| Week | Old process | Old | Calculations | Week | New process | New | Calculation | | Interpretations | | These are for the new process | | | | Using the make believe data |
| 1 | | Slope | ERROR:#DIV/0! | 13 | | Slope | ERROR:#DIV/0! | | For every increase of one unit there is a decrease of 1.12 in predicted y |
| 2 | | Intercept | ERROR:#DIV/0! | 14 | | Intercept | ERROR:#DIV/0! | | When there is an absence of x values the predicted y value is 56.0 |
| 3 | | correlation coefficient | ERROR:#DIV/0! | 15 | | correlation coefficient | ERROR:#DIV/0! | | There is a negative and weak association between the variables |
| 4 | | coefficient of variation | ERROR:#DIV/0! | 16 | | coefficient of variation | ERROR:#DIV/0! | | 11.2% of the variability present in predicted y can be explained by variability present in |
| 5 | | | | 17 | | | | | the model. Another interpretation ends with th phrase: variability present in x. |
| 6 | | | | 18 |
| 7 | | | | 19 | | | | | In this example the equation for the LSRL is : y = -1.12x + 56.0 |
| 8 | | | | 20 |
| 9 | | | | 21 |
| 10 | | | | 22 |
| 11 | | | | 23 |
| 12 | | | | 24 |
| | | | | | | | Only fill in the yellow |
| | equation should look like: y = slope (x) + intercept |
AT2000
LinearRegression2018_1__2__3_1.xlsx
