Assume that you are interested in investigating the common underlying dimensions to well-being. You recruit 250 participants who complete a questionnaire that is supposed to tap into different aspects of well-being styles. Items on this questionnaire are rated on a 1 – 10 point likert scale. The items included the following characteristics of well-being: sleep quality, exercise, eating habits, satisfaction with work and family and so forth. The data are in the file: well-being.sav. Perform a factor analysis (with principal axis factoring), using appropriate rotation, to investigate the latent structure of the data. Each of the questions carries 5 marks. (a) Comment on the appropriateness of this dataset for FA, in terms of sample size and sampling adequacy. Provide relevant statistics to justify your answers. (b) How many factors should be retained? What method of rotation have you used and why? Explain your answers quoting any relevant statistics. (c) Which items load on which factors? How might you interpret these results in terms of the latent structure underlying well-being? (d) Is each of the factors reliable? Report on the reliability figures for each subscale. 2. GCSE marks have been argued to be influenced by gender. Assume that you were interested in examining this hypothesis by looking at how Gender predicts GCSE marks using a Mixed Linear Model analysis that takes into account random variability brought about the type of school pupils attended: Independent, State, and Other. Could type of school be influencing the effect of gender? Carry out the Mixed Effects analysis by adding random variability to the intercept only. What do you conclude? The data can be found in the file: school.sav. [15 marks] 3. In a study that looked at the effects of adaptive cognitive training using a dual n-back task on cognitive flexibility and emotional vulnerability, you are interested to see how the rate of improvement on the training task mediates the relationship between self-reported emotional vulnerability (assessed using a 0 – 100 mm scale: higher scores indicate greater emotional vulnerability), and avoidance of threat (higher scores indicate greater avoidance) as measured by attentional bias. Conduct a mediation analysis on the data in the file Trainingimprovement.sav to show if rate of training improvement mediates the relationship between emotional vulnerability and avoidance of threat. Report on the effects of interest. What conclusions can you draw from this mediation analysis? [22 marks] 4. There is continuing debate over whether stressful life events in interaction with genetic risk factors for depression predict depressive severity in children with clinically diagnosed depressed mothers. The data in the file RiskSeverity.sav shows data from 60 children who were genetically screened for risk of developing depression (Risk), the average number of traumatic events they reported they had in the last two years prior to testing (events), and the severity of the depressive symptoms they experienced (Severity). Do ‘events’ moderate the effect of genetically determined risk factors in predicting depressive severity? Corroborate your answer with a simple slope analysis to show the moderating effect (if any) of Events on Risk factors, in predicting Severity. What conclusions can you draw from your moderation analysis? What do your results say about the Environment X Gene interaction in predicting depressive severity? [22 marks] 5. The table below shows the data from 7 experiments that looked at how attentional bias training away from threat (hours spent training per day for a week) would result in a change/reduction in attentional bias for angry faces in socially anxious individuals (change/reduction: post – pre test, with higher scores indicative of greater reduction in attentional bias). You intend to use this information to conduct a meta analysis based on effect sizes from the studies obtained. Each question carries 7 marks. Study Training per day (hrs) Reduction/Change in AB (ms) SD t N p A 5.5 15.96 5.4 5.05 40 0.001 B 3.1 14.74 2.1 5.26 55 0.001 C 4.3 15.62 4.7 4.10 60 0.0001 D 2.1 13.78 3.6 1.78 14 0.05 E 3.3 13.44 3.2 5.23 30 0.012 F 7.8 18.32 3.3 6.20 32 0.0001 G 1.9 -2.93 3.2 -0.13 18 0.66 (a) Calculate and interpret a composite measure of significance using Fisher’s Combined Test. Show all your working. You will need to consult a table of the ?2 statistic, which can be found in the appendices of most statistics textbooks (and the class handout). . (b) Calculate Cohen’s d for each of the studies. Calculate the composite effect size (expressed in Cohen’s d) and its 95% confidence interval. (c) Make a scatterplot showing the relationship between the number of hours of training provided by each experiment and effect size. What is the minimum training hours needed for a meaningful change?
