- Paste Required Information For Example Scatter Plot Model Summary And Anova Table From Sas Output Onto The Template 1 (116.08 KiB) Viewed 92 times
- Paste Required Information For Example Scatter Plot Model Summary And Anova Table From Sas Output Onto The Template 2 (131.81 KiB) Viewed 92 times
https://drive.google.com/drive/folders/ ... sp=sharing
paste required information (for example: Scatter Plot, Model Summary and ANOVA table from SAS output) onto the template. Dataset A lecturer would like to study the relationship among the overall score of Integrated Study, exam scores of two Research Methods subjects and exam score of other two subjects. Thirty students are randomly selected and their academic performance is recorded. The variables are listed as follows. Variable name Variable description RM1 (x) Exam score of Foundation Research Method subject RM2 (x2) Exam score of Advanced Research Method subject PSY (x3) Exam score of Psychology subject ENG (x4) Exam score of English subject INT (y) Overall score of Integrated Study subject Objectives Complete the following tasks in this project using the given dataset. (1) Perform exploratory data analysis (a) Make four scatter plots of each independent variable against dependent variable. From the output constructed by SAS, paste the scatter plots. (b) Compute Pearson's correlation coefficients between each independent variable and dependent variable. From the output constructed by SAS, paste the results. (c) Based on the scatter plots and Pearson correlation coefficients in parts (a) and (b), comment on the relationship between each independent variable and dependent variable.
(2) Perform linear regression analysis (a) Fit the full model using all independent variables. Report the coefficient of multiple determination R2 and interpret the meaning. From the output constructed by SAS, paste the model summary and ANOVA table. (b) Perform the model selection using each of the following criteria and report the "best" model chosen for each criterion. Explain your answer. You should use 40% as the level of significance. For the "best" model chosen in each criterion, paste the model summary, ANOVA table and other relevant outputs. (1) Adjusted coefficient of multiple determination Ra (ii) Forward selection SEHH2313 Project Page 3 of 4 (c) For the "best" model chosen in above part (b-ii), perform the following analysis: (i) Determine whether there is a multicollinearity problem. Explain your answer. (ii) Perform the Durbin-Watson test to assess whether the model errors have negative autocorrelation at the 5% level of significance. From the output constructed by SAS, paste the relevant output(s). (iii) If the first-order autocorrelation of errors is significant from Durbin-Watson test, then use the Cochrane-Orcutt method to remove the first-order autocorrelation of errors. And then report the corresponding final model. From the output constructed by SAS, paste the model summary and ANOVA table. Otherwise, if the first-order autocorrelation of errors is not significant, then report the final model and from the output constructed by SAS, paste the residual plot and discuss whether the residual plot shows the sign of autocorrelation.