Content uploaded by Soomi Lee
Author content
All content in this area was uploaded by Soomi Lee on Aug 21, 2018
Content may be subject to copyright.
Better previous night sleep is associated with less next day work-to-
family conflict mediated by higher work performance among female
nursing home workers
Katie M. Lawson,PhD
a,
⁎,Soomi Lee,PhD
b
a
Department of Psychological Science, Ball State University, United States, 122 North Quad Building, Muncie, IN, 43706
b
School of Aging Studies, University of South Florida 13301 Bruce B Downs Blvd, MHC 1344 Tampa, Florida 33612-3807
abstractarticle info
Article history:
Received 22 April 2018
Received in revised form 22 June 2018
Accepted 12 July 2018
Available online xxxx
Keywords:
Work-to-family conflict
Sleep duration
Sleep quality
Daily telephone interview
Work performance
Female employees
Objectives: Cross-sectional research has found that shorter and poorer sleep are associated with lowerwork
performance and greater work-to-family conflict (WTFC). However, we know little about daily mecha-
nisms linking sleep, work performance, and WTFC. This study tested whether previous nights' sleep was
linked to next day WTFC, mediated by work performance.
Design: Daily interview methodology.
Setting: US extended-care workplaces.
Participants: One hundred seventy-one female employees with children aged 9 to 17 years.
Measurements: In telephone interviews on 8 consecutive evenings, participants reported their daily work
performance (work productivity, work quality), WTFC (e.g., “how much did things you wanted to do at
home not get done because of the demands your job put on you?”), and previous nights' sleep duration
(in hours) and sleep quality (1 = very badly, 4 = very well).
Results: Multilevel models revealed a significant association between previous night's sleep with next-day
work performance. More specifically, on days following better sleep quality than usual, participants re-
ported better work productivity than usual. Moreover, higher work productivity was associated with less
WTFC on that day. A mediation test revealed that poorer previous night's sleep quality predicted less
work productivity the next day, which, in turn, predicted more WTFC on the same day.
Conclusion: Results provide evidence for the downward spiral of resource losses starting from poor sleep.
Better quality sleep, as a replenished resource, may promote next-day productivity at work, which may
bring less interference from work to the home.
© 2018 National Sleep Foundation. Published by Elsevier Inc. All rights reserved.
Introduction
Approximately 1 in 3 employed individuals report not getting
enough sleep
1
and working more hours per week is associated with
reduced hours of sleep per night.
2
Moreover, approximately half of
US employees report that they experience non-refreshing sleep a
few nights per week or more.
3
Research has also found that work-
to-family conflict (WTFC) –which occurs when work demands inter-
fere with personal and family life –has been increasing over the
years.
4,5
Although evidence has emerged from cross-sectional
research that individuals who report poorer sleep are more likely to
report higher levels of WTFC,
6–8
less research has focused on under-
standing the short-term processes linking sleep and WTFC.
9
Cross-sectional evidence has emerged that poorer sleep has nega-
tive implications for work performance,
3,10
and that poorer work per-
formance is associated with more WTFC.
11,12
Thus, it is possible that
work performance is a mechanism underlying the established link
between sleep and WTFC.Yet, thus far, no studies have tested the ef-
fects of sleep on WTFC mediated by work performance. According to
the Work-Home Resources (W-HR) model,
9
improved personal
resources –which includes physical recovery resources such as
sleep –may improve work and home outcomes, ultimately reducing
WTFC. In contrast, an initial loss in a personal resource may result in a
downward loss spiral of resources across domains. For example,
starting the day off following nights with poorer recovery sleep
may result in lower performance at work (i.e., lower levels of work
productivity and work quality), ultimately leading to more interfer-
ence with family/personal activities after work (i.e., more WTFC).
Thus, we could expect negative effects of shorter and poorer sleep
Sleep Health xxx (2018) xxx–xxx
⁎Corresponding author. Tel.: +1 765 285 1760.
E-mail address: kmlawson4@bsu.edu (K.M. Lawson).
SLEH-00295; No of Pages 7
https://doi.org/10.1016/j.sleh.2018.07.005
2352-7218/© 2018 National Sleep Foundation. Published by Elsevier Inc. All rights reserved.
Contents lists available at ScienceDirect
Sleep Health
Journal of the National Sleep Foundation
journal homepage: sleephealthjournal.org
Please citethis article as: Lawson KM, Lee S, Better previous night sleep is associated with less nextday work-to-family conflict mediated by
higher work performance among fem..., Sleep Health (2018), https://doi.org/10.1016/j.sleh.2018.07.005
on work performance and further on WTFC, however, limited re-
search has tested short-term processes linking sleep ➔at work ➔
after work constructs.
The present study extends the research on links between sleep
and WTFC by examining daily short-term processes linking sleep to
WTFC mediated by work performance. To better understand these
associations, we draw upon a sample of employees in the ex-
tended-care industry (i.e., nursing homes). The extended-care indus-
try is an industry that is experiencing faster than average job growth
in the United States due in part to the aging population.
13
Most em-
ployees in this industry are women and also mothers.
14
Working
mothers in the extended-care industry may be particularly vulnera-
ble to sleep deficiencies and WTFC as a result of unpredictable care
demands from their patients, varying work schedules, and other
caregiving responsibilities at home.
