statistics: 6.4 LINEAR REGRESSION IN JASP

how to run a simple linear regression on jasp? 7 pts

is PA/IV a significant predictor of BMI/DV:

1. open data set in jasp

2. click on regression

3. under classical click on linear regression

4. select BMI then dependent variable

5. select PA then covariates

6. under statistics select estimates, confidence intervals, and model fit

7. see the results on the right side

interpreting regression tables: model summary table? 2 pts

1. (R - H1)- a coefficient that is always positive and tells us about how strong the correlation between both variables is rather than about the direction of the correlation

2. R2 - H1)- it is the coefficient of determination for regression; regression equivalent of the r2 and tells us the proportion of variance of the dependent variable accounting for the independent variable

interpreting regression tables: ANOVA table? 6 pts

1. linear model

2. tests the regression model as a whole and indicates whether the regression model provides a better fit to the data than a model that contains no independent variables

3. F value- a tool used together with the p-value to answer the questions of whether variance between means of two populations differ significantly

4. P value- If p < .001 (sifnificant), the independent variable helps predict the dependent variable much better than simply using the average value meaning there is a statistically significant relationship between the independent variable and the dependent variable; If the p-value is not significant, the independent variable does not improve our predictions and prediction could be done using the average value of the dependent variable alone

5. regression of degrees of freedom (df)- the number of IVs in our regression model; for simple linear regression there is only one

6. residual degrees of freedom (df)- total number of observations/rows of the data set subtracted by the number of variables being estimated; for simple linear regression the number of variables is 2 (n-2)

interpreting regression tables: coefficients? 9 pts

1. tells us how the model works (anova tells us if it works or not)

2. standardized coefficient- equal the correlation R but can be positive or negative (standardized covariance)

3. unstandardixed coefficient- values in the original units of measurement

4. regression line formula- y = a + (b x X) + error

5. intercept in jasp = unstandardized H1 in the table and is the starting point of the regression line

6. slope in jasp = unstanrdized vs PA

7. T and P- if the t-test in the table is not statistically significant (p < .05) then that predictor does not add to your model

8. if coefficient is positive- then for every one unit increase in the predictor value the outcome variable increases by b points

9. if coefficient is negative- for every one unit increase in the predictor variable the independent variable decreases by b points