Childrearing practices and developmental expectations for Mexican-American mothers and the developmental status of their infants.

Physical therapists rely, to some extent, on parents to promote the health and development of children with developmental delays and disabilities. Mandates for provision of developmental, health, and educational services, such as those from the Individuals With Disabilities Education Act amendments of 1997, (1) also define the roles of physical therapists by identifying them as crucial in supporting parenting and caregiving efforts in children's natural environments. The philosophy behind the designation of the physical therapist as a supportive figure is that, in theory, because parents are a constant in the child's life, (2) educating them about health and developmental needs will result in long-lasting positive results for the child. (3)

Although parents are expected to adapt their parenting behaviors and childrearing practices in response to educational programs, the impact of parent education on child outcome has not been consistent across socioeconomic and ethnic groups. (4,5) Differing outcomes have been observed between children from ethnic majority and minority families. (6-9) Ethnic minority families demonstrated inconsistent, less predictable, or less favorable outcomes. For example, teaching strategies that are believed to promote learning in children, such as the use of praise and directives, were correlated with greater gains in task performance in the case of Anglo-American children, but not in Mexican-American children. In other studies, (8,10) low birth weight appeared to have a differing relationship to the cognitive outcome of African-American children as compared with Anglo-American children from similar socioeconomic backgrounds.

The variability of impact of parent-focused education programs suggests that differing child and family outcomes may be due to differences in childrearing and parenting behaviors. Shonkoff et al. (11) found that family characteristics were stronger predictors of child and family change such as parenting stress and social support network size than were any specific aspects of early intervention services. For example, factors such as higher-quality home environments (based on the score on the Home Observation for Measurement of the Environment [HOME] Inventory (12)) were more predictive of positive developmental outcomes in children receiving early intervention than the type of early intervention service provided. Although familial or environmental factors have been implicated in physical therapy efficacy studies, follow-up research to investigate the nature or impact of these factors on child development and outcome has been sparse. Very little is known about the interplay between parental childrearing practices and how the quality of the child's immediate environment interacts with the child's characteristics and influences the selection of physical therapy interventions.

Research on programs that are designed to be used for parent education with minority families also suggests there is an impact on cultural childrearing processes. The sociocultural environment and the child's abilities are believed to be inextricably interdependent. (13,14) Culture is thought by many experts to mediate both the physical and environmental factors in development. (14) What is unclear is how the nuances of culture influence the interaction between parenting behaviors and the childrearing environment, presumably to generate optimal child development. Information on culture and childrearing is largely based on cross-cultural comparisons. (15) Cross-cultural findings provide valuable information about how groups differ. They also, however, can mask within-group variations that may exist in minority populations such as Hispanics. These factors, I believe, need to be understood to find out why some children do, or do not, exhibit optimal development. For example, one group studying school-age children found that, across cultures, the amount of time children spent on homework and the parental assistance they received were positively related to achievement, but this relationship was not statistically significant within cultural groups. (16)

Culture also is believed to play a role in determining the level of confidence with which childrearing ideas are held by parents, and it seems to influence the extent to which parents may be prepared to be flexible in the light of new information (such as that provided during parent education). (15,17,18) Thus, parents are likely to change or to resist change depending on whether the new information they receive is in accordance with the views held by people who are similar to them in terms of ethnicity, culture, or socioeconomic status. Whether or how new knowledge helps the parent to achieve culturally or locally valued goals may determine the success of physical therapy interventions that entail parent education.

In my view, therefore, examining how culture and childrearing practices are related to optimal or non-optimal child development in minority families is an essential first step toward investigating whether these practices may be enhanced through interventions such as physical therapy and may lead to positive developmental outcomes. Childrearing practices and parenting behaviors are defined as "specific, goal-directed behaviors through which parents perform their parental duties." (19(p488)) Parenting behaviors relate to unique parental behaviors that a child directly or indirectly experiences and that affect the child's development (eg, nurturing). (20) Some of the parenting behaviors are expressed through specific childrearing practices such as sensitivity to the child's cues during teaching sessions. (19)

The within-culture study discussed in this article focused on the relationships between various aspects of children's immediate caregiving environment and infants' developmental status among Mexican Americans. I examined the relationship between the mothers' developmental expectations and parenting behaviors, the quality of the home environment, and parent-infant interaction and how these variables are related to infants' developmental status.

Studying the link among ideas (expectations), behavior, and child outcome, I contend, requires a conceptual or theoretical framework regarding the interconnectedness of these variables. Two theoretical models provided a framework that I deemed suitable for investigating the relationship between childrearing and parenting behaviors and infants' developmental status within the cultural context. These are the "developmental niche" model (18) and Barnard's Child Health Assessment Interaction model. (21,22) The developmental niche proposed by Harkness and Super (18) identifies 3 components centrally related to parents and childrearing: (1) the physical and social settings of the child's life, (2) culturally regulated customs and practices of child care and childrearing, and (3) the psychology of the caretakers (eg, ideas about parenting). Barnard's model is related to parent-child interaction and contends that an infant's developmental outcome is based, in part, on the interaction between maternal and infant characteristics and the qualities of the environment. This model focuses on observations obtained from 2 familiar exchanges between the caregiver and child: feeding and teaching. The model also includes the contribution of the child to the parenting and childrearing process. The interaction between the child and the characteristics of the parent is believed to shape the child's developmental outcome. (21)

Inherent in these models are the assumptions that parents form childrearing beliefs and that acceptable behaviors on the part of the child are based, among other things, on sociodemographic and cultural factors. These beliefs are often expressed as ideas or expectations about child development. (23,24) In turn, childrearing ideas and expectations are believed to affect the child through numerous parenting behaviors or practices such as parent-child interactions (21) and the way in which parents structure the child's immediate caregiving environment. (12)

