Explore the associations between gender (FEMALE) and children’s academic, social, and behavioral characteristics at kindergarten entry (C1RSCALE, C1MSCALE, WKINCOME, P1AGEENT, T1LEARN - or other continuous variables of interest to you). Use a standardized (z-scored) version of these variables when appropriate. How do male and female kindergarteners differ? Be sure to distinguish between statistical significance and practical significance. Place your findings into a single table.
*I can't make office hours on Thursday afternoon, so if you need extra help, I'll be hanging out in the library, 2nd floor from around 11am to 1pm on Thursday morning (some time slots within this are already booked...). Or email me.
-Megan
Subscribe to:
Post Comments (Atom)
10 comments:
Hey Megan, I'm not entirely clear on the three decimal place rule. When listing means would I list -0.762 for a z-scored reading test variable, and then 52331 (with no decimals) for an income variable in the same table. That seems odd to me, because I always thought I should be consistent about how many decimals I list for multiple variables in the same table. Let me know. Thanks, Alex
Hi Alex,
I'm not sure it makes sense to include decimal points when talking about income... But, if you have a system that you feel really strongly about, go ahead - use your best judgment.
-Megan
oops, forgot to add that your understanding of Doug's decimal point guideline is right on. up to you if you want to use it.
Hi Megan, Did I understand correctly that Doug only wants us to explore continuous variables of our own choosing if they are normally distributed? It makes some sense because of the t-tests' validity I guess. But I just feel confused about whether I heard that, I guess since WKINCOME is of course not normal at all....Thanks, Karen
Hi Karen,
Yes, you're right - t-tests assume that the variable is normally distributed. For now, don't worry about the fact that the income measure isn't normal, but if you're adding other variables try to only include normally distributed ones. (Doug will be talking about transforming non-normal variables and I think specifically income later this semester.)
Megan
Hi Megan,
Is Table 1 supposed to look like Table 1 on page 29 of Doug's gender paper, or is it supposed to more closely reflect that output table (the long one with Levene's test, the t-test, and confidence intervals) from SPSS?
Thanks in advance.
Hi -
The table should be like Doug's - mean, sd, and stars to indicate significance - don't include all the other extra information in the output.
-Megan
Hi - I've received a couple of questions via email that I thought I'd post.
1) Do we need to include confidence interval in the 'finding' section.
No, not required.
2) Should we put 'mean difference' or 'means of boys and girls in two rows' in the table ? What about SD?
Use two columns, Boys and Girls - see Doug's table in his k literacy/gender paper, p 29 for an example.
3) Should we add 'stars' in the table to indicate statistical significance?
Yes, always. See Doug's paper for examples.
4) What to say about 'practical significance'? Should we say there is practical significance because the difference is substantial?
You don't have to use the words "practical significance," but just make sure you're addressing both the significance and the magnitude in your discussion of the relationships.
5) Should we z-score Approaches to Learn (it seems that Prof. Ready's example didn't but did that make sense?)
Up to you. Does the scale make sense to you or does it make more sense to talk about it in terms of standardized units?
Hi, I would like to know how do we put those stars on the table? I saw that Doug has put some next to the mean and some next to the SD, how do we determine where do we put them?
Also, since we include SD in our table, do we have to mention SD in our results section?
Thanks for your help!
Hi anonymous,
Look at the key at the bottom of the table. Stars are for significant differences - between the two means or between the two SDs. Definitely include them for the means, up to you if you want to include them for the SDs. Mention the SD in your results if it's interesting.
Post a Comment