|Home | About | Journals | Submit | Contact Us | Français|
The purposes of this study were to investigate the effects of maternal employment, maternal employment attitude/behavior consistency, and degree of choice and satisfaction with the employment decision on family functioning and preterm infant development and to describe changes in family functioning over time. Data were collected in the family’s home (N = 79) when the infant was 3, 9, and 18 months old. Parents in nonemployed-mother families were more satisfied with their families at 18 months than parents in employed-mother families. Decreases in family cohesion and/or adaptability from 9 to 18 months were seen for fathers in employed-mother families, for mothers in nonemployed-mother families, and for mothers in families where the mother’s employment attitudes and behaviors were consistent. Degree of choice was positively related to the child’s mental development, mother’s perception of family cohesion, and mother’s and father’s satisfaction with family.
Employment of mothers with young children is increasingly common, and issues surrounding maternal employment have surfaced in many areas, including the political arena. Concem for the well-being of the child and the marital relationship has produced a significant amount of research on the impact of maternal employment. Few, if any, negative effects have been found for the marital dyad or the healthy child (Hoffman, 1989). However, children who are born early are at risk for developmental delay and may be more sensitive to stressors in their environments (Tobey & Schraeder, 1990). In addition, premature birth of an infant is generally a stressful event for the parents and may be a source of strain in their relationship. Thus, the purposes of this study were: (a) to investigate the effects of maternal employment, employment-related attitudes, and attitude/behavior consistency on family functioning and preterm infant development when the infants were 18 months of age and (b) to describe changes in family functioning and infant development over time.
Three studies have been reported in which the effects of maternal employment on children born early were investigated. Prematurely born toddlers with employed mothers scored lower on cognitive development testing than toddlers of nonemployed mothers (Cohen, 1978). Controlling for birth weight and number of parents in the home separately did not account for these differences. However, Cohen did not control for both influences concurrently. Youngblut, Loveland-Cherry, and Horan (1991, 1993) studied correlates of maternal employment in a sample of two-parent families with preterm infants, controlling for degree of prematurity. Number of hours employed per week at 3 months was positively related to the preterm infant’s motor development at 3 months of age, but not at 9 or 12 months. Degree of choice about the mother’s employment decision at 3 months was positively related to motor development at 9 and 12 months. Mothers with consistent employment attitudes and employment status at 3 months had infants at 9 and 12 months with higher motor development. Thus, when degree of prematurity and number of parents in the home were controlled, maternal employment did not negatively affect preterm infant development.
Maternal employment attitude/behavior consistency was found to be a more important factor in the healthy child’s development than either the mother’s employment attitudes or her employment status alone. Based on mother-child observations and interviews conducted in the home at 8 months, greater maternal separation anxiety was related to greater infant proximity seeking and more infant wariness with a stranger for nonemployed mothers, but to less proximity seeking and wariness for employed mothers (Hock, 1980). Sons with insecure attachments to their mothers at 18 months had mothers with lower integration of the employee and mother roles compared to sons with secure attachments (Benn, 1986). Mothers employed by choice held more positive attitudes about their preschool children than mothers employed for financial need only (Alvarez, 1985). Higher job satisfaction for employed mothers was related to greater self-control and fewer behavioral problems in their preschool daughters (Barling & van Bart, 1984). Kindergarten children whose nonemployed mothers preferred employment consistently scored lower on a battery of achievement tests than children with nonemployed mothers who preferred not to be employed and children with employed mothers regardless of their employment preference (Farel, 1980).
To resolve some of the conflict in the findings on maternal employment, attention has turned to identifying family process variables that could mediate its effects. Youngblut et al. (1991, 1993) investigated the effect of maternal employment, employment attitudes, and employment attitude/behavior consistency on families with preterm infants at 3, 9, and 12 months of age. For the total sample, family measures at 3, 9, and 12 months were not related to number of hours employed. However, in analyses with employed-mother families alone, higher number of hours employed per week was related to greater family adaptability perceived by fathers at both 9 and 12 months. Greater satisfaction with family by fathers at 9 months was related to greater choice and satisfaction with the employment decision by mothers at 3 months; greater satisfaction with family by fathers at 12 months was related to greater maternal satisfaction with her employment decision at 3 months.
Research regarding the effects of maternal employment on the family system has been limited to the two studies by Youngblut et al. (1991, 1993) mentioned above. Researchers have investigated its effects on the marital dyad with conflicting results. Some studies found no relationship between women’s employment and their husbands’ well-being (Fendrich, 1984), depression (Ross, Mirowsky, & Huber, 1983), and marital satisfaction (Locksley, 1980; Staines, Fleck, Shepard, & O’Connor, 1978). Others found greater depression and lower self-esteem for husbands when their wives were employed (Kessler & McRae, 1982). However, the role that employment attitude/behavior consistency may play in the process was not considered.