14,15
Thus, focusing on mothers
employed in the extended-care industry may provide better insight
into whether and how insufficient and poor sleep is associated with
experiencing more negative WTFC in a particularly vulnerable, at-
risk population.
Thus far, a small but emerging line of research examining tempo-
ral associations between nightly sleep anddaily experiences suggests
that sleep quality and sleep duration are more likely to be predictors
of next-day experiences, includingWTFC. For example, Lee et al.
16
re-
ported that poorer sleep quality and shorter sleep duration predicted
more next-day WTFC in a sample of IT workers. Such pattern wasalso
observed in another study that revealed poorer sleep quality pre-
dicted lower emotional well-being and greater odds of encountering
stressors the following day.
17
Guided by these studies and also the
W-HR model
9
that suggests good sleep –as a personal recovery re-
source –may improve work performance and reduce WTFC, we ex-
pected that previous night's better sleep would predict better work
performance, and further, lower WTFC the next-day. Specifically,
the following daily associations were predicted:
Hypothesis 1: Poorer sleep quality and shorter sleep duration (than
usual) during the previous night will be associated with lower levels
of work productivity and work quality the next day.
Hypothesis 2: Lower levels of work productivity and work quality
(than usual) on a given day will be associated with experiencing
more conflicts from work-to-family on that day.
Hypothesis 3: Poorer sleep quality and shorter sleep duration (than
usual) the previous night will be indirectly associated with higher
levels of work-to-family conflict the next day, mediated by lower
levels of work productivity and work quality.
Method
Participants
Data were part of a larger study examining the impact of work
conditions on organizational outcomes and the health of employees
and their families.
18,19
For the current study, participants included
US employees in the extended care industry (i.e., nursing homes;
30 work sites located in the northeastern United States) who partic-
ipated in an 8-day telephone interview sub-study. Eligible employees
for the larger study were involved in direct patient care, typically
worked at least 22 hours/week, and did not do regular night work. El-
igible participants for the daily telephone interview sub-study in-
cluded individuals who had a child between 9 and 17 years of age
living at home (to recruit the ta rget child in a child daily teleph one in-
terview as well). A total of 182 employees completed the daily inter-
view portion of the study. The extended-care industry is comprised
mostly of female workers, and only 8 male workers completed daily
interviews; they were dropped from the analyses. In addition, 3
more individuals were removed because they did not report working
at least 1 day during data collection, resulting in a final sample size of
171 employed women.
A majority of participants were married/cohabitating (64%) and
reported having at least a high school degree: 6% reported some
high school, 30% reported graduating high school, 54% reported
some college or technical school, and 10% reported graduating from
college. On average, participants were 38.54 (SD = 6.40) years old,
worked at the company for 6.38 (SD = 5.75) years, worked 36.70
(SD = 8.09) hours per week, and reported 2.24 (SD = 1.11) children
living at home. Sixty-two percent were White, 15% were Hispanic,
13% were Black, and the rest 10% included Asian Indian, “Other”
Asian, “Other”Pacific Islander, “some other race,”or multiracial.
More than half of participants worked a standard daytime schedule
(59%), 21% worked regular evening shift schedule, 13% reported a
schedule that changes from day to day or rotating shifts, and 7% re-
ported split shifts or long/double shifts.
Procedures
The larger study included a workplace interview for employees,
conducted by trained interviewers using computer-assisted personal
interviews (CAPI). Participants were recruited through study posters
and informational material posted throughout the workplace, letters
and brochures that were sent as an insert with their paycheck, and re-
search personnel participated in workplace meetings and organized
several “meet and greet”sessions to provide information about the
study to employees.
At the end of the workplace interview, employees with a child
aged 9–17 were recruited (via computer-assisted scripts and a bro-
chure) to participate in the daily interview portion of the study.
Daily assessment methods, which require participants to report on
their experiences, moods, stressors, etc. for multiple consecutive
days, may be particularly useful methodologies to better understand
short-term processes linking sleep and WTFC.
20
This method allows
for the examination of the commonly studied between-person asso-
ciations (e.g., On average, do individuals who report better sleep
quality than others in the sample report lower WTFC on average
across days?) in addition to within-person associations (e.g., When
participants report better sleep quality than usual the previous
night, do they also report less than usual WTFC the next day?).
Data collection began with informed consent/assent procedures,
which were approved by the Institutional Review Boards of the
project's principal investigators. A series of eight, consecutive nightly
phone calls with the employee-parents were scheduled and con-
ducted by trained personnel at the University's survey research cen-
ter. This center specializes in collecting daily telephone interview
data, and all interviewers receive training by research personnel. A
total of 373 female employees were eligible for the daily interview
portion of the study (i.e., had a child age 9–17), and 182 employees
(174 female employees) chose to participate. To determine if those
who chose to participate differed from those who chose not to partic-
ipate in the daily interview portion of the study (i.e., eligible mothers
with a child age 9–17 who completed the larger part of the study, but
chose not to participate in the daily interviews), t-tests and chi-
square analyses were conducted to compare the groups on basic de-
mographic information. Results revealed that those who chose to
participate did not significantly differ from those who chose not to
participate in terms of education, age, income, number of children
in the home, tenure at work, race/ethnicity, and marital status.
