Table 3.

Model results for four negative binomial generalized linear models, testing the effect of journal impact factor (JIF) and publication year on the contribution of the four main online platforms to AASs.

platform	variable	estimate	2.5%	97.5%
Twitter	intercept	−884.623	−937.199	−832.174
	year	0.440	0.414	0.466
	JIF	0.360	0.299	0.422
Facebook	intercept	−621.090	−895.124	−629.177
	year	0.308	0.311	0.443
	JIF	0.147	0.162	0.469
news	intercept	−1667.202	−2159.574	−1167.099
	year	0.827	0.579	1.072
	JIF	0.443	−0.153	1.029
blogging	intercept	−1276.000	−1498.542	−1054.870
	year	0.633	0.523	0.744
	JIF	0.347	0.062	0.628