To be considered for inclusion, the study had to examine the relationship
between at least one of the two forms of flexibility (i.e., flextime,
flexplace) and at least one directional indicator of work–family conflict
(i.e., WIF, FIW). Because our intent was to isolate and begin to deconstruct
flexibility, effect sizes based on composite measures that included
other benefits that often fall under the “family-friendly” umbrella (e.g.,
dependent care) were excluded. Also excluded were studies that isolated
part-time work/reduced workloads as well as those that investigated satisfaction
with flexibility. Studies based on self-focused management of
time were excluded. Finally, we excluded studies that investigated organizational
time demands (e.g., having to be on call after leaving the
workplace)
The study also had to include the information necessary to calculate a
correlation between the flexibility variable and work–family conflict. For
studies that met the inclusion criteria but did not report usable statistics,
we attempted to obtain relevant data by contacting the study authors.
Care was taken to identify studies based on the same data set and to
eliminate effects based on the same sample. When studies were based on
the same sample, we used the study with the largest sample size and/or with
usable data. Table 2 presents a list of studies excluded due to overlapping
samples. To enable comparisons of our studies with those of previous
meta-analyses, a list of studies included in three previous meta-analyses
based on information provided to us by the authors is presented in the
Appendix. We indicate whether or not each study was included in our
analyses and if not the reason for the exclusion. We emphasize that the
information provided in this table is not intended to be critical of the
inclusion decisions made by other researchers. Rather, its purpose is to be
transparent with regard to the decisions that form the database used in this
study. A list of the complete references for excluded articles that appear
in Table 2 and in the Appendix is available upon request.
A total of 61 independent samples from 58 articles met the inclusion
criteria. These articles are designated in the Reference section by an asterisk.
The total sample size included in the primary analyses ranged
from 12,853 to 112,834.
Genderwas coded as the percentage of male participants
from each sample. Parental status was coded as the percentage of parents
from each sample. It should be noted that the way in which these data
are presented varies somewhat from study to study. For example, some
studies report the percentage of the sample with children, some report the
percentage of the sample with children living at home, and some report
percentage of the sample with children living at home with children under
the age of 13. Marital status was coded as the percentage of married
participants in the sample. Work hours was coded as the average weekly
work hours of participants in the sample.
Meta-Analytic Effect Sizes Between Flexibility Variables and WIF
Note. IBM= International Business Machines; NSCW = National Study of the Changing
Workforce; QES = Quality of Employment Survey.