Homework 2: SPSS
Twelve twins from around the country are located. In an effort to promote the virtues of Spam™, a particular food by-product company conducts an experiment. For each pair of twins, one twin is fed Spam™ twice a year, and the other twin is fed congealed pond scum (controlling for consistency and general appeal). After the year, the two hapless twins’ IQs are measured. Here are the results, where each row shows the data from one pair of twins.
Pair No. /Spam™
/Scum
01 / 104 / 11202 / 110 / 108
03 / 96 / 91
04 / 113 / 114
05 / 120 / 128
06 / 111 / 115
07 / 103 / 108
08 / 93 / 97
09 / 102 / 107
10 / 109 / 107
11 / 110 / 118
12 / 115 / 121
Note: Please hand in printouts of all data (input) files, all output files, and all tables and graphs.
1) Make a scatterplot of the twins’ IQs. Label the X and Y axes "Spam-fed Twin" and "Scum-fed Twin" respectively.
2) Are the IQs for twins from the same pair significantly correlated? Find the coefficient of correlation.
3) Is there a significant difference in IQs due to diet differences? Use a paired samples t-test.
4) Is there a significant difference in IQs due to diet differences? Use an independent samples (unpaired) t-test.
(Hint: you will have to include a grouping factor to do this – see Example 2 of the tutorial).
5) You should have gotten opposite answers to the question "Is there a significant difference" in Questions 3) and 4). How is this possible, and why did this happen in the current example?
6) Is there a significant difference in IQs due to diet differences? Use an ANOVA (either a One-way ANOVA or a univariate General Linear Model). Again, you will have to include a grouping factor to do this.