kparrish92
commented
3 years ago Great work, Patrick! Congrats on the QP and on taking the next step in the process. It has been a pleasure to have had classes with you these two years. Hard to believe they're over! Please see my comments, questions and suggestions below.
Comments:
-
The document knitted - yay!
-
The generate tidy data script all works on my machine - yay!
-
As always, your writing quality is very good. All sections were very well explained and organized.
Questions:
- You mentioned in the manuscript that lexical frequency was standardized. How was this done?
- Did you consider log transforming lexical frequency, due to the wide range in values reported?
Suggestions:
-
You reported main effects, but these seem to be the parameter estimates of the models you ran. My understanding is that main effects and interactions are derived from nested model comparisons, starting with either a null or inclusive model and either adding or subtracting one variable at a time. This way, it is possible to find the best fitting model and you can report those parameter estimates.
-
Based on your later nested model comparisons for production, it seems like the best fitting model is one that includes only frequency and proficiency, in which there is a main effect for frequency, but not for proficiency, but there is a proficiency*frequency interaction.
-
This is what you reported, but the parameter estimates for the best fitting model would be different. Also, reporting main effects and interactions in prose should come from the anova table (or I guess it’s called the analysis of deviance table), rather than the parameter estimates of the model.
Great job, Patrick! It all looks great! Below, I´m leaving some comments.
Slides:
Everything knitted well. Also, the presentation looks very clean and clear. Good job!
Manuscript I ran into some issues when trying to knit the papaja file for the manuscript. First, I got this error regarding line 123.
Then, when I looked at the name of the column, I saw that it had a c in front of “Regulares”. However, I am not sure how that got there, because I went back to your repo on Github and I cannot see this “ï…” there.
After I finally managed to knit your file to Word, I saw the following:
The references and the appendix are combined (I believe you mentioned that that was happening…)
The words that you wrote with tilde in R got changed to very weird symbols.
I read the manuscript and it looks great – very clear! Good job!