Nightly interviews were conducted at the times most convenient
for participants to increase compliance and accommodate partici-
pants' busy schedules. After participants were asked to move to a
quiet and private location, employees reported on their daily experi-
ences including stressors, interactions with family members, physical
health (including previous night sleep), affect, and time use. On
2K.M. Lawson, S. Lee / Sleep Health xxx (2018) xxx–xxx
Please cite this article as: Lawson KM, Lee S, Better previous night sleep is associated with less next day work-to-family conflict mediated by
higher work performance among fem..., Sleep Health (2018), https://doi.org/10.1016/j.sleh.2018.07.005
average, the daily interviews lasted around 25 minutes. Families
received $150 for their participation in the daily interview study.
Out of the 1368 possible days (171 participants × 8 daily interview
days), a total of 1218 days (89.04%) were completed by participants.
Analyses were restricted to the 736 work days that participants
reported working. A majority of participants (n= 147, 84.48%)
reported working 3 days or more during the daily interview portion
of the study.
Measures
Previous night's sleep
To assess previous night's sleep quality, participants responded to
one item using a 4-point scale (1 = Very badly, 4 = Very well): “How
would you rate last night's sleep quality overall?”This item reflects
the questions used to assess sleep quality found in sleep assessments,
such as the Pittsburgh Sleep Quality Index.
21
To assess previous
night's sleep duration, participants reported the time they went to
bed the previous night and the time they woke up in the morning,
and sleep duration was calculated (in hours) with this information.
These sleep variables can inform how participants perceived their
last night's sleep quality and quantity. Daily self-reported sleep has
good agreement with actigraphically-measured sleep, especially
when comparing the measurement of bedtimes, wake times, and
sleep duration.
22
Following previous studies that analyzed nightly
sleep measures from daily interview data,
16,17
we entered sleep qual-
ity and sleep duration as separate variables in our models.
Work performance
On work days, participants reported both work productivity and
work quality by responding to items from the National Study of
Daily Experiences using a 4-point scale (1 = Not at all, 4 = A lot).
23
To measure work productivity, participants responded to the item:
“How much did you cut back on your normal paid work activities
today?”To measure work quality, participants responded to the
item: “How much did the quality of your work suffer today?”Re-
sponses on these items were reverse coded such that higher scores
reflect better work performance. Work productivity and work quality
were analyzed separately in analyses given that they are often de-
fined as distinct constructs,
24
and that past researchcommonly treats
them as separate constructs.
25
Work-to-family conflict
At the end of each work day, WTFC was evaluated using 5 items
from Netemeyer et al.'s
4
work-to-family conflict scale (e.g., How
much did things you wanted to do at home not get done because of
the demands your job put on you?; How much did your time spent
at work make it difficult to fulfill your family or personal responsibil-
ities?). Items were modified in order to assess daily WTFC, the extent
that what happened at work interferes with family/personal activi-
ties after work. Participants responded using a 4-point scale (1=
Not at all, 4 = A lot). Items were averaged so that higher scoresreflect
more daily WTFC. Reliability was calculated at the between- and
within-person levels.
26
Both were found to be adequate (between-
person α= .84, within-person α= .73).
Covariates
Sociodemographic, family characteristic, work characteristic, and
physicalhealth variables were included as covariates because past re-
search has found them to be related to sleep, work performance, and
WTFC.
16,27,28
Physical health symptoms were measured using a 10
modified items from Larsen and Kasimati's
29
physical symptoms
checklist. Items were summed each day to indicate the total number
of physical health symptoms experienced each day (e.g., headache,
stomach problems), and an average score was created across the 8
daily interview days. Other covariates included marital status (0=
Not married/cohabitating, 1 = Married/cohabitating), employee edu-
cation level (1 = Grade 1 through 8, 2=Some high school, 3=High
school graduate, 4=Some college or technical school, 5=College grad-
uate), employee age (in years), work hours (in hours), work schedule
(0 = Standard daytime schedule, 1=Nonstandard schedule, e.g.,vari-
able schedule, regular evening), and day in study (0 = day 1, 7=day
8). All continuous variables were centered at the grand mean.
Analytic strategy
To account for the nested structure of the data (days nested
within individuals nested within worksites), multilevel models
were conducted using SAS Proc Mixed (Version 9.3). Prior to analy-
ses, ICCs were conducted to examine the variance at the different
levels. Because there were relatively small (and non-significant) var-
iance at the worksite level compared to the variances at the daily and
person levels, we usedtwo-level models to simplify the model struc-
ture. Covariates were included in all models.
