, which ones? We probably can’t include all college students, even in one institution, so we’ll need to figure out how to create a subsample.
   
6. Will this be an observational study, based on interviews, or surveys? Each method has its strengths and weaknesses.
   
   In observing students’ behavior, we will understand more of the context in which scrolling takes place. We might not know if they are using Instagram or another app. Interviews will allow us to get students’ perspectives on why they are using Instagram before class, or whatever they happen to be scrolling. Both of these methods would likely limit our number of participants, since we can only observe and interview so many people within our study’s time frame.
   
   A survey is often a student’s favorite method, because they are familiar. But getting people to respond to surveys is difficult. We wouldn’t get the depth we would with the other two methods, but we might get a large sample that is more geographically diverse.
   
7. Setting aside the challenges and limitations, let’s start by conducting interviews, which is an excellent way to learn about participants’ experiences. We would have to write questions that encourage them to share whether they scroll before class, what apps they look at and why, and perhaps whether they ever make conversation with classmates. We can also ask them to describe how often they are in contact with close friends and learn more about those connections.
   
   We wouldn’t ask our hypothesis in the form of a question (do you think that students who scroll Instagram before class will have fewer friends overall?) because we really aren’t looking for opinions about our hypothesis, but whether data support it. This might also push them to say they don’t scroll Instagram before class, because there is a suggested negative connotation within our hypothesis. Remember—we started off very critical of mobile phone users!

Let’s pause here for a moment and think about all of the considerations we have taken, before even conducting a study. This process reminds us that drawing sweeping conclusions is not something social science encourages; instead, we look at small connections, bit by bit, building on other people’s research on the same subject.

If we had done any prior reading of previous research, we might have explored some of the many studies conducted by Pew Research Center, a nonpartisan think tank. One particular study of how teens view their social media use reported that apps like Instagram and Snapchat are how they stay connected to their friends and family. A large portion of teens noted that Snapchat in particular helped their friendships.

It’s easy to make unsupported conclusions about the world around us, especially other people’s social media use. How can you use these seven steps to think more critically about an assumption you or others might have about our social world?