Table 2.
Traditional ANOVA and GLM Models of Pause Duration in Seconds by Age in Years and Experimental Condition
Main effects only | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
ANOVA model (categorical age) | GLM model (continuous age) | |||||||||||
Summary table | Summary table | |||||||||||
Predictor | SS | df | MS | F | p | [95% CI] | SS | df | MS | F | p | [95% CI] |
Age | 123.75 | 4 | 30.94 | 1.59 | .184 | .06 [.00, .14] | 138.98 | 1 | 138.98 | 7.41 | .008 | .07 [.01, .18] |
Condition | 44.57 | 1 | 44.57 | 2.86 | .134 | .02 [.00, .11] | 48.51 | 1 | 48.51 | 2.59 | .111 | .03 [.00, .11] |
Error | 1852.83 | 95 | 19.50 | 1837.59 | 98 | 18.75 | ||||||
Parameter estimates | ||||||||||||
b | [95% CI] | SE | t | p | ||||||||
Intercept | 6.73 | [5.52, 7.94] | 0.60 | 11.07 | < .001 | |||||||
Age | 1.02 | [0.28, 1.76] | 0.37 | 2.72 | .008 | |||||||
Condition | 1.39 | [−0.32, 3.10] | 0.86 | 1.61 | .111 | |||||||
Main effects with interaction | ||||||||||||
ANOVA model (categorical age) | GLM model (continuous age) | |||||||||||
Summary table | Summary table | |||||||||||
Predictor | SS | df | MS | F | p | [95% CI] | SS | df | MS | F | p | [95% CI] |
Age | 122.82 | 4 | 30.70 | 1.54 | .197 | .06 [.00, .14] | 138.79 | 1 | 138.79 | 7.33 | .008 | .07 [.01, .18] |
Condition | 51.69 | 1 | 51.69 | 2.59 | .111 | .03 [.00, .12] | 47.60 | 1 | 47.60 | 2.51 | .116 | .03 [.00, .11] |
Interaction | 38.94 | 4 | 9.74 | 0.49 | .744 | .02 [.00, .06] | 1.21 | 1 | 1.21 | 0.06 | .801 | < .01 [.00, .04] |
Error | 1813.88 | 91 | 19.93 | 1836.38 | 97 | 18.93 | ||||||
Parameter estimates | ||||||||||||
b | [95% CI] | SE | t | p | ||||||||
Intercept | 6.74 | [0.91, 7.47] | 0.61 | 11.01 | < .001 | |||||||
Age | 0.92 | [−0.14, 1.98] | 0.53 | 1.73 | .087 | |||||||
Condition | 1.38 | [−0.35, 3.10] | 0.87 | 1.59 | .116 | |||||||
Interaction | 0.19 | [−1.30, 1.68] | 0.75 | 0.25 | .801 |
Note. F tests for model effects are based on Type III SS for all models. Age was centered by subtracting the mean age from all ages to facilitate interpretation. Condition is coded 0 = control, 1 = experimental.