Mexican Americans were chosen for this study because evidence suggests that the developmental expectations and interactional processes used by Mexican Americans may differ from those of the majority culture. (25-28) Authors have described these families as nurturing, (27,29) egalitarian, and indulgent (29) toward their children. Unlike members of the predominant culture in the United States, Mexican-American parents are believed to place greater emphasis on the development of proper demeanor and a sense of dignity than on early achievement or attainment of developmental milestones. (29) How these childrearing expectations and values influence parenting behaviors (eg, discipline) and childrearing practices (eg, teaching interactions) and the child's developmental outcome is not clear. Mexican Americans have been under-represented in or excluded from many of the larger, longitudinal studies of early environmental and biological influences on the development of infants. (30-32)

Studies of maternal views of child development also suggest that there is a "complex picture" of diversity among Mexican-American mothers that may be associated with acculturation. (33-34) In the United States, acculturation has been defined as the extent to which a family has adopted the mainstream US culture. (35) Acculturation contributes to adaptations in lifestyles and childrearing behaviors in families. (7,36-38) For example, Mexican-American women (born in the United States) have been reported to demonstrate more prenatal health-related problem behaviors compared with Mexican-immigrant women. However, Mexican-American women also demonstrated more optimal teaching behaviors with their children. (39) Circumstances under which some families acquire certain aspects of the majority culture, or retain values and beliefs from their own culture, are less understood.

Pertinent to pediatric physical therapy and developmental research are the findings that Mexican-American children are over-represented among those exhibiting some of the deleterious consequences of impoverishment. (40) Poor functioning at school, (40) increased health risk, (41,42) and poor immunization status (37) are all more frequent in Mexican-American children than in children from other Hispanic groups. Researchers (14,34) have shown that these children also experience developmental risk from multiple rather than single sources and that the effects tend to be cumulative. Mexican American children represent the most rapidly growing group in the US population, and recent demographic trends suggest that Mexican-American children under the age of 18 years will comprise the largest minority group in the 21st century. (43)

In my study, I specifically addressed the following questions:

1. What is the nature of the relationships among maternal developmental expectations, maternal nurturing and discipline behaviors, the quality of parent-infant interaction, and the quality of stimulation in the home environment (childrearing and parenting behaviors) of Mexican-American families? How are these factors related to infants' developmental status?

2. Is the relationship between these childrearing and parenting behaviors (ie, maternal developmental expectations, maternal nurturing and discipline behaviors, the quality of parent-infant interaction, and the quality of stimulation in the home environment) and infants' developmental status affected by mothers' socioeconomic status (SES), years of education, and acculturation?

3. Which childrearing and parenting behaviors are important predictors of infants' developmental status among Mexican-American families?

4. Are there differences in childrearing and parenting behaviors among mothers based on their SES, their level of education and acculturation, and whether or not the infant has a diagnosis of developmental delay?

On the basis of findings from studies of children from the mainstream culture, studies of childrearing beliefs and practices among Mexican Americans, and theoretical frameworks previously discussed, I hypothesized that:

1. The developmental expectations and nurturing and discipline behaviors of Mexican-American mothers and the quality of the parent-child interaction and stimulation in the home environment would be positively correlated with infants' developmental status.

2. Mothers' SES, years of education, and acculturation would modify the relationships among these factors.

3. Maternal nurturing behaviors and developmental expectations would be strong predictors of infants' developmental status.

4. Differences would be observed in the mothers' mean scores on measures of childrearing practices based on their SES, their level of education and acculturation, and whether the infant has a diagnosis of developmental delay. Specifically, I hypothesized that: (1) mothers with middle SES or a high level of education would score higher on childrearing measures than those with low SES or a low level of education, (2) mothers who are not acculturated would report higher nurturing and score lower on measures of childrearing compared with those who are acculturated, and (3) mothers of infants with developmental delays would score lower on measures of childrearing and parenting behaviors compared with mothers of infants without developmental delays.

Method

Subjects

Participants were 62 mother-infant pairs who met the following criteria: (1) the infants had to be between 9 and 12 months of age and (2) at least one parent had to be Mexican or Mexican American. The sample was divided into categories according to level of acculturation (ie, monocultural Hispanic [not acculturated], monocultural non-Hispanic [acculturated], or bicultural) and level of SES (middle or low). The Bidimensional Acculturation Scale for Hispanics (BAS) was used to determine the level of acculturation. (44) The BAS provides an acculturation score for 2 cultural factors (ie, Hispanic and non-Hispanic), as well as a score representing biculturalism. The Hollingshead scale was used to determine SES. (45) Teenage mothers (less than 18 years of age when their infants were born) were excluded.

Mother-infant dyads were recruited from the greater Chicago metropolitan area by a bicultural recruiter who was fluent in Spanish and English. Flyers about the study (approved by the University of Illinois at Chicago Institutional Review Board-Human Subjects Review Committee) were distributed at centers that provided primary health care, Early Head Start program, or early intervention primarily to Mexican and Mexican-American families. Presentations about the study also were made to church groups and other community organizations. As part of the recruitment process, an initial interview form was used to gather demographic data such as parental marital status, level of education, type of occupation, number of children, household structure, and information on infants' medical, health, and developmental histories.

