- Demonstrating your understanding of Bayesian inference and Bayesian hypothesis testing.
- Reporting your data analysis plan, results and conclusions as if you are writing a lab report.
The original data for this assessment come from ongoing teaching research. This research is examining different teaching methods in Psychology Introduction 1 (PSYC1010) at The University including enabling students to take a quiz multiple times, giving students relevant practice quizzes, and other methods. In addition to pedagogical methods, various other factors were examined in this research including sleeping behaviours, resilience factors, psychological distress, and other enrolment factors.The three research questions for this assessment are:
- Do reslience and psychological distress contribute to good academic performance?
- Do exam scores differ between Ourimbah, Callaghan, and Online students?
- Do sleep behaviours and psychological distress jointly influence academic performance?
If you wish, you may abbreviate research questions 1, 2 and 3 to RQ1, RQ2 and RQ3, respectively, in your assessment. Use the following approaches to address the three research questions.Research Question 1.
Use Bayesian correlations to test the hypotheses that overall course mark (aggregate mark of all assessment items) is correlated positively with the total resilience score and that the overall course mark is correlated positively with the total psychological distress score.Research Question 2.
Perform a series of Bayesian t-tests with an appropriate correction for multiple comparisons (e.g., null control) to exhaustively compare the Exam assessment item score between Callaghan, Ourimbah and Online students to determine if there is any difference in Exam scores between the campuses/online.Research Question 3.
Using the Psychological Distress scores, categorise people as high or low distress individuals using a score of >=30 as the cutoff for high distress individuals. The newly created "Distress" variable with two levels (Low, High) will be the first factor in this analysis. The second factor will be Sleep Quality. This was measured with four levels (Very bad, Fairly bad, Fairly good, Very good) but for your analysis create a new sleep quality variable with two levels: combine Very bad and Fairly bad into a "Bad" level, and combine Very good and Fairly good into a "Good" level. Using the appropriate factorial Bayesian ANOVA, examine the effect of Distress and Sleep Quality on the overall course mark.How to write your assignment.
Your assignment should read like sections of a lab report. Use the following headings in your assignment and include the following information in each section:Data Analysis.
For each research question:
- Briefly describe your data preparation. This includes how many data points were missing and what you did with missing data.
- Briefly describe the Bayesian analysis you conducted. This includes the type and details of the statistical test, directional hypotheses of the test if relevant, prior distribution, and corrections for multiple comparisons if relevant.
Note: This section is like the Data Analysis section of a Method. It does not include the outcome of statistical tests.Results.
For each research question:
- Re-state the research question, as you would in a lab report.
- Visualise your data using appropriate descriptive plot/s. Include measures of uncertainty where relevant. Briefly describe the trends in the plot/s.
- Describe the outcome of appropriate assumption checks and make appropriate adjustments as required.
- Report the outcome of your Bayesian analysis and post-hoc tests (if relevant) using appropriate language and supporting statistics. Determine the robustness of the analysis where relevant (i.e., robustness check).
- Briefly draw a conclusion from each of your analyses that refer back to each of the relevant research questions. This should be about one paragraph in total.