Skip to main content
. 2013 Sep 21;69B(1):85–94. doi: 10.1093/geronb/gbt075

Table 4.

Standardized Coefficient Between Covariates and Loneliness Scores Over Time

Covariates Intercept Slope
Age (0 = <70 or 1 = ≥70) 0.37 (p = .01) ns
Sex (1 = male, 2 = female) ns ns
Race (0 = white; 1 = other) −0.12 (p = .05) ns
Employed (0 = no; 1 = yes) ns ns
Education (at least college degree vs. less education) ns ns
Religion (Mormon or not) ns ns
Living arrangement (0 = alone; 1 = not alone) ns ns
Financial status (0 = not good to 2 = very good) ns ns
Happy marriage (1 = not happy to 7 = very happy) −0.14 (p = .02) ns
Death expected (1 = expected; 2 = unexpected) ns ns
General health (1 = poor to 7 = excellent) −0.16 (p = .01) ns
Depressive symptoms (0–15) 0.55 (p < .001) ns
Person to share thoughts with (0 = no, 1 = yes) −0.23 (p < .001) ns
Adequate opportunity to express self (0 = no, 1 = yes) −0.51 (p < .001) ns
Participation in social event (0 = none to 4 = often) −0.16 (p = .01) ns
Made new friend in group (0 = no; 1 = yes) −0.24 (p < .001) ns
Made a best friend in group (0 = no; 1 = yes) −0.17 (p = .01) ns
Study condition (support group A vs. B) ns ns
Attendance in support group (0–14 sessions) ns ns
Study site (San Francisco or Salt Lake City) ns ns

Notes. ns = not significant at p < .05.

All models were estimated using a single covariate as potential modifier of both slope and intercept of loneliness. Refer back to Model B of Figure 1.