There are no absolute standards in the literature concerning sample size estimation for correlation and regression analysis. (46) Therefore, for this study, I used a sample size that I believed would permit simultaneous analysis of the childrearing variables. Based on these specifications, a target sample size of 60 (10 subjects per variable) was considered sufficient to estimate the parameters of the regression model. (47)

The average maternal age at the time of testing was 27 years (SD=5.4, range=18-40). Fifty-two mothers (52%) had completed high school, and 74% of the mothers were married. The infants' average age (adjusted for prematurity) at the time of testing was 12 months (SD=1.7, range=9-14). The mean gestational age at birth was 39 weeks (SD=1.6, range=32-42). Forty-seven percent of the infants were male, and 53% were female. Nearly a third of the infants had been previously diagnosed as having developmental delays and were receiving health care and developmental services. Forty-three percent of the mothers were classified as monocultural Hispanic, 37% were classified as bicultural, and 20% were classified as acculturated. The majority of the mother-infant dyads were categorized as having low SES (61%).

Instruments

In selecting the measurement tools I used, I gave preference to what I believe are comprehensive, standardized scales designed to measure a wide range of child-rearing activities and parenting behaviors. They also need to be based on areas specified in the conceptual models previously described. In addition, the scales had to: (1) have been shown to predict, or to be highly correlated with, child outcome in groups other than Mexican Americans; (2) have very good psychometric properties; and (3) be currently used by professionals to plan prevention and early intervention services. I reviewed the scales with other professionals who worked in Mexican-American communities. A comprehensive child assessment measure was selected (48) because multiple domains of an infant's development are interconnected, and a deficit in one area of development is likely to affect other domains.

Parenting Behaviors

The Parent Behavior Checklist (PBC) (20) was used to solicit and measure maternal child-rearing practices and expectations based on parental report. The PBC is a 100-item Likert-type scale that focuses on multiple rather than single aspects of childrearing and provides information on parents' beliefs and the childrearing behaviors that appeal to their sense of logic. The items also are meant to represent day-to-day activities and expectations that parents and their children experience (eg, "I allow messy play," "I spend at least 1 hour playing with or reading to my child," "My child should be able to play alone for 30 minutes"). The scale consists of 3 subscales that measure parents' developmental expectations (expectation subscale), parents' discipline strategies (discipline subscale), and how parents promote children's psychological growth (nurturing subscale). Raw scores for each subscale are converted into normalized t scores, with a mean of 50 and standard deviation of 10. Higher scores are associated with higher parental expectations, more frequent punishment, and more positive nurturing by parents. The PBC subscales demonstrate good internal consistency based on the values reported in the manual. (20) The scale also is available in Spanish. (49) Each subscale was analyzed separately using the t scores.

Childrearing Practices

The HOME Inventory (12) was used to measure the quantity and quality of stimulation in the home environment. The 45 binary items are organized into 6 subscales that are scored using semistructured interviews and observations of mother-child interactions during a home visit. Among the family characteristics assessed with the HOME Inventory are the amount of developmental and vocal stimulation, avoidance of restriction of motor and exploratory behavior, availability of play materials, frequency and stability of adult contact, and other home characteristics indicative of parental concern with achievement. The HOME Inventory is widely used in research and in planning interventions and has sound psychometric properties. (12,50) The subscales were based on a series of factor analyses that were subsequently tested with families from diverse backgrounds. Cronbach alpha coefficients for the 6 subscales ranged from .49 to .78. The point-biserial correlation coefficients between items and their subscales were greater than .25. Information on validation studies also has been reported. (50,51) Correlations between the HOME Inventory subscale scores and socioeconomic indices such as mother's education and father's occupation ranged from .24 to .50. (12) The HOME Inventory total raw score was used for analysis in this study. A recent factor analysis of the HOME Inventory revealed that Hispanic parents of preschool children showed a different factor structure than that of African-American and Caucasian mothers. Cultural factors were suspected but were not investigated. (51)

The Nursing Child Assessment Teaching Scale (NCATS) (22) was used to assess and score dyadic parent-child interactions. The NCATS is a binary scale of 73 items designed to assess interactions between caregivers and children during a teaching episode. The scale consists of 2 subscales that can be used to assess the parent's and the child's contributions to the interaction concurrently. The NCATS has sound psychometric properties. Cronbach alpha coefficients for the 6 subscales ranged from .52 to .87, with the highest coefficient observed for the total summary score (.87). The internal consistency was similar among Hispanic, African-American, and Anglo-American samples (.88-.90). Coefficients for test-retest reliability were .85 for the total parent score and .55 for the total infant score. (22) Studies on the criterion and construct validity of the NCATS have indicated differences in scores based on substance abuse by mothers and mothers who are abused. (22) The NCATS has been used extensively in research and practice. (52) Scores on the HOME Inventory and the NCATS have been found to correlate with the results of naturalistic observations of caregiving and parent-child interactions (r=.44-.69 for maternal behaviors and r=.19-.44 for child behaviors). (53) The NCATS caregiver and child total raw score was used for analysis in my study.

Infants' Developmental Status

The Bayley Scales of Infant Development II (BSID II) (48) was used to assess infants' overall developmental status. The original BSID (54) was widely used in developmental research for children 0 to 3 years of age. I used the revised version in my study because it includes more items and more representation of Hispanic children and covers a wider age span (0-42 months) than the original BSID. Because the BSID II contains items from the language domain, a separate language test was not used. There is currently no reliable Spanish-language test for Spanish-speaking infants and toddlers. Both the BSID II mental developmental index (MDI) and psychomotor developmental index (PDI) were used for analysis in this study.