Data for this study were part of a longitudinal study of families (N = 125) who experienced the preterm birth of an infant. The convenience sample was recruited from two Level III Neonatal Intensive Care Units (NICUs) in the Midwest. Inclusion criteria for the infants were: gestational age less than 37 weeks at birth, weight appropriate for gestational age, absence of congenital anomalies that would preclude developmental progress, and hospitalized in a Level III NICU for more than 1 week but less than 3 months. In addition, the mother was living with a male partner in the father role. The study reported here excluded mothers who identified their employment status as leave of absence at any time point. Thus, the sample for this study consisted of 79 families who were interviewed when their infants were 3, 9, and 18 months old.
Mothers’ ages ranged from 18 to 40 (M = 28.0, SD = 4.92); fathers, from 19 to 47 (M = 30.5, SD = 5.72). Most of the parents were married (96.2%), white (94.9%), and had completed high school (97.5%). The majority of fathers were employed throughout the study. One father at 3 months, 4 fathers at 9 months, and 3 fathers at 18 months were unemployed. The number of employed mothers increased steadily over the 15-month period, from 31 at 3 months, to 39 at 9 months, to 44 at 18 months. Mean number of hours employed at 3 months was 34.9 (SD = 10.25); at 9 months, 31.9 (SD = 13.97); and at 18 months, 31.4 (SD = 13.28). Thus, the average number of hours employed was fairly stable. At 18 months, mothers’ stated occupations included 30 housewives; 13 unskilled, skilled, or semiskilled positions; 6 clerical or sales positions; and 30 semiprofessional or professional positions. The occupational breakdown for fathers was 41 unskilled, skilled, or semiskilled positions and 35 semiprofessional or professional positions. At 3 months, 18 families reported incomes of less than $20,000; 39 reported incomes between $20,000 and $40,000; and 22 reported incomes of $40,000 or higher. Compared to the 3-month income category, 6 families were in lower income categories, 52 were in the same category, and 21 were in higher categories by 18 months.
There were 41 male and 38 female infants. Gestational age at birth ranged from 27 to 36.5 weeks (M = 32.7, SD = 2.33). Average weight was 1814.6 g (SD = 538.83, range = 975 to 3990 g) at birth, 8.0 kg (SD = 1.11) at 9 months, and 10.7 kg (SD = 1.35) at 18 months. Infants spent an average of 31.5 days (SD = 16.01) in the NICU. Nine infants experienced an intraventricular hemorrhage; seven were grade I and two were grade II. None of the infants had necrotizing enterocolitis; 49 (62%) were discharged from the NICU with an apnea monitor.
Mother’s employment behavior was measured at 3, 9, and 18 months. Mothers reported the number of hours they were employed per week and identified their employment status as employed, not employed, or leave of absence. Employment status groups were constituted based on the mother’s reported employment status at 18 months. Mothers who reported their employment status as leave of absence at any one of the three times were excluded from the analyses reported here.
Attitudes about maternal employment were measured by Home/Employment Orientation (HEO), degree of choice about the employment decision, and degree of satisfaction with that decision. Each was measured at all three time points. The HEO Scale (Youngblut, Loveland-Cherry, & Horan, 1990) contained 8 items that mothers rated on an 8-point Likert-type scale ranging from 1 (strongly agree) to 8 (strongly disagree). The other points on the scale were not explicitly identified with word anchors. Items were chosen to tap the mother’s perception of how her own employment would affect her husband/partner, her children, and herself. Factor analysis supported the extraction of one factor, and all items loaded strongly (≥ .50) on that factor (Youngblut, 1993). Higher summative scores represented stronger employment orientations or more positive attitudes toward employment. Coefficient alphas were .80, .85, and .82 at 3, 9, and 18 months, respectively.
Degree of choice in the employment decision and degree of satisfaction with that decision were each measured with a single item. Mothers rated how much choice they had and how satisfied they were with their decision on 10-point Likert-type scales ranging from 1 (no choice or not at all satisfied) to 10 (totally my choice or very satisfied). Construct validity of these measures was supported by positive correlations with measures of joy and contentment and by negative correlations with depression and hostility (Youngblut & Casper, 1993). Intercorrelations of choice (r = .52 to .60) and satisfaction (r = .45 to .65) across all three times indicated stability in these measures.
