In Chapter 14, Problem 18, p. 375, an experiment was conducted to evaluate the effect of…

In Chapter 14, Problem 18, p. 375, an experiment was conducted to evaluate the effect of decreases in frontalis muscle tension on headaches. The number of headaches experienced in a 2-week baseline period was recorded in nine subjects who had been experiencing tension headaches. Then the subjects were trained to lower frontalis muscle tension using biofeedback, after which the number of headaches in another 2-week period was again recorded. The data are again shown here.

In that problem, the sampling distribution of  was assumed to be normally distributed, and the analysis was conducted using the t test. For this problem assume the t test cannot be used because of an extreme violation of its normality assumption. Use the Wilcoxon signed ranks test to analyze the data. What do you conclude, using a = 0.052 tail? clinical, health

Chapter 14, Problem 18

Since muscle tension in the head region has been associated with tension headaches, you reason that if the muscle tension could be reduced, perhaps the headaches would decrease or go away altogether. You design an experiment in which nine subjects with tension headaches participate. The subjects keep daily logs of the number of headaches they experience during a 2-week baseline period. Then you train them to lower their muscle tension in the head region, using a biofeedback device. For this experiment, the biofeedback device is connected to the frontalis muscle, a muscle in the forehead region. The device tells the subject the amount of tension in the muscle to which it is attached (in this case, frontalis) and helps them achieve low tension levels.After 6 weeks of training, during which the subjects have become successful at maintaining low frontalis muscle tension, they again keep a 2-week log of the number of headaches experienced. The following are the number of headaches recorded during each 2-week period.

a. Using α = 0.052tail, what do you conclude? Assume the sampling distribution of the mean of the difference scores is normally distributed. Assume a nondirectional hypothesis is appropriate because there is insufficient empirical basis to warrant a directional hypothesis.

b. If the sampling distribution of is not normally distributed, what other test could you use to analyze the data? What would your conclusion be? clinical, health