Reliability

Three testers (including the author) participated in the study. Because I have used each of the tests described in both clinical practice and research, I provided training to the other 2 testers in the administration and scoring of the BSID II. In addition to testing infants, the training sessions also included a didactic session on psychometrics and discussions of item administration. Inter-rater reliability was conducted on 6 infants who met the age criteria of the sample but were not part of the study. Intraclass correlation coefficients (ICC [2,1]) for the BSID II ranged from .94 to .97 among the testers. The HOME Inventory manual (12) does not specify the training level required to administer and score the scale reliably. Ideas for reliability training were discussed with Robert H Bradley, one of the scale developers (personal communication, April 1999). A training videotape and scoring package provided by Bradley were used for reliability training. The ICCs for the HOME Inventory ranged from .92 to .95. To use the NCATS for research purposes, certification is required by the Nursing Child Assessment Satellite Training (NCAST) program at the University of Washington in Seattle. To use the scale for research, a percentage of agreement of 90% for inter-rater reliability is required according to the developers of the scale. As a certified NCAST instructor, I provided training to the other testers prior to reliability certification. The training course for reliability certification consists of approximately 40 hours. The testers were subsequently certified and recertified by the NCAST faculty at the University of Washington School of Nursing. All 3 scales were administered to each tester for the mother-infant dyads assigned to her.

Procedure

The recruiter obtained parental consent for children to participate and obtained information on the family intake form and the BAS. This information was used to classify mother-infant dyads. Thereafter, the recruiter scheduled a home visit to complete the PBC with the family, using a semistructured interview format. Although the PBC can be self-administered, the interview format was used to minimize variations due to varying levels of maternal education.

Within a month after the interview, a testing session was scheduled. Two experienced pediatric physical therapists and a Child Development Specialist supervisor conducted the testing sessions. The sequence of testing was as follows. At the beginning of the session, mothers were asked to select and teach infants a novel task or activity. They chose an activity from the list provided on the NCATS activity sheet, representing everyday tasks that parents teach their children at one point during the child's development. The teaching session was videotaped and scored at a later time. At the end of the teaching session, mothers were asked if they considered the teaching and interaction session typical of their usual experience in teaching their child a new task or activity.

Following observations of the interaction, the BSID II MDI was administered and scored according to procedures provided in the manuals. This was followed by the administration of the BSID II PDI. Instructions were given in the child's spoken language (English or Spanish), and the language used was noted on the form. The home environment was observed and scored throughout the testing session, or scoring was completed after the administration of the BSID II. The testing procedure described took an average of 1.5 to 2 hours (approximately 45 minutes to 1 hour involved the infant directly). A session was terminated if infants showed signs of fatigue or distress. In these cases, a second visit was scheduled within a week to complete the testing.

At the end of the testing session, the mothers were given $25 and a certification of appreciation and a toy or book for their infants. At the end of the data collection period, the recruiter conducted a brief exit interview by telephone with the mothers to solicit their reactions to the use of the questionnaires and the testing process. The testers shared the results of the developmental assessment with the parents. The 2 trained testers and I coded the mother-infant interaction during the teaching session from the videotapes using instructions provided in the NCATS manual. (22)

Data Analysis

The data were first examined for skewness. The relationships among mothers' nurturing and discipline behaviors, developmental expectations, mother-infant interaction, the quality of stimulation in the home environment, and infants' developmental status were examined using the Pearson product moment correlation. Partial correlations were used to examine the contribution of the effects of SES, years of education, and acculturation to the relationship between the mothers' childrearing practices and their infants' developmental status. The Fisher z transformation (55) was used to analyze the magnitude of differences between the 2 correlation coefficients.

To identify childrearing and parenting behaviors that were important predictors of the infants' developmental status among Mexican-American families, several multiple regression analyses were conducted. The first step was to select maternal childrearing and parenting behaviors and maternal demographic factors that explained the greatest amount of variance in the infants' developmental scores. There is no agreement in the literature regarding the selection method of variables that would maximize the predictive ability of a regression model. (55,56) According to Pedhazur, (56) the criteria for selecting the number of variables should take into consideration factors unique to the study, current theories, and ease of administering the measures and obtaining the measurements.

In this study, those childrearing behaviors and practices (as measured with the HOME Inventory, NCATS, and PBC nurturing subscale [PBC-N]) that were correlated with the BSID II were selected for inclusion in the regression model. Maternal and child characteristics that have been reported to be related to child outcome such as SES, maternal age, years of education, and gestational age at birth (GA) also were included in the regression model. (6,8,32,57) The data obtained for these variables also were correlated with the BSID II scores.

The next step was to use stepwise regression to determine variables that explained most of the variance in the BSID II scores. Residual and diagnostic plots were used to examine the presence of outliers and multicollinearity. Three outliers were found in the BSID II MDI scores. Their removal did not change the overall results of the multiple regressions. Multicollinearity was found between the NCATS and HOME Inventory scores.

To examine whether the scores on these childrearing scales differed based on the mothers' SES or level of education and whether their infants had developmental delays, a t test was used. An analysis of variance was used to determine whether the PBC, HOME Inventory, NCATS, and BSID II scores differed on the basis of acculturation.

Results

Table 1 presents descriptive statistics for the study variables. The PBC expectation subscale (PBC-E) t scores for the mothers ranged from 23 to 54. The t scores ranged from 34 to 62 for the PBC discipline subscale (PBC-D) and from 25 to 81 for the PBC-N. The mothers' mean PBC-E and PBC-D t scores were lower than the published normative mean t score of 50 (SD=10), whereas the mean PBC-N t score was similar to that of mothers in the norming group. The mean PBC-E and PBC-D t scores, although below the published mean, are considered to lie within the average range for infants of this age (average t scores=36-50). (23)

The HOME Inventory mean score of 37 (SD=5.3, range=19-44) is within the upper fourth of the range for the normative sample (range=37-45). (12) The NCATS total scores ranged from 29 to 66. The mean score of 53 (SD=7.7) is also below the mean scores reported in the NCATS manual (22) for non-Hispanic mothers, but within the average specified range.

