• Sophie H. Janicke, Ph.D.

Nut’s and Bolts in Understanding Social Media Research

Understanding Social Media Research

"Does social media cause loneliness?'" Does social media lead to depression?" Facebook lowers our self-esteem,”…. these headlines are commonly found in the news media. We are constantly bombarded by the news media about the detrimental effects of social media and apparent new studies that are cited to confirm this. While, indeed some of those studies may be sufficient to be able to make such broad generalizations as they are made within the news media, many studies cited and picked up by journalists do not allow for such general conclusions. In fact, the Washington Post just summarized a study conducted by Common Sense Media that debunks a lot of the detrimental claims of social media for teens and provides tips for parents.

How do those different reports come about? I want to give you a short overview of some basic research methods that we have to consider when reading an academic publication on the effects of social media (or other phenomena in general). In today’s media world (unfortunately) we cannot put all of our bets on the journalist anymore, which is why we need to empower ourselves to know what to watch out for when we read about the effects of such a complex technology as social media.

Five Aspects to Consider when Interpreting Social Media Research

1. How social media exposure is measured

Until now (With Apples screentime app where we can actually track our screen time) it was hard to measure the objective amount of time spent on social media because we only had access to self-report. How good is your estimate of how much time you spend on your phone really? Do you think you over estimate? Underestimate? Or hit the time you actually spend on social media or your phone exactly? Every time I do this experiment with my students and have them download the screentime tracking app Moment or Breakfree and then reflect on if they were surprised by the number, most students are astonished by the amount they underestimated or (and yes, that happens too) overestimated their usage. Thus, any generalizations of a study that assesses screen time with a best guess, can lead to biased results.

Another way that oftentimes social media exposure is measured is by asking participants how much time they spend on all their different social media. The different platforms are then listed: Snapchat, Instagram, Facebook, Youtube. Twitter, Pinterest, Tumblr, Yikyak, you name it. As you can imagine, people greatly overestimate their time when indicating time spent on each platform. So, a bias in many studies out there comes from unobjective measures of time spent with social media.

2. The complexities of social media use are rarely considered

So far, only few studies consider in what CONTEXTS social media is used and HOW it is used. Social Media can be used in the context of work, for private reasons, for networking, on weekdays, weekends, during the day, during the night, and consequently can have different positive or negative effects. For example, researchers from the University of Oxford found in a study with a representative sample of 120.000 UK adolescents, that 2 hours of smart phone use during the week lead to increases in well-being, and after 2 hours to a steep decrease. On the weekend, the researchers found a slight negative trend for well-being for 4 hours or less and a steep negative trend after spending more than 4 hours on the smart phone.

Additionally, social media can be used with different intentions (consciously or unconsciously). For example, it can be used to connect with relatives far away, or as a substitute for f2f communication. It can be used to engage in communities and offline activism, or only to lurk around, which again determines if the effects from social media are positive or negative.

A review article published in the Journal of Social Issues and Policy Review showed that when using social media more actively, rather than passively, people feel greater positive emotions and life satisfaction. Another, longitudinal study of a representative survey of Americans, however, found that FB use in specific, independently if used passively or actively, decreases well-being.

A recent experimental study published in the Journal of Cyberpsychology, Behavior and Social Networking found that when people socially compare themselves in an assimilative way to people on Instagram, meaning they look for similarities between the Instagram personality and oneself, they feel inspiration, which in turn increased their positive emotions rather than negative emotions. The authors outline that envy as a result of social comparison does not always have to be bad. We can also find inspiration in other people’s accomplishments and become motivated to try to achieve similar things. So social comparison does not necessarily have to be bad. It can also be a good thing!

As you can see, the context and how we use social media matters greatly for the effects. While naturally not all studies can account for the context and how people use social media, it is important to keep in mind that a lot of variables interact and thus produce unique effects of social media on our well-being.

3. Check the Validity of Generalizations

Additionally, most of the research that investigated the effects of social media on well-being looked at Facebook use. This is not surprising, given that Facebook is the oldest and most successful social network in the world. Well, one of the problems is that Snapchat and Instagram are used as often as Facebook by the millennial generation and such mediums are very different from Facebook.

In fact, two researchers at the University or Oregon published a study in which they showed a relationship between Instagram and Snapchat use and decreased levels of loneliness and increased levels of positive emotions and life satisfactions, but no such a relationship was found for the use of YikYak and Twitter. The authors argue that image-based media, based on their visual nature to create more intimacy, may be more beneficial for well-being then text-based platforms. While these studies still need further support, it is important to acknowledge that we cannot generalize the results of a study that uses a particular platform as their main object of investigation to ALL social media platforms. Nor can we generalize effects to the whole population when the study is using only a specific convenience sample (i.e., university students) of the population.

So, when the news media makes generalizations such as: “Social media increases loneliness” we have to ask: for whom and what platform was studied.

4. Correlation does not mean causation

I know we all have heard this one multiple times before, but I think it is worth repeating. Most of the studies conducted on the effects of social media use on well-being are correlational in nature, which means that lower well-being, loneliness and mental health can also predict the amount of time people spend on social media, and not that social media is causing depression, loneliness and unhappiness, which is, however, what seems to be propagated through the media a lot. While some research indeed seems to provide some evidence for a causal effect using longitudinal and experimental designs with representative samples, I am not sure if we are at a point yet to make conclusive statements about the cause and effect here. While the methods with which phenomena are studied matter for the establishment of causal effects, alternative explanations need to be ruled out as well. I will outline a systematic review of the effects we have so far regarding causality of social media effects in another post.

Some researchers argue there is enough evidence for the claim that social media causes mental health issues, and not the other way around: Christine Carter, a social scientist herself outlines in her Greater Good Science Article that indeed there is enough evidence for a causal relationship. I don’t think there is anything wrong to err on the conservative side and accept her claim, because all it does (or should do, once causal effects have been proven) is making us even more cautious about social media. And knowing the addiction techniques social media companies are using to get us hooked, it makes sense that rather negative effects are the result.

5. Small effects do not mean no effects

Counter arguments from the no-causal-effects side pertain to the effect sizes that the longitudinal and experimental studies on social media effects report. That is, the effects are rather small. For example, compared to the effect of sleep on our mental and physical health (d=.58), or regular breakfast on our well-being (d=.54), the effect of smart phone use on our well-being (negative or positive) is not more than 1/3rd of that (d=.18) according to the Oxford Study mentioned earlier.

However, it is important to note that small effects over time can have a big effect. We are all witness now of how a temperature change of only 1.5 Fahrenheit had our icecaps melting, full rivers and lakes vanishing from the earth; we had to witness extreme drought, fires, hurricanes and tsunamis, all caused by our own environmentally disastrous habits.

Thus, the argument that a small effect means a negligible effect is hardly sustainable.


As you can see, the interpretation of the effects of social media on our well-being and mental health is rather complex. In order to exercise our critical thinking skills, though, it is important to not take everything for face value that we read in the news about the effects of social media on us or the iGeneration in general.

Keep those nuts and bolts in mind when you read your next article on the effects of social media on our well-being. Because this subject matter really is complicated.

#socialmedia #research #media

76 views1 comment