Hi all:

I'm sorry, again, that I was not present on Wednesday.  I hope the audio recording was audible and clear enough.

I wanted to briefly address your first critiques (Castro & Rice) in class, but since we didn't meet, I'll make my comments here.

Everyone received full credit (2/2), although there was some variability.  For this first assignment, I really accepted all efforts, but wanted to provide some feedback on the characteristics of stronger vs. less strong critiques.  Since you all have your next critique to turn in on Wednesday, I think this is timely for you :-).

(One small note...if you went beyond one single-space page, strive for more "economy" next time.  Not because these are hard to read--they are an enjoyable breeze, actually--but to force yourself to get succinct with the critiques.  That is a skill that will definitely help you if you ever do journal or grant reviews!)

The best critiques were those that were really INTERNAL to the other words, that used some of the tools we had explored in class and readings to address ways that the analyses and write-up could have been done better.  Less effective critiques (just because they were generic, and could really be applied to any article) were those that focused on aspects of presentation (e.g., graphs).  These latter critique points may be appropriate!  I am not trying to discourage them!  But, they don't so deeply delve into the actual paper, so you might want to wean yourself from the habit of these "easy" critiques that apply to most papers.

Some examples:

Stronger points
- Would Pillai's have been more appropriate than Wilks?
- Subscale multicollinearity needed to be more critically evaluated
- recruitment strategies were inconisitent across minority groups
- test of homogeneity (e.g., Box-M) absent
- self-reported GPA could not be validated
- normality assumptions not tested
- failed to control for mediators (more detail was offered)
- scales not validated across ethnicity groups
- subscale reliabilities do not seem high
- interactions not tested (homogeneity of covariance)

More generic points
- Was sample size even adequate for this kind of paper?
- No corrections for multiple comparisons (true...but most readers could just do that in their head)
- Gender distribution unequal
- effect sizes are small
- generalizability to the genral population not known
- sample not random
- university students are not generalizable
- self-report is biased
- absence of graphics

Now, all these comments are good and valuable.  But the second set is sort of the "boilerplate" critique that applies to 90% of the psychology literature :-).  The first set of critique points emerged from a deeper reflection on the specific analyses presented in the paper, and the way those assumptions were tested and justified.

So, as you move forward with your weaknesses, try to aim for more of the former than the latter.