For each model, both the between- and within-person effects
were examined. In order to obtain the between-person estimates,
we averaged individuals' scores across all days of the study, then cen-
tered the variable at the sample mean. For example, positive values of
between-person sleep duration indicated sleeping more than others
in the sample. These between-person variables were entered as
Level 2 predictors. To obtain within-person estimates, we subtracted
the person mean from each daily score. Thus, positive values of
within-person sleep duration, for example, indicated days with
more sleep than the person's average across days. These within-per-
son variables were entered as Level 1 predictors. An example simpli-
fied model of work productivity predicted by previous night's sleep
quality (without covariates) follows:
Next Day Work Productivitydi ¼β0iþβ1iWP Prior Night Sleep QualityðÞ
d−1i
þediβ0i¼γ00 þγ01 BP Average Sleep QualityðÞ
i
þμ0iβ1i¼γ10 þμ1i
Because sleep quality reported on a given day was about last
night's sleep, the day of sleep quality was d-1 in relation to daily
work productivity. Here, β
0i
represents person i's intercept, which is
the function of the sample mean (γ
00
, intercept), the effect of the be-
tween-person (BP) level sleep quality (γ
01
), and random deviations
of person i's mean from the sample mean (μ
0i
). β
1i
represents the ef-
fect of within-person (WP) sleep quality, whether prior night sleep
quality predicts the next-day work productivity. Residual error, e
di
,
denotes random variation of person ion the d
th
day from person i's
mean. Similarly, the associations of work performance variables
with WTFC were tested at the BP and WP levels. For example, the
WP effect of work productivity on WTFC evaluates how todays' pro-
ductivity at work is associated with experiencing WTFC after work.
Several models were con ducted to address the research questions.
First, the link between previous nights' sleep and next-day work per-
formance was examined (Path A in Fig. 1; two models were con-
ducted to account for the two mediators –work productivity and
work quality). Results for the first step can be viewed in Table 2
(Model 1 assesses sleep ➔work productivity; Model 2 assesses
sleep ➔work quality). Second, the link between work performance
and WTFC after work was examined (Path B in Fig. 1), while control-
ling for previous night's sleep. Results for the second step can be
viewed in Table 2; Model 3 assesses work productivity ➔WTFC and
Model 4 assesses work quality ➔WTFC. To calculate effect sizes, un-
standardized gamma coefficients (γ) were converted into standard-
ized regression weights (β). Gamma coefficients were multiplied by
3K.M. Lawson, S. Lee / Sleep Health xxx (2018) xxx–xxx
Please citethis article as: Lawson KM, Lee S, Better previous night sleep is associated with less nextday work-to-family conflict mediated by
higher work performance among fem..., Sleep Health (2018), https://doi.org/10.1016/j.sleh.2018.07.005
the standard deviation of the predictor, and that product was divided
by the standard deviation of the outcome variable.
Based on the recommendations of Krull and MacKinnon,
30
the
estimates obtained in Paths A and B were used to calculate the in-
direct effects of sleep on WTFC through work performance. More
specifically, the estimate of the associations between sleep and
work performance (Path A) were multiplied by the estimates of
the associations between work performance and WTFC (Path B).
The significance of the indirect effect was calculated using the as-
ymptotic normal distribution method, which accounts for the
non-normal distribution of the product of estimates obtained in
Paths A and B.
31
Results
Descriptive results
Descriptive statistics and correlations for study variables can be
seen in Table 1. Overall, the intra-class correlations, calculated by di-
viding the between-level variance by the total variance, indicated
that sleep, work performance, and WTFC vary greatly from day to
day (ranging from .13 to .61), justifying the use of multilevel model-
ing with daily measures. Interestingly, for four out of five of the study
variables, over half of the variance occurred at the within-person
level. For example, only 36% of the variance for sleep quality occurred
at the between-person level, whereas 64% of the variance occurred at
the within-person level, supporting the use of a daily interview
methodology to examinedaily associationsbetween variables. In ad-
dition, sleep variables (quality and duration) were correlated at the
within- and between-person levels but the correlation coefficients
were not high (r= .43 to 45, Pb.001), meaning that they are inde-
pendent constructs. Similarly, work performance variables
(productivity and quality) were also moderately correlated (r=.24
to 43, Pb.001).
Hypothesis 1: Sleep and performance at work
Table 2 shows results of multilevel modeling examining sleep dura-
tion and quality as predictors of work productivity (Model 1) and work
quality (Model 2). In Model 1, sleep duration was not associated with
work productivity at either the BP or WP level. Sleep quality, however,
was significantly associated with work productivity at the WP level (ef-
fect size = .14). On days following nights with better sleep quality than
usual, participants reported greater work productivity than usual, after
adjusting for the BP association (not significant). The effect of WP sleep
quality on work productivity was 12 times greater than the effect of
work hours. In Model 2, sleep duration and sleep quality were not sig-
nificantly associated with work quality at the BP or WP level.
Hypothesis 2: Performance at work and work-to-family conflict
Next, we examined whether work performance was associated
with WTFC, after controlling for previous night's sleep. Model 3 in
Table 2 shows that work productivity was significantly associated
with WTFC at both BP (effect size = .22) and WP levels (effect
size = .13). At the BP level, participants who reported higher work
productivity, on average, reported less WTFC across days. At the
WP level, higher work productivity than usual predicted less
WTFC than usual. The effect of WP work productivity on WTFC
was five and a half times greater than the effect of work hours on
WTFC.
Furthermore, better work quality was associated with less WTFC
at the BP level (effect size = .18), but not at the WP level (Model 4,
Fig. 1. Conceptual model of the interconnections between previous nights' sleep, workplace performance, and work-to-family conflict.
