Refer to the Behavioral Neuroscience (Feb. 2004) study of the drug scopolamine’s effects on memory for word-pair associates, Exercise 14.13. Recall that a completely randomized design with three groups was used—group 1 subjects were injected with scopolamine, group 2 subjects were injected with a placebo, and group 3 subjects were not given any drug. The response variable was number of word pairs recalled. The data for all 28 subjects are reproduced in the table.

a. Rank the data for all 28 observations from smallest to largest.

b. Sum the ranks of the observations from group 1.

c. Sum the ranks of the observations from group 2.

d. Sum the ranks of the observations from group 3.

e. Use the rank sums, parts b–d, to compute the Kruskal–Wallis H statistic.

f. Carry out the Kruskal–Wallis nonparametric test (at α = .05) to compare the distributions of number of word pairs recalled for the three groups.

g. Recall from Exercise 14.13 that the researchers theorized that group 1 subjects will tend to recall the fewest number of words. Use the Wilcoxon rank sum test to compare the word recall distributions of group 1 and group 2. (Use α = .05.)

Data from Exercise 14.13

The drug scopolamine is often used as a sedative to induce sleep in patients. In Behavioral Neuroscience (Feb. 2004), medical researchers examined scopolamine’s effects on memory for word-pair associates. A total of 28 human subjects, recruited from a university community, were given a list of related word pairs to memorize. For every word pair in the list (e.g., robber-jail), there was an associated word pair with the same first word but a different second word (e.g., robberpolice). The subjects were then randomly divided into three treatment groups. Group 1 subjects were administered an injection of scopolamine, group 2 subjects were given an injection of glycopyrrolate (an active placebo), and group 3 subjects were not given any drug. Four hours later, subjects were shown 12 word pairs from the list and tested on how many of the associated word pairs they could recall. The data on number of pairs recalled (simulated based on summary information provided in the research article) are listed in the next table. Prior to the analysis, the researchers theorized that the mean number of word pairs recalled for the scopolamine subjects (group 1) would be less than the corresponding means for the other two groups.

Transcribed Image Text:

SCOPOLAMINE Group 1 (scopolamine): 5 8 8 6 6 6 6 8 6 4 5 6 8 10 12 10 9 7 9 Group 2 (placebo): 10 Group 3 (по drug): 8 9 11 12 11 10 12 12