The mean BSID II MDI score was 94 (SD=16.2, range=50-118), and the mean BSID II PDI score was 86 (SD=16.6, range=50-121). Although below the published norms, the means would be considered within the average range described in the test manual. (48)

Relationship Between Mothers' Childrearing and Parenting Behaviors and Infants' Developmental Status

Table 2 presents a matrix of correlations between mothers' scores on the PBC, HOME Inventory, and NCATS and their infants' scores on the BSID II. Statistically significant correlation coefficients were found among the PBC-N t scores, HOME Inventory scores, NCATS total scores, and BSID II MDI scores. The correlation coefficients between the PBC, HOME Inventory, and NCATS scores and the BSID PDI scores were low and not statistically significant, as had been hypothesized.

The results of the partial correlation analysis revealed differential effects of years of education, SES, and level of acculturation on the relationship between the childrearing measures and infant developmental status (Tab. 3). The coefficients between the PBC-N and BSID II MDI scores were not statistically significant when the effects of years of education (r=.11, P=.46), SES (r=.26, P=-.06), and acculturation (r=.24, P=-.09) were removed. The magnitude of the difference in the coefficients was statistically significant when the effect of maternal years of education was removed (Fisher z=-1.98), but not when the effects of SES (Fisher z=-.20) and acculturation (Fisher z=-1.03) were removed, indicating that years of education considerably modified the relationship between nurturing behaviors and infants' BSID II MDI scores.

The correlation coefficients between the HOME Inventory and NCATS scores and the BSID II MDI scores were slightly reduced after the effects of SES, years of education, and acculturation were controlled for (Tab. 3). However, the difference in the magnitude of coefficients was not statistically significant (Fisher z=-.20--1.03), indicating that influence of these demographic factors on the observed relationship could not be justified based on the results of this study.

Childrearing and Parenting Behaviors as Predictors of Infants' Developmental Status

Tables 4 and 5 summarize the results of multiple regression equations. Collectively, the PBC-N scores, HOME Inventory scores, NCATS scores, SES, maternal age, and GA predicted 45% of the variance in the infants' BSID II MDI scores. Of these variables, the HOME Inventory scores, NCATS scores, SES, and GA accounted for 43% of the variance in the infants' BSID II MDI scores (Tab. 4). Infants' GA, mothers' SES and age, and NCATS scores collectively explained 32% of the variance in the infants' BSID II PDI scores (Tab. 5). Only the infants' GA and the mothers' SES were predictors of the variance in infants' BSID II PDI scores (28%).

Differences Within Groups Based on Demographic Factors

Differences were found between the mother-infant dyads based on maternal SES, years of education and acculturation, and whether the infant was diagnosed as developmentally delayed (Tab. 6). As hypothesized, mothers with middle SES and those who had completed high school scored higher on the PBC-N, HOME Inventory, and NCATS, indicating that they endorsed more nurturing behaviors and displayed higher-quality mother-child interaction and higher quality and quantity of stimulation in the home environment compared with mothers with low SES and those who had not completed high school. No differences were observed between mothers' scores on the PBC, HOME Inventory, and NCATS based on whether or not their infants were diagnosed with developmental delays, as had been hypothesized.

Differences also were observed in the mean scores on the PBC-N, NCATS, and HOME Inventory based on the level of acculturation. The analysis of variance for the PBC-N revealed differences within the group based on level of acculturation (F=5.82; df=2,59; P=.005). Post hoc analyses using the Tukey honestly significant difference test for multiple comparisons revealed that mothers who were bicultural reported more nurturing behaviors and higher developmental expectations than mothers who were not acculturated. Differences also were found in the HOME Inventory and NCATS scores based on level of acculturation (F=3.49; df=2,50; P=.04 and F=4.08; df=2,50; P=.05, respectively). Mothers who were acculturated (English-speaking) scored higher on the HOME Inventory and NCATS than mothers who were not acculturated.

Discussion

Professionals engaging in early intervention often care about the relationship among parents' childrearing ideas, their actions, and child development, In this study, the link among infants' developmental status, determinants of parental behaviors (maternal SES, level of education, and developmental expectations), childrearing behaviors (nurturing and parent-child interactions), and childrearing practices (such as those reflected by the quality of the home environment), were examined in the context of culture. The results are consistent with the belief that there is a link between the childrearing behaviors and infant development in non-Hispanic mother-child dyads, (10,11,30) but not between parental expectations and child developmental status. The findings of this study also provide insight into the nature of this relationship. For example, the differential influence of the mother's SES and level of education and acculturation on the observed relationship raises questions about circumstances under which childrearing practices in Mexican-American families become culture-bound. These within-group differences have often been masked in cross-cultural studies. (25,51)

Mothers in this study tended to endorse higher nurturing behaviors than developmental expectations. Based on the findings of the correlation analyses, it would appear that Mexican-American mothers who strongly agreed that taking time to provide nurturing experiences to their children believe that it is important to have children who exhibit average to above-average cognitive functioning. The results of the partial correlation analysis, however, indicate that this relationship is influenced by the association that these 2 variables have with maternal level of education. It appears that whether or not the infants ok" mothers who reported high nurturing behaviors scored high on the BSID II MDI also may depend on mothers' level of education. The present data, therefore, suggest that among Mexican Americans, high nurturing behaviors may not be sufficient in terms of promoting optimal infant development.