Table 1
Means (standard deviations) and correlations for study variables (N = 171 participants, N = 736 work days)
Mean (SD) 12345
Sleep
1. Sleep Quality (1 to 4:higher) 2.91 (.83) .36 .43*** .09* .02 −.20***
2. Sleep Duration (in hours) 6.16 (1.58) .45*** .35 .06 .05 −.09*
Work Performance
3. Productivity (1 to 4:higher) 3.71 (.72) .05 .01 .13 .24*** −.20***
4. Quality (1 to 4:higher) 3.64 (.75) .01 .01 .43*** .30 −.20***
5. Work-to-Family Conflict (1 to 4:higher) 1.83 (.94) −.26*** −.06 −.28*** −.26*** .61
Note. * Pb.05, ** P b.01, *** Pb.001.Descriptive statistics reported for workdays only. The numbers 1 to 5 correspond to the study variableslisted in the left-hand column.Diagonals
(bold) show intra-class correlations (ICC= between-personlevel variance/total variance) of the variable. Numbers below thediagonal represent between-personlevel correlations
N = 171) and numbers above the diagonal indicate within-person level correlations (N = 736).
4K.M. Lawson, S. Lee / Sleep Health xxx (2018) xxx–xxx
Please cite this article as: Lawson KM, Lee S, Better previous night sleep is associated with less next day work-to-family conflict mediated by
higher work performance among fem..., Sleep Health (2018), https://doi.org/10.1016/j.sleh.2018.07.005
Table 2). These effects occurred while controlling for sleep duration
and sleep quality both at the BP and WP levels.
Hypothesis 3: The indirect effects of sleep on work-to-family conflict
The indirect effect of sleep on WTFC was tested for the pathway in
which both the A and B Paths were significant at the WP level. The
pathway was: sleep quality →work productivity (Path A) and work
productivity →WTFC (Path B). There was a significant indirect effect
of previous night sleep quality on the next-day WTFC, through work
productivity, β=−.0132, CI [−.0261 to −.0003]. After adjusting for
cross-day average (BP) associations, better sleep quality the previous
night was associated with greater work productivity the next day,
which, in turn, was associated with less WTFC.
Supplementary analyses
We further conducted supplementary analyses to test alterna-
tive WP temporal directions. First, we examined whether today's
work productivity →WTFC →tonight's sleep. Today's work pro-
ductivity was associated with WTFC on the same day (β=−.11,
SE = .04, Pb.01), however, WP WTFC was not significantly associ-
ated with sleep quality (β=.03,SE =.05,PN.10) or sleep dura-
tion that night (β=−.02, SE =.12,PN.10).
We also examined whether today's WTFC ➔tonight's sleep ➔
next day work productivity. WP WTFC was not significantly associ-
ated with sleep quality (β= .03, SE =.05,PN.10) or sleep duration
that night (β=−.02, SE =.12,PN.10). Previous night WP sleep
quality did, however, significantly predict productivity at work the
next day (β=.12,SE =.04,Pb.01), but WP sleep duration did not
predict productivity at work the next day (β= .01, SE =.02,PN.10).
Discussion
Guided by the Work-Home Resources Model,
9
the present study
examined a short-term process linking sleep and WTFC among
women working in the extended care industry –a growing occupa-
tional field with unpredictable care demands and varying work
schedules that have implications for sleep and WTFC.
13,14
The results
found a daily level indirect effect of sleep on WTFC mediated by work
performance: poorer sleep quality the previous night was associated
with greater WTFC the next day, through less productivity at work.
The results provide insight into the day-to-day mechanism underly-
ing work-family connections
9
–by identifying previous night's
sleep as the antecedent of next-day performance at work and WTFC.
The present study provides evidence of a resource loss spiral –that
the lack of recovery sleep during the previous night may have a
“ripple”effect for both work and family outcomes the next day.
Sleep, work performance, and work-to-family conflict
The results of the present study support the Work-Home
Resources model that describes when the initial loss of a personal
resource “induces further loss because there are fewer personal
resources available to effectively deal with the chronic demands or
to collect contextual resources”(p. 551).
9
First, sleep was found to
be associated with poorer work performance at the daily level.
