Skip to main content
. 2005 Sep 15;2(4):A04.

Table 6.

Linear Regression Analysis for Belief Subscales, Walk Missouri Campaign, St Joseph, Mo, 2003

Independent Variables Social Benefitsa Pleasure Benefitsb Health Benefitsc All Beliefsd
B (95% CI)e SE P B (95% CI) SE P B (95% CI) SE P B (95% CI) SE P
Age −0.005 (−0.01 to 0.003) .004 .20 Excludedf     Excluded     Excluded    
School Excluded     0.06 (0.003 to 0.12) .03 .04 0.08 (−0.02 to 0.18) .05 .11 Excluded    
Sex Excluded     Excluded     0.51 (0.16 to 0.87) .18 .01 1.46 (0.15 to 2.77) .67 .03
Health status 0.11 (−0.02 to 0.25) .07 .10 0.14 (0.04 to 0.25) .05 .01 0.22 (0.04 to 0.40) .09 .02 1.26 (0.61 to 1.92) .33 <.001
Told overweight −0.28 (−0.59 to 0.03) .16 .08 Excluded     Excluded     Excluded    
Advised by doctor to walk 0.02 (−0.29 to 0.32) .15 .90 Excluded     Excluded     Excluded    
Injury Excluded     Excluded     Excluded     −2.58 (−4.68 to −0.49) 1.07 .02
Walking environment 0.05 (0.01 to 0.08) .02 .01 0.05 (0.02 to 0.08) .01 .001 0.08 (0.03 to 0.12) .02 .001 0.44 (0.27 to 0.61) .09 <.001
Campaign dose exposure 0.16 (0.05 to 0.28) .06 .01 0.14 (0.04 to 0.24) .05 .006 0.11 (−0.05 to 0.28) .08 .17 0.44 (−0.16 to 1.05) .31 .15
a

Belief in social benefits = 2.73 ? (.005 × Age) + (.11 × Health status) ? (.28 × Told overweight) + (.02 × Advised by doctor) + (.05 × Walking environment) + (.16 × Campaign dose exposure). Adjusted R2 = 0.09.

b

Belief in pleasure benefits = 2.67 + (.06 × School) + (.14 × Health status) + (.05 × Walking environment) + (.14 × Campaign dose exposure). Adjusted R2 = 0.11.

c

Belief in health benefits = 3.59 + (.08 × School) + (.51 × Sex) + (.22 × Health status) + (.08 × Walking environment) + (.11 × Campaign dose exposure). Adjusted R2 = 0.09.

d

All walking beliefs = 30.88 + (1.46 × Sex) + (1.26 × Health status) ? (2.58 × Injury) + (.44 × Walking environment) + (.44 × Campaign dose exposure). Adjusted R2 = 0.18.

e

B indicates unstandardized regression coefficient; CI, confidence interval.

f

Variables were included in analyses only when they were associated with the dependent variable at the bivariate level.