Skip to main content
. Author manuscript; available in PMC: 2017 Apr 1.
Published in final edited form as: Int Psychogeriatr. 2015 Nov 10;28(4):613–619. doi: 10.1017/S104161021500174X

Table 2.

Steps for building final ANCOVA model for insomnia severity.

F df p η2 Δη2
Basic ANOVA 7.1 2,132 .001*** .10
Model 1-demographic variables added .01
    Age 0.2 1,129 .69 .001
    Gender 1.0 1,129 .32 .008
    Education 0.1 1,129 .81 <.001
    Group 6.7 2,129 .002** .09
Model 2-health variables added .03
    Mini-Mental Status Exam 0.1 1,126 .77 .001
    Depressed mood (HAM-D factor) 0.0 1,126 .97 <.001
    Anxiety (HAM-D factor) 0.5 1,126 .47 .004
    Alcohol abuse in the past month (SCID-I/P) 1.8 1,126 .19 .01
    Physical illness burden (CIRS-G) 2.3 1,126 .13 .02
    Group 5.1 2,126 .008** .07
Final model .03
    Physical illness burden (CIRS-G) 2.4 1,129 .12 .02
    Alcohol abuse in the past month (SCID-I/P) 1.9 1,129 .18 .01
    Group 7.3 2,129 .001** .10

Note:

*

p <.05,

**

p <.01,

***

p <.001.

Δη2 = indicates difference compared to the basic ANOVA model, HAM-D=Hamilton Rating Score for Depression, SCID-I/P = The Clinical Interview for DSM-IV Axis I Disorders–Patient Edition, Version 2. CIRS-G = Cumulative Illness Rating Scale for Geriatrics.