Woman with LaptopTwo research articles were recently published that shed more light on the so-called “Internet addiction”, a concept we’ve long lampooned here due to its continuing lack of scientific validity.

The first study (Dowling & Quirk, 2008) looked at one of the common measures of “Internet addiction,” used by nearly all researchers who’ve examined this phenomenon — the Young Diagnostic Questionnaire (developed by the originator of the disorder, Kimberly Young). The researchers administered the quiz to 424 Australian college students and discovered no statistical difference between “Internet addicts” and people who scored below the cutoff for “Internet addiction” (those considered “at risk”) in the amount of time spent online or in general psychological distress.

This means a researcher studying Internet addiction could be basing his or her results on a measure that cannot differentiate between people who supposedly have the disorder, and those who may not have the disorder. “At risk” is one of those squishy terms often used in research with quizzes like this to give a clinician some leeway in making the final determination about whether to diagnose a problem. But in research, such a group muddies the empirical waters and calls into question the validity and usefulness of such a measure.

The problem with the measure is likely its sensitivity. With only 8 questions, it can’t do a very good job in its attempt to differentiate between “normal” Internet use, and that which may lead to behavioral problems in a person’s life. The test also doesn’t account for different usage patterns in different phases of a person’s life. Today’s young adult is far more “connected” with Internet technologies than, say, most 65-year-olds (or even the young adult’s parents).

The second study examined both video game and Internet use in a group of 813 college students from six different institutions across the U.S. It did not use the Young Diagnostic Questionnaire, instead just asking, “On average, how many hours a day do you spend on the internet?” and then on a 5-point Likert scale to measure how they were spending their Internet time: entertainment (e.g., games, music, movies), headline news (e.g., national events, politics, international affairs), pornography, e-mail/instant messaging (IM), chat rooms, shopping, and school/work activities.

Noticeably absent from the list is social networking tools and sites, such as Facebook (which is arguably where a lot of college students spend their online time these days) and blogs/blogging. I’d also separate out email and instant messaging, since they are hardly the same medium or used for the same purposes (IM is far more social, while email tends to be used for general purposes).

Since I’m focused on Internet use in this entry, I won’t discuss the video game findings. The college students in this study self-reported they spent about 3 1/2 hours a day on the Internet (or nearly 1/6th of their day). By and far, most students spent most of their time on the Internet in email/IM or focused on school/work activities.

Here’s what the researchers further found:

Regarding internet use, results highlighted the need to more closely examine internet use in regard to the medium versus content debate that has existed for years regarding television use (Anderson et al. 2001). While there were findings for internet use in general (e.g., internet use was negatively related to self-worth), our findings were more consistent with content theory in that what the internet was being used for appeared to be significant in understanding its role in the lives of young people. In other words, there were different patterns of findings depending on how the internet was used.

For example, when the internet was used for chat rooms, shopping, entertainment, and pornography, there was a link to negative outcomes including more risk behaviors (both drinking and drug use), number of sexual partners, lower self-perceptions and self-worth, and poorer relationships with friends and parents; but when it was used for schoolwork, it was associated with a plethora of positive outcomes including less drug use, higher self perceptions and self-worth, and positive parent-child relationships for young men.

This is one of those “no-duh” moments in research. Since this was a correlational study, the researchers cannot infer causative relationships — they cannot say, for instance, that Internet shopping causes a person to be more sexually promiscuous. Indeed, one might make the equally valid argument that people who have lower self-esteem and engage in more risk behaviors (such as drinking) go online to find entertaining distractions. There may be specific personality types at work here (or some other unmeasured third factor), but we wouldn’t know it because the study only looked at a slice of a person’s life and their behaviors.

Which the researchers eventually acknowledge:

It is possible that using the internet for a specific purpose may lead an individual to engage in certain behaviors, but it is equally possible that certain characteristics (such as not feeling socially accepted or having low self esteem) might lead one to withdraw into the ‘‘safer’’ social world of chat rooms and pornography on the internet. It may also be possible that the various purposes of the internet are relatively benign and only begin to appear problematic once they start to replace other things that might be beneficial for young people (e.g., class attendance and homework for students, reading, exercise, work, and face-to-face social interactions).

I think that last statement is reaching (and a value judgment on the part of the researchers), however, as there’s been little evidence to suggest that online social interactions are inherently of a lesser quality than face-to-face interactions, or that reading online is somehow less educational or beneficial than reading a book. It all depends on whom you’re interacting with (and for what purpose), and what you’re reading. A hundred face-to-face social interactions in the school’s kitchen isn’t likely to compare, qualitatively, to a single online social interaction with a close friend for an hour.

I think the key from this second study is that it’s not only measuring how often a person uses the Internet, but what specifically they are doing online that’s important. If people are using the Internet’s “pro social” tools (such as email and social networking websites), then it’s no wonder they’re likely to report better social relationships with others.

All Internet use — even heavy Internet use — is not inherently bad or problematic or “addicting.” It’s a far more subtle relationship, and one that most current “Internet addiction” measures are not accounting for. The broad measures commonly used in research to study online use need far more specificity and sensitivity to understand the potential benefits and drawbacks of the Internet. For over a decade, researchers have been using a magnifying glass when they really should’ve been using an electron microscope.


Dowling, N.A. & Quirk, K.L. (2008). Screening for Internet Dependence: Do the Proposed Diagnostic Criteria Differentiate Normal from Dependent Internet Use? CyberPsychology & Behavior, 12(1). DOI 10.1089/cpb.2008.0162.

Padilla-Walker, L.M., Nelson, L.J., Carroll, J.S. & Jensen, A.C. (2009). More than just a game: Video game and Internet use during emerging adulthood. Journal of Youth & Adolescence. DOI 10.1007/s10964-008-9390-8.