Assignment Three, Quiz on Friday Oct 3d

Bring a printouts of your unedited list file to the quiz; Use set width=80 length=none. to avoid unwelcome form feeds. Also bring a printout of your raw data file. Bring a calculator. Don't write on the printouts, except for your name and student number. If anything besides your name and student number is written on your printouts, you will be charged with academic dishonesty and turned over to Dean Krull.

In a study concerned with determining the effectiveness of varying motivational levels on the acquisition of arithmetic skill, twenty-seven children were randomly assigned to 3 groups of equal size. Group 1, the control condition, received a low amount of classroom motivation. Subjects in Group 2 were exposed to motivational levels that independent persons judged to be medium. Finally, the Group 3 subjects were exposed to high motivational levels. Then all the children were given an arithmetic test. Here are the data.

Group 1: 4, 5, 4, 3, 6, 10, 1, 8, 5

Group 2: 12, 8, 10, 5, 7, 9, 14, 4, 1

Group 3: 3, 4, 6, 8, 6, 5, 3, 2, 2

Make an SPSS command file to read the data, provide variable labels and value labels, and do an analysis to determine how motivational level is related to average performance on the arithmetic test. Note that you want to know not only whether there is a significant difference among groups, but also which group means are significantly different from which other means. For the question about differences among groups, you should be able to give an exact p-value. For the question about whether a given pair of groups shows a significant difference, you should be able to give an inequality (like p<0.05 or p>0.05). You should be able to say what proportion of the variation in arithmetic scores is explained by the differences among means. For this you will need a calculator.

You might be asked "What do you conclude from this study?" The answer is a statement about motivational level and performance on the arithmetic tests. Pretend you are talking to a boss who does not like to feel stupid, who knows no statistics and who will NOT be impressed by a lot of statistical terminology. Rather, if you do not answer the question directly in simple language, you are fired. And this is how your answer will be graded.

Again, bring a printout of your list file to the quiz; also bring a printout of your raw data file. Bring a calculator. Don't write on the printouts, except for your name and student number. If anything besides your name and student number is written on your printouts, you will be charged with academic dishonesty and turned over to Dean Krull.