The relationship between both the quality of the home environment and parent-infant interaction and the infants' cognitive development was robust (P=.001). The infants of mothers who scored high on the quality of the home and parent-infant interaction also tended to do so on the BSID II MDI, regardless of the mothers' SES and level of education or acculturation. The differences in how the 3 aspects of childrearing measured in this study are related to infant developmental status may be a function of the extent to which the infant was exposed to their influence. Parenting practices and behaviors that are sustained over time, as is the case with behaviors assessed by the HOME Inventory, may be more important in shaping infant development than parents' ideas, such as the reported expectations of development. This speculation may offer information on how childrearing practices may influence infant development. My findings, however, suggest the presence of an interaction among these variables that requires further exploration. To do so, a bidirectional analysis such as structural equation modeling and larger samples could be used. (58,59)

Both the home environment and parent-infant interaction predicted 37% of the variance in the infants' cognitive scores. The HOME Inventory scores were the largest contributor to this variance. This finding was unexpected because other authors have reported that the HOME Inventory scores were not as predictive of IQ among Hispanic-American preschool children as among African-American and Anglo-American preschool children. (60,61) In my study, the influence of the home environment was much greater than that of parents' SES and the observed parent-infant interactions. Several factors may explain this finding. First is the child's age at the time of prediction. Unlike in previous studies, the infants in my study were approximately 1 year of age. It is possible that, for Mexican Americans, the relationship becomes less clear as the child grows older. To examine this premise, longitudinal studies are necessary. There is also some confusion regarding the use of the term "Hispanics," as has been the case in previous studies. Hispanics are an ethnic group. Mexican Americans are not only culturally different from other Hispanic groups, they are also different in terms of country of origin and level of acculturation. In previous studies, only those parents who spoke English were included, whereas mothers who spoke Spanish only were included in my study. Although the predictive ability of the HOME Inventory is consistent with the abundant research on non-Mexican Americans that supports the use of the HOME Inventory as an independent predictor of cognitive development in children, (57) my findings underscore the importance of caution when interpreting studies that use categories such as "Hispanics" in cross-cultural comparisons.

The correlation coefficients between the childrearing practices and parenting behaviors that were examined in this study and the infants' BSID PDI scores did not reach statistical significance, as had been hypothesized. These findings were unexpected. Mexican-American mothers have been described as more likely to promote nurturing behaviors than early achievement or attainment of developmental milestones. (29) I expected that these preferences, particularly expectations about developmental milestones (PBC-E scores), would be associated with infants' motor development and specifically with delayed achievement of motor milestones. Although the mothers' developmental expectations (PBC-E scores) were indeed lower than those of the non-Mexican-American mothers upon whom the PBC was normed and their infants' mean scores on the BSID II PDI were below those of the normative sample, the correlation between the scores obtained for these 2 variables was low. Perhaps this relationship develops over time as the child becomes more independent in terms of mobility (eg, walking).

The low correlations between motor developmental status and the childrearing measures raise a few issues that warrant further discussion and investigation. One possibility is that factors other than the childrearing behaviors examined in this study may be more related to motor development. A second possibility may be lack of variability among the scores. For example, Abbott and associates (62) found that low correlations between scores on 3 of the HOME Inventory subscales and scores on the Alberta Infant Motor Scale were due to lack of variance in the HOME Inventory scores. In my study, a lack of variance in scores did not appear to be the issue, particularly given the dispersion of the BSID PDI scores. A third possibility is that motor development may be more sensitive to the influence of biological factors such as GA than to the influence of environmental factors such as those assessed with the HOME Inventory. For example, the results of the multiple regressions indicate that GA was an important predictor of the infants' BSID PDI scores. If this is the case, the findings raise a question concerning whether interventions that are intended to enhance maternal childrearing behaviors and the caregiving environment investigated in this study can positively influence motor outcome in infants with motor delays.

Another possibility is that the poor correlation between the scores obtained for the childrearing measures and motor outcome may be a function of how the construct of motor development is operationally defined and measured by various tests. Reports of low correlations between scores obtained for childrearing/environmental measures and scores obtained for motor development have come from studies in which the original BSID was used as an outcome measure. Although the BSID II includes an improved representation of children by race (Hispanics are still under-represented) and more motor items, the original BSID (on the motor subscale) primarily assess infants' ability to change positions. Items that assess reach, grasp, and manipulation comprise the mental subscale of the test. However, these items comprise the fine and gross motor subscales of the Peabody Developmental Motor Scales, (63) suggesting that other motor tests may yield different results. Indeed, using structural equation modeling, Garrett and associates (58) were able to demonstrate a direct relationship between children's developmental status (motor social development and cognition) and the quality of the environment in children aged 0 to 48 months from low-income families. This finding suggests that the impact of the home environment on motor performance should not be readily dismissed in this population.

As hypothesized, there was evidence of differences among mothers' childrearing practices and parenting behaviors based on the level of acculturation. In this study, only mothers with lower levels of acculturation reported lower nurturing, discipline, and developmental expectations of their infants in contrast to acculturated (English-speaking) or bicultural mothers. The expectation that nonacculturated Hispanic mothers would report higher nurturing behaviors was not supported. This expectation was based on previous studies in which immigrant Mexican-American mothers tended to show more protective childrearing behaviors than their native-born counterparts. (39,64) Lower scores on measures of parent-child interaction and the quality of stimulation in the home environment were expected from this group because new immigrant mothers are over-represented among those with lower levels of education and SES than acculturated mothers (bicultural and English-speaking). (7)

Because differences in childrearing and parenting behaviors were related to mothers' level of education and SES, I believe we need to be cautious when interpreting cross-cultural findings. Most of what has been reported in literature concerning early intervention with children from ethnic minority populations has been based on children from low-income and disadvantaged backgrounds. My findings, however, demonstrate that cross-cultural findings can mask within-group variations that may be related to other maternal characteristics or demographics.

