CS103-Sp01 day 18

CS 103 Day18 Friday May 4, 2001

Final problems: You may repair or redo problems I've looked at and marked lower than perfect (check). Ask me for help if you don't know what needs fixing. You can then submit your new work for regrade; no penalty for having done it twice. The goal is to (eventually) achieve completeness and correctness.

Class notes:
We have already learned the concepts and statistical terminology we are using; only need to see how to do it in the new setting.
Chapter 3

Sec. 3.2 p. 20-21 Crosstabs is the SPSS equivalent of Excel's Pivot Table. Contingency Table is another name for the table resulting from crosstabulation. (The result in a box is contingent on the row and the column it is in). You can get the row and column percents all in the same view, unlike Excel. If you know about Chi-square statistics, you do those here too.

Sec. 3.3 pp. 22-24 Descriptives for one variable. Discussed here is the Descriptives command. As in Ch. 2, most descriptives that involve putting things in order are under the Frequencies command. Not in text: SPSS is a historically very old package. Rather than change things people already knew how to do, newer things have been added in separate modules. It will be helpful to make a list of where to find all the descriptive items you've learned about. Other descriptive statistics are under (my favorite) ~~Statistics>Summarize~~ Analyze>Descriptive Statistics/Explore: Statistics: Check Descriptives, Percentiles, and Outliers to see what you can get. (Right-click over each option will tell you what you will get.)
Sec. 3.4 pp.24-28 Descriptives for subsets of a variable. Work through the book's example using Sample.sav. Not in text: Before leaving the Means dialog box (p.27-top-mid) choose Options, and move Medians into the list of Cell Statistics.

Note that the variable here for defining "subsets" is called "independent". In other places it will be called "grouping", or "factor". (SPSS procedures are NOT consistent in their language). One of the strengths of SPSS is its ability to analyze subsets easily.

Sec. 3.5 pp. 28-29. Standard (z) scores The Sample.sav file is a dumb set of data to standardize, with 4 values, but it serves to show the process. Note that this produces a new variable you can analyze just like the old ones.

Ch.4. Graphing data

4.1, pp. 31-35 Bar, pie, Histograms. Height.sav is in the SPSS stuff folder. Go through and repeat their analyses. This will get you started on the graphs. We'll get more elaborate next week.

Assignment 13 Due Wednesday, Day 20, May 9

A. (Crosstabs) Do Practice Exercise p. 22 top. Answer those questions, also, What percentage of married people are male? What percent of males are married? Hand in the table and your written answers.
B. (Descriptives) Do Practice Exercise p. 24 mid. Also find the interquartile range. Use whatever command(s) you need to to get all the results (see notes). Print and circle the desired results to Hand in.
C. (Descriptives for subgroups) Do Practice Exercise p. 28 top. Compute at the same time the medians. (see notes.) Print the table to Hand in.
D. (Z-scores) Do Practice Exercise p. 29 bottom. Write down the means on paper. Why should the mean of the z-scores (for the whole group) be (close to) 0 and the s.d. (close to) 1? Hand in.

E. (Bar graph and histogram) Ch. 4. Do (and print) a histogram and a bar chart for Model-year in Cars.sav (cf. p.18. In SPSS Stuff) Discuss the pros and cons of each, as a representation of the distribution. Hand in.

Hints and comments for Asst 13:

Crosstabs: What percent of A's are B's? All the B's you count must be A's also (B&A's). Divide that number by the number of A's.

Can't find Interquartile range? Hint: IQR, part of "exploratory data analysis," is relatively new in the statistics world.
Standard deviations from the mean: z-scores for the whole group are measured in standard deviations from the mean, so the mean should be 0 standard deviations from the mean, and the standard deviations should be 1 standard deviation from the mean. Roundoff error may produce slight differences.

separate

To Sievers Home Page

CS103-Sp01/day18.htm 5/3/01

This page belongs to Sally Sievers who is solely responsible for its content. Please see our statement of responsibility.