In earlier work, we saw a wide range of citation usage habits, with order of magnitude differences between wikis. As so-called 20% time, I'd like to dig into this effect and try to understand why this might be the case.
The first question is to find the average citation density each article, and its distribution for wikis being considered. This might require a new approach, since we cannot simply look for <ref> tags in wikitext. Once we know the citation density, we can plot a curve of citation clicks per pageview, vs. citation density.
It might also be possible to record the citation's index and Y-position on the page, if we want a finer-grained picture of how reader fatigue interacts with citation usage.
Rough ideas:
- Hypothesis 1: Differences in citation usage across wikis have endogenous explanations. For example, fewer footnotes in an article correlates with fewer reference link clicks.