Clinical Implications

The findings of my study have implications for program planning and development. Infants who are at risk for developmental delay or disabilities are the focus of early intervention. (11) Given the increasing importance placed on family-centered care and the increasing roles of family members in caring for these infants, information that can maximize parental effectiveness is needed. Understanding the nature of the relationships among developmental outcomes and childrearing practices in families of varying cultural backgrounds and SES could be helpful in (1) identifying children who are at risk for health and developmental problems and (2) developing individual family service plans (IFSPs).

Apparently, low scores on measures of nurturing behaviors and home environment among Mexican-American mothers may represent a high-risk caregiving environment for young infants, particularly if mothers have low levels of education. Because the observed link between mothers' nurturing behaviors and their infants' developmental status appears to be mediated by the home environment and the mother's level of education, provision of intervention strategies targeted toward improving the home environment appears central. This is particularly important because the quality and quantity of stimulation in the home environment is believed to mediate the influence of demographic and socioeconomic variables on the child's developmental status. (57,58)

The results of this study suggest that children of Mexican-American mothers from families with low SES who are born prematurely may be at high risk for motor delay. Unlike measures of childrearing behaviors, however, demographic markers of the environment in which children are raised such as SES or family income are limited in terms of program planning. These markers do not provide information that therapists can use to develop prevention or intervention strategies. The association among the mothers' SES, HOME Inventory scores, and BSID II scores, however, may help physical therapists gear interventions toward identifying resources within the community that are free and readily accessible, such as toy-lending libraries and transportation coupons. This may help change childrearing practices for the low-SES families in this group. The low developmental expectations and motor scores observed in this study accentuate the importance of making comprehensive, collaborative assessments of the social and cultural environment in which the child is raised a part of the process culminating in the development of the IFSP.

Many clinicians have emphasized a need to understand culture and its influence on child development. (65) Based on the findings of developmental expectations, it appears that low expectations, especially by nonacculturated Mexican-American mothers, may influence the parents' level of participation in interventions that could promote independence in young infants, particularly self-help skills such as feeding. Previous research suggests that parents are likely to change or resist change depending on whether the new information is in accordance with the views of people who are similar to them. (66) Although parents may vary in their exposure to a culture, and thus their willingness to adopt new ideas, physical therapists may need to consider the characteristics of new information (eg, content of parent education) beyond the child's developmental status, including the context of the social and cultural environment in which the mothers belong.

In this study, childrearing practices were poorly correlated with motor developmental status, which raises the possibility that the link between these factors occurs over time rather than concurrently. If that is the case, the findings may have important implications for what constitutes the sufficient length or duration of intervention for these infants.

Limitations

The data obtained in this study are cross-sectional in nature and therefore do not address causality or create relationships. Longitudinal research is needed to obtain a better insight into causality. The results of the multiple regressions suggest the presence of statistical interactions that could not be explored due to the limited sample. A large sample would permit the use of modeling techniques such as structural equation modeling to concurrently assess the strength of the direct and indirect relationships of PBC-N, HOME Inventor, and NCATS scores to BSID II scores (mediated through SES, level of education, and acculturation). Overall, the findings are encouraging, and continuing research in this area is important, particularly for children with disabilities from cultural minorities and their families. The fact that the variables studied explained only 32% of the variance in the infants' motor development status underscores the need for continued search for determinants of child motor outcome.

Summary and Conclusions

The present study was designed to assess multiple aspects of parenting behaviors and the childrearing environment of young infants, which have been found to contribute to developmental outcome in children from the mainstream culture. Parent reports were combined with direct observations. Taken together, the results of the correlations and multiple regression analyses partially support the hypothesis that there is a link between some aspects of the mothers' childrearing behaviors and their infants' cognitive developmental status. For motor developmental status, the association appears stronger with the infants' characteristics (eg, GA) than with childrearing practices. The findings on motor developmental status suggest the need for specificity in identifying contextual determinants of early motor outcomes. The results also suggest that the BSID II PDI may not be an appropriate outcome measure for studies of the effects of physical therapy interventions that are geared toward promoting motor outcome through childrearing practices.

Although my findings are consistent with those of studies of non-Hispanic families, they do not fully support previous observations of childrearing among Mexican-American mothers of preschool children. The hypothesis that the mothers' nurturing and developmental expectations would predict the infants' developmental status in this population was not supported. Although correlated, these caregiving variables were not predictive of infants' cognitive and motor developmental status at 1 year of age.

