At the individual level of analysis, the mean d for POFS was .
Analyzing data at the group level, the number of work-life practices offered was again only systematically related to POFS and AC (with a few inversions).
Zero-order bivariate relationships were significant at the individual level between number of practices and POFS and AC, at r = .
At the individual level, the number of practices was significantly associated with POFS (Beta = .
Intangible POFS at Time 1 was significantly related to affective commitment and job search behavior measured at Time 2 (r = .
As predicted in Hypothesis 3a and shown in Table 3, correlations between the three types of support (Tangible POFS, Intangible POFS and PSFS) and affective commitment appeared to be stronger at the group compared to the individual level; rs = .
Tangible POFS at Time 1 was not related to either variable at Time 2, which corroborates our cross-sectional findings.
This pattern of results also held up for Intangible POFS and work-interferes-with-family conflict, but not job search behavior.