SWIN BUR * NE * SWINBURNE UNIVERSITY OF TECHNOLOGY Swinburne University of Technology - Department of Statistics, Data Science & Epidemiology STA20006: Analysis of Variance and Regression Module 2: Report writing for correlation and regression
Significant 1. Hypothesis 2. Sample 3. Describe correlation 1 4. Report significance Not significant 5. Report confidence interval Hypothesis is supported 6. Report Slope 7. Conclusion " ... As expected ... " 5. Report confidence interval 6. Conclusion 7. Power Analysis Hypothesis not supported " ... Contrary to expectations" > 80% power v " ... if there is any correlation in the population it is only weak" < 80% power " ... not sufficiently powerful and should be repeated with more participants"
Using the flow chart from the previous page, have a go at writing up correlation & regression reports based on the following scenarios. Note: sample solutions are provided on the last 2 pages. a. Suppose we investigated an extremely corrupt university, MinhTek, to investigate the relationship between bribe amount (in dollars) and student's final grade (out of 100). An investigation into 100 random students from MinhTek was conducted by the Police department. It was hypothesised that the more students spent on their bribe, the higher their final grade will be. Use the output provided below, write a report. Bribe Pearson Correlation 1 Grade Correlations Sig. (2-tailed) N 100 .756 1 Pearson Correlation Sig. (2-tailed) N Bribe .000 100 100 Grade .756 .000 100 Lower Limit Upper Limit 0.95 0.658 0.829 Coefficientsª Standardized Coefficients Unstandardized Coefficients Model 1 (Constant) Bribe B 26.985 4.590 Std. Error 1.225 .866 Beta 4.330 a. Dependent Variable: Grade t 1.633 3.662 Sig. .000 .000 b. Using the same scenario as part (a) and the output provided below, write a report. Bribe Pearson Correlation 1 -. 459 .000 Grade Correlations Sig. (2-tailed) N 100 -. 459 1 Pearson Correlation Sig. (2-tailed) N Bribe .000 100 100 Grade 100 Lower Limit Upper Limit 0.95 -0.601 -0.289 Coefficientsª Standardized Coefficients Unstandardized Coefficients Model 1 (Constant) Bribe -12.556 B 26.985 Std. Error 1.225 .866 -15.923 Beta a. Dependent Variable: Grade t 1.633 3.662 Sig. .000 .000