Section: standards and consistency

[x] "something to keep in mind" way too weak.
[x] Section doesn't really handle the issues of consistency vs standards separately. Before we worry about whether altmetric and impact story count tweets in quite the same way, you should make the very obvious point that almetric gives you a single "magic" number, while impact story gives you a set of metrics (how do we compare altmetric score of 162 to an impactstory report of an article?). This is a kind of fundamental philosophical divide that gets into "what counts" before we get into the technicalities of how to count it. Though you focus on altmetric's individual scores (tweets, cites, etc), it puts quite a bit of emphasis on the combined weighting.
[x] Analysis here looks great, decent size dataset to see both individual outliers and general patterns.
[x] add png versions of the plots, (just for github viewing, not for the tex, note that knitr device can take a list, dev=c("CairoPNG, CairoPDF"))

sckott / isqaltms