Skip to main content
. 2015 Jul 30;2015:819183. doi: 10.1155/2015/819183

Table 2.

Summary of key statistical results.

β in final model 95% CI p value
Δ perceived stress
 Yoga classes attended −1.053 ± 0.383 [−1.850, −0.341] p = 0.010
 Self-practice 3-4 times per week −4.650 ± 2.293 [−9.114, 0.064] p = 0.052()
 Self-practice 5 or more times per week −5.329 ± 2.408 [−10.147, −0.636] p = 0.032
 Psychological distress at T1 −0.450 ± 0.169 [−0.777, −0.109] p = 0.013
Δ psychological distress
 Perceived stress at T1 −0.584 ± 0.253 [−1.176, −0.233] p = 0.017
Δ positive affect
 Positive affect at T1 −0.609 ± 0.139 [−0.874, −0.33] p = 0.001
Δ negative affect
 Yoga classes attended −0.914 ± 0.473 [−1.666, 0.217] p = 0.074()
 Self-practice 5 or more times per week −4.863 ± 2.187 [−9.128, −0.441] p = 0.033
 Gender 9.331 ± 3.672 [1.945, 16.639] p = 0.018
 Educational qualifications achieved −2.913 ± 1.042 [−4.947, −0.825] p = 0.019
 Negative affect at T1 −0.563 ± 0.186 [−0.931, −0.187] p = 0.006

First column shows, in italics, dependent values for each regression model. Beta-values (±bootstrapped standard errors) are shown, together with 95% confidence intervals (CIs) and bootstrapped p values. Only regressors significant at p < 0.05 and () p < 0.1 are presented here (see Supplemental Table 1 for full results of regression analyses). Self-practice regressors are coded against a reference of no yoga self-practise. Note that, due to regression-to-the-mean phenomena, any relationship specifically between baseline psychometrics and key dependent variables should be interpreted with care (see also main text).