Reference no: EM132862504

PSYU3349 Design and Statistics - Macquarie University

Research Questions

1. Some clinicians believe that therapeutic benefit may derive not only from the active component of the therapy but the duration of therapy. In particular, it is believed by many that longer duration of therapy may enhance expectancy of therapy, which translates into better patient outcome in the placebo/nocebo conditions. Determine whether there is evidence to support this belief by reporting on whether the duration of therapy alters the difference between therapy groups (group) with respect to change in somatic symptom score (chg_sss). In particular we are interested in whether the difference between sham placebo and neutral conditions and the difference between sham nocebo and neutral conditions are altered by duration of therapy (durnrx).

2. Understanding factors that account for variance in placebo response with respect to change in quality-of-life (chg_qol)
a. How well does the combination of psychological and demographic constructs (see Table 1) explain change in quality-of-life? Are any of these factors redundant as predictors, and if so which? Do any of these predictors appear to have substantially stronger effects than others in predicting change in quality-of- life, and if so which? Note that model reduction is not required for this question.

You just need to use whatever metrics you have learned to indicate which variables might be redundant, which are contributing relatively little to the explained variance and which are contributing substantially.
b. What is the unique explained variance in chg_qol response for the combined psychological constructs after controlling for the combined demographic constructs?
c. What is the unique explained variance in chg_qol response for the combined demographic constructs after controlling for the combined psychological constructs.
d. Comment on the degree of shared explained variance in chg_qol by psychological and demographic constructs.

Attachment:- Design and Statistics.rar