probable typo in "So, you did some GLMs & compared with AIC. Congrats!"

[ashander]

The visualization example at the end of this post

Line 301 here should be ggplot(mavg.pred) + not ggplot(d) +

ashander.github.io/posts/2015/10/model-selection-glms-aic-what-to-report/index.html

Lines 294 to 310 in c398f07

	m2.pred <- broom::augment(m2)
	m2.weight <- reported.table[2, "weight"]
	m3.weight <- reported.table[3, "weight"]
	mavg.pred <- m2.weight$weight * m2.pred[ , -(1:2)] + m3.weight$weight * m3.pred[ , -(1:3)]
	mavg.pred <- cbind(m3.pred[1:3], mavg.pred)
	library(ggplot2)
	ggplot(d) +
	geom_point(aes(Dept, Admitted / (Admitted + Rejected), color=Gender,
	size=Admitted + Rejected),
	position = position_dodge(width = 0.5)) +
	geom_pointrange(aes(Dept, plogis(.fitted),
	ymin = plogis(.fitted - 2 * .se.fit),
	ymax = plogis(.fitted + 2 * .se.fit),
	shape=Gender),
	position=position_dodge(width = 0.5), data =mavg.pred, alpha=0.4) +
	theme_minimal() + scale_color_manual(values=c(&quot;blue&quot;, &quot;orange&quot;))</code></pre>