Nonmaternal Care's Association With Mother's Parenting Sensitivity: A Case of Self‐Selection Bias?

• Author: Alfred DeMaris,Kei M. Nomaguchi
• DOI: http://doi.org/10.1111/jomf.12022
• Date: 01 June 2013
• Published date: 01 June 2013

KEI M. NOMAGUCHI AND ALFRED DEMARIS Bowling Green State University

Nonmaternal Care’s Association With Mother’s

Parenting Sensitivity: A Case of Self-Selection Bias?

Although attachment theory posits that the

use of nonmaternal care undermines quality

of mothers’ parenting, empirical evidence for

this link is inconclusive. Using data from the

National Institute of Child Health and Human

Development Study of Early Child Care and

Youth Development ( N=1,233), the authors

examined the associations between nonmaternal

care characteristics and maternal sensitivity

during the f‌irst 3 years of children’s lives,

with special attention to selection effects and

moderation by resource levels. Findings from

f‌ixed-effects regression models suggested that,

on average, there is little relationship between

nonmaternal care characteristics and maternal

sensitivity, once selection factors are held

constant. Some evidence of moderation effects

was found, however. Excellent-quality care is

related to more sensitivity for mothers with lower

family income. Poor-quality care is related to

lower sensitivity for single mothers, but not

partnered mothers. In sum, nonmaternal care

characteristics do not seem to have as much

inf‌luence on mothers’ parenting as attachment

theory claims.

Between 1975 and 1998, the labor force

participation rate among mothers with children

Department of Sociology, Bowling Green State University,

Bowling Green, OH 43403 (knomagu@bgsu.edu).

Key Words: child-care arrangements, f‌ixed-effects models,

marital status, parenting styles, selection effects, socioeco-

nomic status.

under age 3 increased from 34.3% to 62.2%.

Since the peak in 1998, there has been no

clear trend (U.S. Department of Labor, Bureau

of Labor Statistics, 2010). The increase in

maternal employment during the f‌irst 3 years of

children’s lives has led to scientif‌ic and public

debates about how early nonmaternal care might

inf‌luence the mother – child relationship and

the development of children (National Institute

of Child Health and Human Development

Early Child Care Research Network [NICHD

ECCRN], 2005). Although most research has

examined the direct link between nonmaternal

care and child outcomes, some researchers

speculate that the extensive use of nonmaternal

care may result in poorer child outcomes in

part indirectly through undermining mothers’

sensitive parenting (Belsky, 1999; Brazelton,

1986). Mothers’ sensitive parenting generally

refers to the extent to which mothers respond

to their children’s needs promptly, remain

patient with their children, provide adequate

stimulation, and engage in playful interactions

(Belsky, 1999). Given that mothers’ sensitive

parenting is a key factor in shaping child

outcomes (NICHD ECCRN, 2001, 2003), this

speculation is important to explore, yet relatively

little is known about how nonmaternal care may

inf‌luence the degree to which mothers exhibit

sensitive parenting.

Attachment theory has led to the popular

premise that daily separations from their chil-

dren due to the extensive use of nonmaternal care

may result in less sensitivity in mothers’ parent-

ing (Belsky, 1999, 2001). Some researchers,

760 Journal of Marriage and Family 75 (June 2013): 760 –777

DOI:10.1111/jomf.12022

Nonmaternal Care and Maternal Sensitivity 761

however, have questioned the universality of

this link, emphasizing variations by contexts of

mothers’ lives (Clark, Hyde, Essex, & Klein,

1997). Furthermore, other researchers are skep-

tical about nonmaternal care having much inf‌lu-

ence on mothers’ parenting quality, suggesting

the possibility of a selection effect; that is,

maternal sensitivity and choices made concern-

ing nonmaternal care might both be dependent

on unobserved characteristics, such as moth-

ers’ child rearing orientation (NICHD ECCRN,

1999).

The purpose of the present research was to

address this debate. Using longitudinal data from

NICHD Study of Early Child Care and Youth

Development (SECCYD), we reexamined the

association between nonmaternal care character-

istics and maternal sensitivity. To address vari-

ations by context, we used a stress – resource

perspective and examined mothers’ resources,

such as family income and marital status, as

moderators of the association. To address the

selection issue, we used f‌ixed-effects models

that control for unmeasured characteristics that

could confound the observed association, and we

compared the results with those from random-

effects models. This study extends prior work in

that it is one of the f‌irst studies to (a) explicitly

use the stress – resource perspective in linking

nonmaternal care to maternal sensitivity and (b)

attempt to rule out self-selection bias.

BACKGROUND

Prior Research

Attachment theory (Ainsworth, 1989) under-

girds the premise that the use of nonmaternal

care during the f‌irst 3 years of children’s lives is

related to poorer quality of mothers’ parenting.

It posits that extended hours of daily separations

from their children may place constraints on

mothers’ ability to learn their children’s subtle

cues, which in turn results in a failure to pro-

vide sensitive parenting (Belsky, 1999, 2001).

Although a few empirical studies have found a

link between the amount of nonmaternal care and

less sensitivity in mothers’ parenting (Belsky,

1999; NICHD ECCRN, 1999), most studies have

not. One study found that the amount of nonma-

ternal care was related to less sensitive parenting

only for mothers with depressive symptoms and

mothers whose child had a diff‌icult temperament

(Clark et al., 1997). A few studies found no

relationships between the use or amount of non-

maternal care and maternal sensitivity (Booth,

Clarke-Stewart, Vandell, McCartney, & Owen,

2002; Sf‌ifter, Coulehan, & Fish, 1993). In addi-

tion, several studies found that full-time use of

nonmaternal care was related to more sensitive

parenting of mothers (Burchinal, Bryant, Lee,

& Ramey, 1992; Crockenberg & Litman, 1991;

Schwartz, 1983), especially when the quality of

care was high (Caruso, 1989; Edwards, Logue,

Loehr, & Roth, 1986; Howes & Olenick, 1986).

These inconclusive f‌indings may be

attributable to methodological differences and

limitations. First, most studies used a small,

local sample. Some studies used samples of

mothers whose children were in a particular

type of care (e.g., high-quality center care;

Edwards et al., 1986); other studies focused on

middle-class families (e.g., Schwartz, 1983).

The variations in sample characteristics indicate

the possibility that the link between nonmaternal

care and maternal sensitivity may depend on

context. Second, most studies did not adequately

control for characteristics that were related

to both nonmaternal care characteristics and

maternal sensitivity, such as mothers’ parenting

values, leaving the question of a selection effect

unanswered.

Another limitation is overreliance on attach-

ment theory. Because it heavily focuses on the

negative inf‌luence of long hours of nonmaternal

care, attachment theory pays little attention to

other effects that nonmaternal care may have on

mothers’ parenting, such as the positive effects

of high-quality care found in prior research

(e.g., Burchinal et al., 1992). Also, it ignores

the possibility of variations in the inf‌luence of

nonmaternal care on mothers’ parenting that are

due to context (e.g., Clark et al., 1997). An

alternative perspective that expands the scope of

the link between nonmaternal care and mothers’

parenting is needed.

In this article, we address the conceptual

and methodological issues in prior research

by using a stress – resource perspective that

provides a broader theoretical scope and by

examining f‌ixed-effects models that take into

account unobserved characteristics. Below, we

discuss how a stress – resource perspective

predicts the association between nonmaternal

care and mothers’ sensitive parenting and how

it further predicts the possibility of variations in

the link by levels of mothers’ resources. Then

we address the issue of selection effects.