## 6.4 Lab Report Instructions

### 6.4.1 R Script

1. Calculate participants’ mean scores on the 5-item WLSE measure. Do this separately for the two measurement occasions.
2. Calculate M and SD for the pre- and post-psychotherapy means, and M, SD, skew, and kurtosis of the difference scores.
3. Use t.test() to conduct a paired t test of the null hypothesis that average pre-to-post change in WLSE scores is zero.
4. Compute Cohen’s dz for the change in WLSE scores.
5. Produce a histogram of the WLSE difference scores.
6. Produce a Q–Q plot of the WLSE difference scores.
7. Conduct a Shapiro–Wilk test on the difference scores.
8. Produce a boxplot of pre- and post-psychotherapy WLSE scores.

### 6.4.2 Results

Write a 1–2 paragraph results section that reports the results of your analyses. Include:

• M and SD of WLSE at each measurement occasion.
• Results of the t test, the minimally sufficient set of statistics for constructing the test.
• Standardized effect size (i.e., Cohen’s dz) with associated confidence interval.
• Results of your assessment of the assumption of normality, including the Shapiro–Wilk test and other statistics you used to make inferences about the distribution of scores in the population.
• Appropriate figure visualizing scores at pre- and post-intervention. This figure should be referenced in the text of the results.

### 6.4.3 Discussion Questions

1. Does this study support the primary hypothesis? Restate the hypothesis in plain language as part of your answer.
2. What is one potential source of bias that could affect how these results are interpreted? I.e., in addition to the theoretical mechanism proposed in the introduction to the study, what else could account for the presence or absence of the effects you observed?
3. Given the results of this study, is it safe to conclude that adding psychotherapy will lead to better health outcomes for participants of weight-loss interventions? Identify what this study does not tell us that we would need to know to make strong conclusions in this regard.