Mothers’ employment attitude/behavior consistency groups were created from HEO scores and employment status groups at 18 months. First, based on the theoretical midpoint of the HEO scale, women were classified as home oriented (scores ≤ 36) or as employment oriented (score > 36). Then HEO categories and employment status categories were combined to form consistency groups. Employed mothers in the employment-oriented group and nonemployed mothers in the home-oriented group were considered “consistent,” while employed mothers in the home-oriented group and nonemployed mothers in the employment-oriented group were considered “inconsistent.”
Family functioning was measured with the FACES III (Olson, Portner, & Lavee, 1985) and the Feetham Family Functioning Survey (FFFS) (Roberts & Feetham, 1982). The FACES III measured family cohesion and family adaptability. Mothers and fathers each rated the 20 items (10 for each concept) on a 5-point Likert-type scale from 1 (never) to 5 (almost always). Higher scores represented greater cohesion or adaptability. Construct validity of the scales was supported by their ability to distinguish between clinical and nonclinical families (Olson, 1986) and by significant correlations between the FACES scales and other measures of family functioning (Thomas & Barnard, 1986). Olson, Sprenkle, and Russell (1979) originally posited a curvilinear relationship of cohesion and adaptability to family functioning. While this relationship holds in problem families, a linear relationship was found for normal families (Olson, 1989). In addition, scatterplots between these two measures and variables of interest in this study did not support the curvilinear hypothesis. Therefore, simple summative scores for cohesion and adaptability were used. Olson et al. (1985) reported internal consistencies of .77 for cohesion and .62 for adaptability. In this study, internal consistencies for cohesion were .80, .79, and .85 for mothers and .80, .75, and .76 for fathers at 3, 9, and 18 months, respectively. Alphas for adaptability were .58, .74, and .77 for mothers and .67, .65, and .71 for fathers at 3, 9, and 18 months, respectively.
The FFFS used a Porter format to measure the family’s satisfaction with relationships among family members and between the family and external persons and systems. Respondents rated each of the 25 stems regarding (a) How much is there now? and (b) How much should there be? Absolute values of the difference between real (a) and ideal (b) were summed to obtain total scores. Thus, lower scores indicated greater satisfaction with family. Construct validity was supported by a strong correlation (r = −.68) between the FFFS and Pless and Satterwhite’s Family Functioning Index (Thomas & Barnard, 1986). Roberts and Feetham reported an internal consistency of .81. Coefficient alphas were .80, .80, and .81 for mothers and .80, .82, and .84 for fathers at 3, 9, and 18 months, respectively, in the current study.
Child development was measured with the Mental Developmental Index (MDI) and the Psychomotor Developmental Index (PDI) of the Bayley Scales of Infant Development (Bayley, 1969). The MDI assessed sensory-perceptual, verbal communication, and early cognitive development, and the PDI measured fine and gross motor control. Raw scores were converted to standardized scores according to Bayley’s recommendations, based on the child’s corrected age. By using corrected age, rather than chronologic age, scores were adjusted for degree of prematurity. Standardized scores ranged from 50 to 150 with a mean of 100. Bayley reported split-half reliabilities of .81 to .93 for the MDI and .68 to .92 for the PDI. In this study, interrater reliabilities ranged from 76% to 86% at 3 months, 74% to 81% at 9 months, and 70% to 73% at 18 months.
Data collection for the parent study occurred in the family’s home when the infant was 3, 9, and 18 months of age using a combination of interview, self-complete, and developmental testing. Data about the infant’s NICU stay were retrieved from the hospital record. All 79 families in this subsample provided data at all three time points. Unless otherwise noted, employment status groups and consistency groups were constituted based on the mother’s 18-month data.
Employed-mother families and nonemployed-mother families were compared on six family measures (three reported by mothers and three reported by fathers) and two child development measures with two-sample t tests. There were no significant differences on the two child development measures. (See Table 1.) Mothers who were not employed at 18 months were more satisfied with their families (had lower FFFS scores) than employed mothers. Likewise, fathers in the nonemployed-mother group were more satisfied with their families than fathers in the employed-mother group. Differences in cohesion, adaptability, MDI, and PDI were not significant. The two groups of infants did not differ on gestational age at birth, birth weight, length of NICU stay, l- and 5-minute Apgars, and proportion on apnea monitors at discharge.
Comparisons of the Mothers’ stated employment status at 3, 9, and 18 months revealed that 24 mothers experienced a change in employment status. Reasons for the changes were not elicited. Changing employment status during the 15 months of the study could confound the results. Therefore, analyses were repeated with the subsample of women (n = 55) who reported the same employment status at 3, 9, and 18 months. The results of comparisons between employment status groups were identical to those with the full sample. Mothers and fathers in the continuously nonemployed-mother group were more satisfied with their families (had lower FFFS scores) than mothers and fathers in the continuously employed-mother group. Differences in cohesion, adaptability, MDI, and PDI scores at 18 months were not significant. Infants in the continuously employed and nonemployed families did not differ significantly on neonatal variables.
