For this first assignment, you're conducting two sets of analyses:
1) Describe the children in your sample in terms of socio-demographic and academic characteristics (see handout for the list of measures).
2) Explore how children who are living in poverty differ (on average) across several dimensions from those who are not. Use t-tests and focus on the continuous measures (see handout for the list).
For examples of tables, take a look at the Ready et al. 2005 paper, page 29. Also see the handout, "Making Basic Tables" on the classweb files.
Submit only Table 1, Table 2 (in word format! not your spss output), and your syntax to Megan on Thursday, October 15th by 5 p.m.
Subscribe to:
Post Comments (Atom)
2 comments:
I'm trying to understand when to use Z scores. I know that they help us to be able to read the test scores even if the scores are reported using a variety of scaling systems but do we also use them for the age of the mother and the number of siblings? If so why or if not why not?
Generally, we use z-scores to compare tests that have different scales. But it is also useful when the metric is unknown. If you think about how you'll be reporting results, does it make more sense, or is it more meaningful, to report the mean on the original scale or in standard deviation units?
For example - reporting that females have a mean of 47 points on the fall test makes less sense than reporting that females have a mean of 0.17 SD. Right? In contrast, reporting that females have a mean age of 58 months makes a lot more sense than reporting that they have a mean age of 1.12SD.
Post a Comment