One possibility is to have students use a commercial statistics program such as SPSS to compute *t* tests. If you want students to use SPSS, you may want to use the following, short (7 page) tutorial that shows students how to set up a *t* test in SPSS and how to interpret the output. Although the tutorial is excellent, students will have fewer problems if you let students know that "test variable" is their dependent variable, that "grouping" is their independent variable, and that, on the grouping variable, they can code the control group as "1" and the experimental group as "2."

**Note: ** If your students do not know how to enter data into SPSS, you might want them to go through the tutorial at

http://www.statsguides.bham.ac.uk/howtoguides/WebPages/HTG/HTG_FileMan.htm before tackling the *t* test.

Rather than using a complex, commercial program, you can have
students calculate statistics using programs that are on the web. Our two favorite online *t* test calculators are

- VassarStats and
- GraphPad.

Alternatively, could download the authors' program that does an independent

- Windows version of statistics program
- Macintosh version of statistics program

- Windows version of tutorial
- Macintosh version of tutorial

If you want to have students use a computer without losing sight of the means and standard deviations, you could have students input the means and standard deviations into this modified *t* test calculator or, you could give students a gentle introduction to *t* tests by using this Excel spreadsheet.

Teach writing, SPSS, or both using Larry P. Wiley's tutorial. His tutorial shows students a brief Results section , and then lets students click on different parts of the Results section to get additional instruction. For example, students can choose to see what the SPSS output of the original analysis looked like and can even choose to learn how to have SPSS analyze data from a simple experiment.

Another way to use the computer is to show them the effects of violating *t* test assumptions. You can use the following applet to show students that, if sample size is adequate, the assumptions of normality and equal variances are not that critical.

Yet another way to use the computer is to have students use a computer to set up a simple experiment. An article showing how this could be done is

Rittle, R. H. (1990). Computer literacy in the psychology curriculum: Teaching a database

language for control of experiments.Teaching of Psychology, 17, 127-129.