Exam 5 - Study guide

Class,

It is apparent from your performance on Exam 4 that you have the mechanics down and you have the write-ups down. For Exam 5, you will be tested on how well you know the SPSS procedures we have used already. I might give you some data and ask you to tell me what procedure to use. I might need to do something and ask you what procedure to use.

For example,
1) I might say that I have a column of continuous data and I want to run a procedure that will give me a boxplot (other than the boxplot command under graph menu - notice that the graph commands are not included in the list below). What is that procedure?
2) I might give you a distribution of data that is skewed, I might ask which procedure do I use to transform to reduce the skew? Depending on the distribution I show you the answer might be - Transform using a log function.
3) I might say that I have 3 independent measures and 2 dependent measures (and I would describe them in sufficient detail so you would know if they were repeating or not) and I want to determine if there is an interaction between the variables. What procedure would I use? Under what circumstances would I conduct simple effects? What post hoc test should I use Etc. Which would be the between-subjects factor? (I probably wouldn't ask all of this in one question. It is just an example of the types of questions I might ask to test your working knowledge of the procedures.
4) I might show you some output and ask which procedure might have generated the output (I will only us the output that we are familiar with from doing our homework and what we covered in class).


There is a good diagram on page 10 of Howell that may be helpful. You will also be tested on how well you know the various assumptions that are required for each statistical procedure. You will not be able to use the computer during the test, but you can bring in a page, single sided HAND WRITTEN notes (which you will turn in with your exam). You should already be very experienced with.

Data:
Sort cases
Select cases
Weight cases
Transform:
Compute
Random number seed
Recode
Analyze:
Frequencies
Descriptives
Crosstabs
One-sample t-test
Independent t-test
Paired sample t-test
One-way anova
Univariate (Anova) (Ancova)
Multivariate (Manova) (Mancova)
Repeated measures
Bivariate correlations
--Pearson's
--Spearman's
Linear regression
Factor analysis
Hiloglinear