PSYM221Z Introduction to Statistics Summative Assessment 2, 2026 | University of Exeter

PSYM221Z Summative Assessment 2

This assignment covers the material in Weeks 6-10, on linear regression.

Generative AI

This is an AI-assisted assessment: see here for full details of what this means https://libguides.exeter.ac.uk/referencing/generativeai . In short, this means you can use AI tools in specific ways if you choose so. It’s important to note that using AI tools is not mandatory or necessary to get a good mark for this assessment. But it is allowed, if you would find it useful, for these tasks:

• To proofread and correct spelling and grammar errors
• To assist with English language expression

Please note this differs from the list of permitted uses for other modules in your MSc. For this assessment you must refer to this list, not any others found elsewhere in your course.

If you do use these tools you MUST declare you use of the tools as part in the Generative AI declaration template table, which is included as part of the Assessment Brief. If you do not use generative AI tools leave the table blank.

Using generative AI tools for purposes other than those listed above is prohibited for this assessment. Using generative AI tools for purposes other than the above is Academic Misconduct and will be treated through the formal university processes. Similarly, using generative AI tools and then not including the declaration table is also considered Academic Misconduct.

 Question 1

A researcher named Sun is examining predictors of subjective health in daily life. 30 participants were tracked over a 2-week period using wearable devices and daily questionnaires. Sun averaged these data to produce a set of variables for each participant: subjective health (on a 1-100 scale where higher numbers indicate better subjective health), step count (average number of steps per day) and interactions with strangers (average number of interactions per day). Ethnicity (East Asian, South Asian, or European) was also recorded. The data are in the Daily.sav file.

Sun hypothesizes that participants with greater average step counts would report better average subjective health over the 2-week period.

a) (15 marks) Test Sun’s hypothesis using correlation analysis. Report the relevant statistical results in full.

b) (15 marks) Run a multiple regression model predicting subjective health from step count and interactions with strangers. What proportion of the variance in subjective health is explained by variance in step count and interactions with strangers combined? Report an appropriate model fit statistic to answer this question.

c) (12 marks) Based on your regression model in Question 1b, which of the two predictors shows the strongest association with the DV? Explain how you obtained this information.

d) (10 marks) For the model in Question 1b, Sun then tests the assumption of normality of residuals and concludes that this assumption has been violated for this model. Why might Sun have come to this conclusion? Use an appropriate graph to support your answer.

e) (8 marks) Two datapoints accidentally get deleted for the step count variable without Sun realising. When Sun then re-runs the model in Question 1b and checks her SPSS output, she notices a change in her ANOVA table that alerts her to the fact that her sample size has been reduced by 2. Where could Sun have found this information in the ANOVA table, and what change might she have noticed?

Based on previous research, Sun decides to include ethnicity as a predictor in the regression model. She wishes to test the hypothesis that both European participants and South Asian participants will report higher subjective health than East Asian participants.

f) (15 marks) Add two dummy coded variables to the dataset that will enable you to test Sun’s hypothesis about ethnicity. Then, run a regression model predicting subjective health from step count, interactions with strangers, and your two dummy coded variables. Write down the resulting regression

g) (8 marks) According to the regression model in Question 1f, what would be the predicted subjective health score for a participant of South Asian ethnicity with an average step count of 7658 steps/day who had an average of 1.02 interactions with strangers per day.

h) (17 marks total) Using hierarchical regression, compare the following two models:

• Model 1: The model from Question 1b predicting subjective health from step count and interactions with strangers.
• Model 2: The model from Question 1f predicting subjective health from step count, interactions with strangers, and ethnicity (the two dummy variables).

[i] (11 marks) If Sun wanted to maximize the efficiency of her model, would you recommend that she keep ethnicity (the two dummy coded variables) in the model, or exclude ethnicity from the model? Report relevant model fit statistics to support your answer.

[ii] (6 marks) The significance of the ‘social interactions’ variable changes when adding the dummy-coded ethnicity variables to model 2. Can you suggest any possible reasons for this change?

SPSS Output

<Please export your SPSS workings and paste here>

The post PSYM221Z Introduction to Statistics Summative Assessment 2, 2026 | University of Exeter appeared first on Students Assignment Help UK.