SPSS 6: Linear Regression

SPSS 6: Linear Regression

1. -Linear -homoscedastic -little outliers

2. Analyze > Regression > Linear. predictor variables to independent panel and predicted to the dependent. click statistics

   • -estimates -confidence intervals -model fit -descriptives -OK

3. OUTPUT. correlation table = laid out differently to bivariate correlations.

4. PREDICTOR/NDEPENDENT = in the table, CRITERION/DEPENDENT = underneath the table

   • R = correlation between IV and DV, does not include the sign or direction of the correlatinon, R2 = variance in the DV explained by the IV
   • ANOVA = test statistics e.g. whether it is a significant predictor.
   • COEFFICIENTS: constant B = intercept a. Attractiveness/ Predictor B = slope b. report the standardized beta. The t value = if predictor is sig

5. COEFFICIENTS = the slope of the line and the intercept. B0 e.g. if 1.5, when attractive rating is 0 the model predicts you've had 1.5 dates. B1= slope of the regression line.

   • B1 = represents the change in the outcome associated with unit change in predictor. e.g. 0.951. if attractiveness =1, .971 x more dates
     • Y = 1.588 + 0.951X
       • = the line of best fit on a scatter graph.
     • F(1,18) = 63.252, p<.001

Images and files in the map