Employed-mother families showed fewer changes in family and child outcomes over time than nonemployed-mother families. A series of repeated measures ANOVAs (RM-anova) were used to compare 3-, 9-, and 18-month scores separately for employed-mother families and for nonemployed-mother families (Table 2). Paired t tests were used for post-hoc comparisons when the RM-anova was significant. In 18-month employed-mother families, fathers had lower cohesion scores and infants had lower MDI scores at 18 months than at 9 months. When RM-anovas were repeated excluding families where the mother’s employment status changed during the study, only the differences for fathers’ cohesion scores remained significant. In 18-month nonemployed-mother families, mothers had lower adaptability scores at 18 months than at both 3 and 9 months. Infants with nonemployed mothers had signficantly lower MDI and PDI scores at 18 months than at 9 months. These findings remained after excluding families where the mother’s employment status changed during the study. However, mothers who were continuously not employed reported greater satisfaction with family (lower FFFS scores) at 18 months than at 9 months, a finding that was not present in the total group of nonemployed mothers.
When compared with two-sample t tests on the six family and the two child development measures obtained at 18 months (Table 1), families in the 18-month consistent maternal employment attitude/behavior group (n = 53) differed from families in the 18-month inconsistent group (n = 19) on only one comparison. Fathers in the inconsistent group were less satisfied with their families (had higher FFFS scores) than fathers in the consistent group. Mothers’ family measures, fathers’ cohesion and adaptability scores, and MDI and PDI scores were not significantly different. Infants in the consistent and inconsistent groups did not differ on neonatal variables.
Of the 79 families in the study, 37 families changed consistency group over the 15-month study period. These 37 families were excluded and analyses based on consistency group were repeated for the 42 families who did not experience a change in consistency group from 3 to 9 to 18 months. As in the full sample, fathers in the continuously inconsistent group were less satisfied with their families at 18 months than fathers in the continuously consistent group. However, when families who changed consistency group over time were excluded, inconsistent mothers also were less satisfied with their families than consistent mothers. No significant differences in infant NICU variables were found.
Contrary to expectations, families in the consistent maternal employment attitude/behavior group at 18 months experienced a decrease in family and child outcomes across time. A series of repeated measures ANOVAs (RM-anova) were used to compare 3-, 9-, and 18-month scores separately for consistent-mother families and for inconsistent-mother families (Table 2). Paired t tests were used for post-hoc comparisons when the RM-anova was significant. In the consistent group, mothers had lower adaptability scores at 18 months than at 3 and 9 months. Infants in consistent families scored lower on the MDI at 18 months than at 9 months. None of the cross-time differences were significant for families in the 18-month inconsistent group.
The pattern of significant cross-time differences in family measures for the continuously consistent families (n = 34) differed from the pattern for the full sample, although the comparisons with infant developmental scores were the same. Mothers in the continuously consistent group were more satisfied with their families at 18 months than at 3 months, although the full sample of consistent mothers reported lower adaptability scores across time. As with the full sample, infants in continuously consistent families scored lower on the MDI at 18 months than at 9 months. Again, none of the within-subject comparisons with continuously inconsistent families ( n = 8) were significant. However, the small number of families in this group dictates caution in interpreting these findings.
Two sets of hierarchical multiple regression analyses were conducted. In the first set, each of the 18-month family and child measures were regressed on number of hours employed, HEO scores, and degree of choice and satisfaction with the employment decision at 18 months. In the second set, 18-month family and child measures were regressed on consistency (consistent = 1, inconsistent = 0) and degree of choice and satisfaction with the employment decision at 18 months. Gestational age was controlled in all regressions, either by entering it first when a family measure was the dependent variable or by using corrected age MDI and PDI scores when a child development measure was the dependent variable. Mother’s education was entered first into the child development regressions as a control variable, since Alvarez (1985) found a moderating effect for mother’s education and since education is frequently found to be related to child development (Hoffman, 1989). Thus, in regressions where a family measure was the dependent variable, gestational age was entered in Stage 1 and the employment-related variables in Stage 2. In regressions where a child development measure was the dependent variable, mother’s education was entered in Stage 1 and the employment-related variables in Stage 2. Regression analyses with mother’s adaptability, father’s adaptability, father’s cohesion, and PDI as dependent variables were not significant in either set of regressions. The regression equation in the second set was not significant when MDI was the dependent variable.