More specifically, when individuals reported poorer sleep quality
the previous night, they also reported being less productive at work
the next day. Unlike past cross-sectional studies,
3,10
the present
study did not find between-person associations of sleep with work-
place performance. Our analyses based on multiple days of interview
data were able to separate variance in sleep and work performance
variables into between- and within-person levels, and by doing so,
we found that the associations were mostly at the within-person
Table 2
Sleep duration, sleep quality, work performance, and work-to-family conflict
Research Question 1 Research Question 2
Model 1
(Path A)
Model 2
(Path A)
Model 3
(Path B)
Model 4
(Path B)
Work productivity
(1 to 4:higher)
Work quality
(1 to 4:higher)
Work-to-family conflict
(1 to 4:higher)
Work-to-family conflict
(1 to 4:higher)
Fixed Effects
Intercept 3.69 (.07) *** 3.54 (.08) *** 1.84 (.10) *** 1.81 (.10) ***
Covariates
Marital Status
1
−.06 (.07) −.03 (.08) −.00 (.11) .02 (.11)
Age (in years) −.01 (.01) −.001 (.01) −.01 (.01) −.01 (.01)
Work Hours .01 (.004) .01 (.005) * −.02 (.01) ** −.02 (.01) *
Education
2
.02 (.05) −.12 (.06) * .15 (.08) * .12 (.08)
Day in Study (0 to 7) .02 (.01) .04 (.01) *** −.03 (.01) ** −.03 (.01) **
Work Schedule
3
−.10 (.07) −.09 (.08) .36 (.11) ** .36 (.11) **
Physical Symptoms (number) −.03 (.02) −.09 .03 ** .10 (.04) * .09 (.04) *
Main Variables
BP Sleep Duration (in hours) .06 (.03) .03 (.04) −.01 (.05) −.06 (.05)
WP Sleep Duration (in hours) −.01 (.02) .04 (.02) −.03 (.02) −.03 (.02)
BP Sleep Quality (1 to 4:higher) −.04 (.06) −.10 (.07) −.22 (.10) * −.23 (.10) *
WP Sleep Quality (1 to 4:higher) .12 (.05) * .05 (.04) −.06 (.04) −.02 (.02)
BP Productivity (1 to 4:higher) −.35 (.11) **
WP Productivity (1 to 4:higher) −.11 (.04) **
BP Work Quality (1 to 4:higher) −.25 (.10) *
WP Work Quality (1 to 4:higher) −.07 (.04)
Random Effects
Intercept .05 (.02) ** .14 (.03) *** .37 (.05) *** .37 (.05) ***
Residual .45 (.03) *** .39 (.02) *** .34 (.02) *** .35 (.02) ***
Note. Betasand standard errors reported. BP= between-person. WP = within-person. The following associations were examined in each model: Model 1: Sleep ➔work produc-
tivity; Model 2: Sleep ➔work quality. Model 3: Work productivity ➔WTFC. Model 4: Work Quality ➔WTFC. Covariates:
1
Marital status was coded as 0 = Not married, 1 = Co-
habiting/Married.
2
Education level was coded as 1 = Gra de 1 through 8, 2 = Some high school, 3 = High school graduate, 4 = Some college or technical school, and 5 =
College graduate.
3
Work schedule was coded as 0 = Standard daytime schedule, 1 = Nonstandard schedule. Continuous variables were centered at sample mean.*pb.05, **pb
.01, ***pb.001.
5K.M. Lawson, S. Lee / Sleep Health xxx (2018) xxx–xxx
Please citethis article as: Lawson KM, Lee S, Better previous night sleep is associated with less nextday work-to-family conflict mediated by
higher work performance among fem..., Sleep Health (2018), https://doi.org/10.1016/j.sleh.2018.07.005
level. This may suggest that person-to-person differences in the
reports of sleep and work performance were not associated on
average, such that employees who reported poorersleep than others
did not report lower work performance than others. However, day-
to-day variations in these variables helped us tease apart the tempo-
ral association between sleep and work performance; better previous
night's sleep quality than a person's usual was associated with higher
next day's work performance than the person's usual. The findings
shed light on the importance of considering multiple levels of analy-
ses when examining the inter-relations among resources across work
and home domains.
32
Second, work productivity was found to predict WTFC at both
the between- and within-person levels. Similar to past research
examining associations at the between-person level,
11,12
individ-
uals who reported being more productive at work were less likely
to report WTFC. Our study adds to the literature by demonstrating
within-person daily association, such that on days when individ-
uals reported being more productive than usual, they also re-
ported less conflicts from work to the family domain than usual.
The consistent result at the within-person level provides stronger
support for the association between work productivity and WTFC
than the between-person level result alone. For example, a be-
tween-person effect could be the result of third variable con-
founds, such as personality; optimistic individuals may report
higher work productivity and less WTFC than pessimistic individ-
uals. By examining within-person associations, we can reduce
concerns about some potential confounds because we are compar-
ing individuals to themselves –whether employees experienced
more WTFC than their usual on days when they reported more
work productivity than their usual. The strong associations
between work productivity and WTFC at both between- and
within-person levels highlight the importance of work experi-
ences for the level of interference from work to the home, which
may be a particularly serious issue for working mothers who
have family responsibilities.
Finally, the present study further advances the understanding
of the short-term mediation pathways linking sleep to WTFC. Ev-
idence emerged that poorer sleep quality than usual the previous
night was indirectly associated with higher WTFC, through lower
productivity at work. Although the present research focused on
day-to-day linkages from prior night's sleep to the next-day
WTFC, it can also be viewed in light of reciprocal processes be-
tween work and family domains. Interestingly, though, addi-
tional analyses did not find WTFC to predict sleep quality or
sleep duration that night. Past research, however, has found
that daily WTFC has implications for the time it takes individuals
to fall asleep at night.
16
Therefore, it is possible that there could
beacontinued,cyclicalspiralloss,beyondtheeffectsseenin
the current study.
Limitations and future research directions
Although the present study has several strengths, including the
use of daily telephone interview data, there are limitations that
provide direction for future research. First, the present study fo-
cused on working mothers in the extended care industry to exam-
ine how their nightly sleep –which may be particularly vulnerable
in this work context –is associated with daily work-family experi-
ences. Future research is needed that includes both men and
women in other industries, particularly given that the Work-
Home Resources model argues that WTFC is less likely if a person
has more contextual resources.
9
For example, non-extended care
occupationsmayhaveresourcessuchasschedulecontroland
the ability to telecommute that may ultimately have implications
for work-family linkages. Second, the present study utilized self-
reported sleep measures. Future research is needed with addi-
tional sleep measures, such as actigraphy measures, to account
for mono-method bias.
