5.2.2 Analytic Data Quality Implications & Solutions

In the above, I’ve summarized the ways scholars have estimated validity and reliability in social network data, and a number of common findings from that literature. The remainder of the chapter turns to discussing some of the analytic implications of those patterns, and solutions for mitigating their impact. When asking how some phenomenon is (causally) related to another, the reality is that our findings can tolerate some noise in the data we rely on. What follows does not attempt to provide a systematic road map of when you can and cannot expect your analyses to be robust to such limitations.112 Instead, the aim will be to provide a few illustrative examples each from cases where data limitations were either particularly damaging or robust to the scientific uses to which network data were applied.