0
votes
Accepted
We agree that effect estimates with confidence intervals can be great summaries to report, as you suggest in the first paper. As we mention in our paper, there has been some research already on human intuition for confidence intervals (http://psycnet.apa.org/?&fa=main.doiLanding&doi=10.1037/1082-989X.10.4.389). We...
10 November 2014
0
votes
In a lot of scientific applications, yes. In general though, it might depend on if you’re more interested in inference (trying to describe a mechanism in a population) or prediction (trying to anticipate characteristics of a new sample). For some scientific applications,...
10 November 2014
0
votes
This is a great point. Any difference in visual stimulus may affect user’s perception of significance. Ideally, all visual stimuli should be controlled for. A better approach might have been to scale each sample to have the exact same maximum and minimum value....
22 October 2014
0
votes
To elaborate on our survey, students were specifically asked “Suppose you fit a linear model relating the two variables displayed in the plot below. Would the P-value for the regression coefficient be significant at the 0.05 level (P less than 0.05)?” The...
0
votes
This could be a great way to account for variability within plot category. We wrote up a modified version of our code that implements a simple version of this idea, assuming the same variance for the random intercept terms in all plot styles. In our calls to the...
22 October 2014

Sort order

Top subject areas - People