Conclusion
The connections between work and family are becoming more
important in today's society, given recent societal-level changes like
increases in the number of dual-earner couples and single-parent
families.
5
In particular, given recent research indicating that even
relatively minor daily events can have long-term implications for in-
dividuals' mental and physical health,
33,34
it is imperative to under-
stand the daily connections between sleep, work performance, and
WTFC. This study contributes to the sleep health literature by demon-
strating that poor sleep on a given night may lead to the loss spiral of
resources across work and family domains the next day.
9
Having
good quality sleep is important for being productive at work, which
may have implications for reducing the daily levels of WTFC experi-
enced by female nursing-home workers.
Disclosure
The author has neither conflict of interest or disclosures necessary
to report.
Grant support
This research was conducted as part of the Work, Family
and Health Network (www.WorkFamilyHealthNetwork.org),
which is funded by a cooperative agreement through the
National Institutes of Health and the Centers for Disease
Control and Prevention: Eunice Kennedy Shriver National
Institute of Child Health and Human Development (Grant #
U01HD051217, U01HD051218, U01HD051256, U01HD051276),
National Institute on Aging (Grant # U01AG027669), Office of
Behavioral and Social Sciences Research, and National Institute
for Occupational Safety and Health (Grant # U01OH008788,
U01HD059773). Grants from the National Heart, Lung, and
Blood Institute (Grant #R01HL107240), William T. Grant Foun-
dation, Alfred P. Sloan Foundation, and the Administration for
Children and Families have provided additional funding.
References
1. Centers for Disease Control and Prevention. 1 in 3 adults don't get enough sleep
[press release]. Retrieved from https ://www.cdc.gov/media/releases/2016 /
p0215-enough-sleep.html;2016.
2. Basner M, Fomberstein KM, Razavi FM, Banks S, William JH, et al. American time
use survey: sleep time and its relationship to waking activities. Sleep. 2007;30
(9):1085–1094.
3. Swanson LM, Arnedt JT, Rosekind MR, Belenky G, Balkin TJ, Drake C. Sleep disor-
ders and work performance: findings from the 2008 National Sleep Foundation
sleep in America poll. J Sleep Res. 2011;20:487–494. https://doi.org/10.1111/j.
1364-2869.2010.00890.x.
4. Netemeyer RG, Boles JS, McMurrian R. Development and validation of work-fam-
ily conflict and family-work conflict scales. J Applied Psychol. 1996;81:4 00–410.
5. Nomaguchi KM. Changes in work-family conflict among employed parents be-
tween 1977 and 1997. JMarriageFam. 2009;71(1 ):15–32. https://doi.org/10.
111/j.1741–3737.2008.00577.x.
6. Buxton OM, Lee S, Beverly C, Berkman L, Moen P, et al. Work-family conflict and
employee sleep: evidence from IT workers in the work, Family and Health
Study. Sleep. 2016;39(10):1871–1882. https://doi.org/10.5665/sleep.6172.
7. Crain TC, Hammer LB, Bodner T, Kossek EE, Moen P, et al. Work-family conflict,
family-suppor tive supervisor beha viors (FSSB), and slee p outcomes. JOccup
Health Psychol.2014;19(2):155–167. https://doi.org/10.1037/a0036010.
8. Jacobsen HB, Reme SE, Sembajwe G, Hopcia K, Stoddard AM, et al. Work-family
conflict, psychological distress, and sleep deficiency among patient care workers.
Work Health Saf. 2014;62:282–291.
9. ten Brummelhuis LLT, BakkerAA. A resource perspective on the work-home inter-
face: the work-home resources model. Am Psychol. 2012;67(7):545–556. https://
doi.org/10.1037/a0027974.
6K.M. Lawson, S. Lee / Sleep Health xxx (2018) xxx–xxx
Please cite this article as: Lawson KM, Lee S, Better previous night sleep is associated with less next day work-to-family conflict mediated by
higher work performance among fem..., Sleep Health (2018), https://doi.org/10.1016/j.sleh.2018.07.005
10. Doi Y, Minowa M, Tango T. Impact and correlates of poor sleep quality in Japanese
white-collar employees. Sleep. 2003;26(4):467–471.
11. Allen TD, Herst DEL, Bruck CS, Sutton M. Consequences associated with work-to-
family conflict: a review and agenda for future research. J Occup Health Psychol.
2000;5(2):278–308. https://doi.org/10.1037//1076-8998.5.2.278.
12. Amstad FT, Meier LL, Fasel U, Elfering A, Semmer NK. A meta-analysis of work-
family conflict and various outcomes with a special emphasis on cross-domain
versus matching-domain relations. J Occup Health Psychol. 2011;16(2):151–169.
https://doi.org/10.1037/a0022170.
13. Bureau of Labor Statistic s, U.S. Department of Labor. Occupational Outlook
Handbook. Nursing Assistants and Orderlies; 2017 https://www.bls.gov/ooh/
healthcare/nursing-assistants.htm.
14. Lee S, Davis KD, McHale SM, Kelly E, Kossek E, Crouter AC. When mothers work
matters for youths' daily time use: implications of evening and weekend shifts. J
Child Fam Stud. 2017;26(8):2077–2089. https://doi.org/10.1007/s10826-017-0731-7.
