A study just published in the journal CyberPsychology and Behavior has reviewed all of the available scientific studies on internet addiction and found them to be mostly crap. And not just slightly lacking, really pretty awful.
To quote from the research summary:
The analysis showed that previous studies have utilized inconsistent criteria to define Internet addicts, applied recruiting methods that may cause serious sampling bias, and examined data using primarily exploratory rather than confirmatory data analysis techniques to investigate the degree of association rather than causal relationships among variables.
Rather disappointingly though, the authors just suggest that better research is needed when it's quite obvious that the whole concept is fundamentally flawed.