The four significant regression equations in the first set are represented in Table 3. Number of hours employed was a significant, positive predictor of MDI scores, but not of mothers’ cohesion, mothers’ FFFS, or fathers’ FFFS. Home/employment orientation was a significant, negative predictor of MDI scores, but not of mothers’ cohesion, mothers’ FFFS, or fathers’ FFFS. Higher maternal satisfaction with the employment decision was related to less satisfaction with family for fathers. As degree of choice increased, mothers rated their families as more cohesive, and both mothers and fathers reported greater satisfaction with family.
The pattern of results from the second set of regressions was similar to that obtained in the first set (Table 3). In the three significant regressions, choice in the employment decision was a significant predictor of mothers’ cohesion, mothers’ FFFS, and fathers’ FFFS. Thus, as degree of choice increased, these family functioning scores also improved. Consistency group and satisfaction with the employment decision were significant predictors of the father’s FFFS, but not of the mother’s FFFS or cohesion. Being in the consistent group was associated with greater satisfaction with family for fathers, but greater maternal satisfaction with the employment decision was associated with lower satisfaction with family for fathers.
Differences in child development between employed and nonemployed mother groups were not significant. This is consistent with results from earlier waves with this sample (Youngblut et al., 1991, 1993). Although Cohen (1978) found that prematurely born toddlers of nonemployed mothers scored higher on developmental testing than toddlers of employed mothers, many studies with healthy children have found no differences between groups (Doyle, 1975; Hock, 1980; Stith & Davis, 1984). Based on the values for the infant NICU variables, our sample included few severely ill infants; therefore, it is not surprising that these results are consistent with those from studies of healthy children. For infants in employed-mother families, in nonemployed-mother families, and in consistent-mother families, the 3- and 18-month child development scores were similar , but the 9-month scores were significantly higher than the 18-month scores. This pattern is similar to the pattern of scores obtained by other researchers (Barrera, Rosenbaum, & Cunningham, 1987) and appears to be unrelated to employment or consistency status. Thus, employment status alone had no negative effect for these healthier preterm infants from 3 to 18 months of age.
Parents in the nonemployed-mother group were more satisfied with their families than parents in the employed-mother group. Again, this was consistent with an earlier wave from this sample. While differences in satisfaction were not significant at 3 and 9 months, nonemployed mothers reported more satisfaction with family at 12 months (Youngblut et al., 1993). Cohesion for fathers in employed-mother families and adaptability for mothers in nonemployed-mother families and for mothers in families where the mother’s employment attitudes and behaviors were consistent were lower at 18 months compared to earlier time points. This was not expected. Although these lower family scores may be an artifact of the data or the number of comparisons computed, it may reflect a change in family dynamics related to the age of the child. Perhaps as the child entered the age of autonomy, the family responded by changing their parenting style in light of the child’s changing developmental needs (Olson, 1989).
Choice was a significant predictor of mother’s cohesion and satisfaction with family and of father’s satisfaction with family. Degree of choice has been consistently related to family and/or child outcomes at each time point. At 3 months, choice was positively correlated with motor development in employed-mother families. For nonemployed-mother families, choice was negatively correlated with motor scores and positively correlated with cohesion (Youngblut et al., 1991). Choice at 3 months was a significant predictor of motor development at 12 months regardless of employment status. Greater choice at 3 months was related to greater satisfaction with family for fathers in the employed-mother group at 9 months. For the nonemployed-mother group, greater choice at 3 months was related to greater cohesion at 12 months for fathers, to greater satisfaction with family at 9 and 12 months for mothers, and to higher mental and motor developmental scores at 9 and 12 months (Youngblut et al., 1993). Thus, consistent with other studies, choice regarding employment status was important for family and child outcomes.
In summary, maternal employment had no negative or positive effects on child development or on family functioning in families with 18-month-old preterm infants. Degree of choice in the mother’s employment decision was positively related to family functioning and child development at 18 months as it was at 9 and 12 months. Research on maternal employment effects needs to include families with older prematurely born children and families whose children represent a broader range of health states.
The data for this study were collected as part of a larger study funded by the National Center for Nursing Research, NIH, R01-NR01390, awarded to the second and third authors.
JoANNE M. YOUNGBLUT, Frances Payme Bolton School of Nursing, Case Western Reserve University, Cleveland, OH.
CAROL J. LOVELAND-CHERRY, Division of Health Promotion and Risk Reduction, School of Nursing. The University of Michigan, Ann Arbor; MI.
MARY HORAN, Kirkhof School of Nursing, Grand Valley State University, Allendale, MI.