1. Why not just compute t-tests among all pairs of means instead computing an analysis of variance?

2. If an experiment is conducted with 5conditions and 26 subjects in each condition, what are dftotal and dfwithin?

3. An important study in the memory literature is an old study by Eysenck (1974) in which he compared the recall of older participants under two levels of processing of the material (low and high). He demonstrated that when asked to perform a higher level of processing of a list of words, participants showed better recall at a later time. Another aspect of Eysencks study compared Younger and Older participants on their ability to recall material. Oneway to look at the Eysenck study mentioned is to compare four groups of participants. One group consisted of Younger participants who were presented the words to be recalled in a condition that elicited a Low level of processing. A second group consisted of Younger participants who were given a task requiring the Highest level of processing. The two other groups were Older participants who were given tasks requiring either Low or High levels of processing. The data follow:

Younger/Low8 7 4 6 7 6 5 7 9 7

Younger/Hi21 19 17 15 22 17 22 21 18 21

Older/Low9 8 7 8 10 4 6 5 7 7

Older/Hi10 19 14 5 10 11 14 15 11 10

Conduct a one-way analysis of variance on these data and interpret the results.

4. Langlois and Roggman (1990) took facial photographs of males and females. They then created five groups of composite photographs by computer-averaging the individual faces. For one group the computer averaged 32 randomly selected same-gender faces, producing a quite recognizable face with average width, height, eyes, nose length, and so on. For the other groups the composite faces were averaged over either 2, 4, 8, or 16 individual faces. Each group saw six separate photographs, all of which were computer-averaged over the appropriate number of individual photographs. Langlois and Roggman asked participants to rate the attractiveness of the faces on a 15 scale, where 5 represents very attractive. The data have been constructed to have the same means and variances as those reported by Langlois and Roggman.

Data on rated attractiveness

Group 1: 2.201 2.411 2.407 2.403 2.826 3.380

Group 2: 1.893 3.102 2.355 3.644 2.767 2.109

Group 3: 2.906 2.118 3.226 2.811 2.857 3.422

Group 4: 3.233 3.505 3.192 3.209 2.860 3.111

Group 5: 3.200 3.253 3.357 3.169 3.291 3.290

a) Run the appropriate analysis of variance.

b) What do these data tell us about how people judge attractiveness?