15. DePasquale N, Davis KD, Zarit SH, Moen P, Hammer LB, Almeida DM. Combining
formal and informal caregiving roles: the psychosocial implications of double-
and triple-duty care. J Gerontol B Psychol Sci So c Sci. 2016;71(2):20 1–211.
https://doi.org/10.1093/geronb/gbu139.
16. Lee S, Crain TL, McHale SM, Almeida DM, Buxton OM. Daily antecedents and con-
sequences of nightly sleep. J Sleep Res. 2016;26(4):498–509. https://doi.org/10.
1111/jsr.12488.
17. Sin NL, Almeida DM, Crain TL, Kossek EE, Berkman LF, Buxton OM. Bidirectional,
temporal associations of sleep with positive events, affect, and stressors in daily
life across a week. Ann Behav Med. 2017;51(3):402–415. https://doi.org/10.
1007/s12160-016-9864-y.
18. Bray JW, Kelly EL, Hammer LB, Almeida DM, Dearing JW, King RB, Buxton OM. An
Integrative, Multilevel, and Transdisciplinary Research Approach to Challenges of
Work, Family, and Health. RTI Press Publication No. MR-0024–1301 . Research
Triangle Park, NC: RTI Press ; 2013. https://doi.org/10.3768/rtipress.2013.mr.
0024.1303.
19. King RB, Karuntzos G, Casper LM, Moen P, Davis KD, Berkman L, Durham M, et al.
Work-Family Balance Issues and Work-Leave Policies. In: Gatchel RJ, Schultz IZ,
editors. Handbook of Occupational Health and Wellness. New York, NY: Springer;
2012.
20. Almeida DM. Using Daily Diaries to Assess Temporal Friction between Work and
Family. In: Crouter AC, Booth A, editors. Work–Family Challenges for Low Income
Parents and their Children. Hillsdale, NJ: Erlbaum; 2004. p. 127–136.
21. Buysse DJ, Reynolds CF, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh sleep
quality index: a new instrument for psychiatric practice and research. Psychiatry
Res. 1988;28:19–213.
22. Lockley SW, Skene DJ, Arendt J. Comparison between subjective and actigraphic
measurement of sleep and sleep rhythms. J Sleep Res. 2002;8(3):175–183.
https://doi.org/10.1046/j.1365-2869.1999.00155.x.
23. Almeida DM. National survey of midlife development in the United States (MIDUS I)
National Study of Daily Experiences (NSDE), 1996–1997. ICPSR03725-v4. Ann Arbor,
MI: Inter-university Consortium for Political and Social Research [distributor]; 2015.
24. Singh J. Performance productivity and quality of frontline employees in service or-
ganizations. JMark. 2000;64(2):15–34.
25. MaroofiF,Navidinya F. The measurement of job performance and its impact on ef-
fectiveness. Int J Bus Perform Manag. 2011;12(3):217–227. https://doi.org/10.
1504/IJBPM.2011.039887.
26. CranfordJA, Shrout PE, Iida M,Rafaeli E, Yip T, BolgerN. A procedure for evaluating
sensitivity to within-person change: can mood measures in diary studies detect
change reliably? Pers Soc Psychol Bull. 2006;32(7):917–929.
27. Byron K. A meta-analytic review of work- family conflict and its antecedents. J
Vocat Behav. 2005;67:169–198. https://doi.org/10.1016/j.jvb.2004.08.009.
28. Folkard S, Tucker P. Shift work, safety and productivity. Occup Med. 200 3;53:
95–101. https://doi.org/10.1093/occmed/kqg047.
29. Larsen RJ, Kasimatis M. Day-to-day physical symptoms: individual differences in
the occurrence, duration, and emotional concomitants of minor daily illnesses. J
Pers. 1991;59:387–423. https://doi.org/10.1111/j.1467-6494.1991.tb00254.x.
30. Krull JL, MacKinnon DP. Multilevel modeling of individual and group level medi-
ated effects. Multivariate Behav Res. 2001;36(2):249–277.
31. Tofighi D, MacKinnon DP. RMediation: an R package for mediation analysis confi-
dence intervals. Behav Res Methods. 2011;43:692–700.
32. Casper WJ, Eby LT , Bordeaux C, Lockwood A, Lambert D. A review of research
methods in IO/O B work-family research. J Appl Pyschol. 2007;92(1):28–43.
https://doi.org/10.1037/0021-9010.92.1.28.
33. CharlesST, Piazza JR, MogleJ, Sliwinski MJ, Almeida DM. Thewear and tear of daily
stressors on mental health. Psychol Sci. 2013;24(5):733–741. https://doi.org/10.
1177/0956797612462222.
34. Piazza JR, Charles ST, Sliwinski M, Mogle J, Almeida DM. Affective reactivity to
daily stressors and long-term risk of reporting a c hronic physical health condi-
tion. Ann Behav Med. 2013;45(1):110–120. https://doi.org/10.1007/s12160-
012-9423-0.
7K.M. Lawson, S. Lee / Sleep Health xxx (2018) xxx–xxx
Please citethis article as: Lawson KM, Lee S, Better previous night sleep is associated with less nextday work-to-family conflict mediated by
higher work performance among fem..., Sleep Health (2018), https://doi.org/10.1016/j.sleh.2018.07.005