Table 1.
Mean Scores and Standard Deviations for Parent Behavior Checklist
(PBC) t Scores, Home Observation for Measurement of the Environment
(HOME) Inventory, Nursing Child Assessment Teaching Scale (NCATS),
and Revised Version of the Bayley Scales of Infant Development
(BSID II) (a)

                                          Acculturation
                       SES
              Total                                         English-
Scale         Sample   Low      Middle    Hispa-   Bicul-   Speaking
                                          nic      tural
PBC-E
  [bar.X]     37       36        39       34       40        38
  SD           8.9      8.4       9.6      7.8      8.8      10.0
  Range       24-54    25-54     24-54    25-52    24-54     25-54

PBC-D
  [bar.X]     41       43        39       44       40        38
  SD           7.8      8.2       6.9      8.7      7.4       4.5
  Range       34-63    34-63     35-56    35-63    34-56     35-48

PBC-N
  [bar.X]     51       48        55       45       57        52
  SD          12.1     11.1      12.6     10.0     13.1       9.2
  Range       25-81    25-81     37-81    25-65    30-81     40-73

HOME
Inventory
  [bar.X]     37       35        40       35       38        40
  SD           5.3      5.5       2.4      6.3      3.5       3.7
  Range       19-44    19-44     37-44    19-44    30-43     32-44

NCATS total
  [bar.X]     53       50        58       49       55        56
  SD           7.7      7.6       5.0      7.5      5.5       9.0
  Range       29-66    29-64     49-66    29-62    48-63     33-66

BSID II
  MDI
  [bar.X]     94       92       101       88       98       103
  SD          16.2     17        12       17       13        12
  Range       50-119   50-119    64-115   50-111   62-119    77-119

PDI
  [bar.X]     86       83        92       83       86        93
  SD          17.6     16.1      16.6     17       15        17
  Range       50-118   50-117    71-118   50-117   61-114    57-118

              Developmental Status

                              No              Level of Education
              Developmental   Developmental
Scale         Delay           Delay           Low      High

PBC-E
  [bar.X]     39               40             35       39
  SD           7.7              8.0            9.0      8.7
  Range       25-54            24-54          25-54    24-54

PBC-D
  [bar.X]     41               41             42       40
  SD           8.3              7.7            8.5      7.2
  Range       35-56            34-63          34-63    35-57

PBC-N
  [bar.X]     41               52             46       56
  SD           10.5             9.9           11.0      11.4
  Range       25-73            30-81          25-78    33-81

HOME
Inventory
  [bar.X]     35               37             35       39
  SD           7.3              4.3            6.4      2.9
  Range       19-39            29-44          19-44    32-44

NCATS total
  [bar.X]     51               53             50       56
  SD           8.9              7.2            8.4      5.7
  Range       29-54            33-66          29-66    41-64

BSID II
  MDI
  [bar.X]     81              100             91       99
  SD           18.6            11.4            18.4     12.5
  Range       54-98            77-119         50-119   62-119

PDI
  [bar.X]     67               93             83       88
  SD           12.1            11.7            17.1    3.10
  Range       50-87            57-118         50-117   57-118

(a) SES=socioeconomic status, PBC-E=Parent Behavior Checklist
expectation subscale, PBC-D=Parent Behavior Checklist discipline
subscale, PBC-N=Parent Behavior Checklist nurturing subscale,
MDI=mental developmental index, PDI=psychomotor developmental
index.

Table 2.
Pearson Correlation Coefficients (r) Among the Parent Behavior
Checklist (PBC), Home Observation for Measurement of the
Environment (HOME) Inventory, Nursing Child Assessment Teaching
Scale (NCATS), and Revised Version of the Bayley Scales of Infant
Development (BSID II) (a)

                                          HOME
                                          Inven-              BSID
              PBC-E     PBC-D   PBC-N     tory      NCATS     II-MDI
PBC-E
PBC-D         .07
PBC-N         .44 (b)   -.14
HOME          .13        .03    .62 (b)
  Inventory
NCATS         .03        .04    .48 (b)   .68 (b)
BSID II-MDI   .04        .01    .36 (b)   .53 (b)   .55 (b)
BSID II-PDI   .08        .03    .11 (b)   .27       .23       .55 (b)

(a) PBG-E = Parent Behavior Checklist expectation subscale,
PBC-D = Parent Behavior Checklist discipline subscale, PBC-N = Parent
Behavior Checklist nurturing subscale, MDI = mental developmental
index, PDI = psychomotor developmental index.

(b) P<.01.

Table 3.
Results of Pearson r Partial Correlations (a)

                                             Correlation
Variable Removed     Variables               Coefficients   z

Level of education   HOME Inventory x MIDI   .51 (P=.001)   -0.20
                     NCATS x MIDI            .44 (P=.001)   -1.03
                     PBC-N x MIDI            .11 (P=.46)    -1.98 (b)

SES                  HOME Inventory x MIDI   .53 (P=.001)   -0.01
                     NCATS x MDI             .49 (P=.001)   -0.58
                     PBC-N x MDI             .26 (P=.06)    -0.78

Acculturation        HOME Inventory x MIDI   .53 (P=.001)   -0.01
                     NCATS x MIDI            .48 (P=.001)   -0.67
                     PBC-N x MIDI            .24 (P=.09)    -0.93

(a) SES = socioeconomic status, PBC-N = Parent Behavior Checklist
nurturing, subscale, MDI = mental developmental index of revised
version of the Bayley Scales of Infant Development (BSID II),
PDI = psychomotor developmental index of the BSID II. NCATS = Nursing
Child Assessment Teaching Scale.

(b) P<.05, z>-1.96

Table 4.
Results of the Multiple Regression Equations for Prediction of Bayley
Scales of Infant Development-Revised Version (BSID II) Mental
Developmental Index Scores From the Home Observation for Measurement
of the Environment (HOME) Inventory, Nursing Child Assessment Teaching
Scale (NCATS), Maternal Socioeconomic Status (SES), and Infants'
Gestational Age at Birth

                         Standardized
                         Regression
                         Coefficient
Predictor Variables      (Beta)         t      P

HOME Inventory (R=.69,   .26            1.64   .10
  P=.001, adjusted
  [R.sup.2]=.43)
NCATS                    .18            1.20   .23
Gestational age          .22            1.96   .05
SES                      .25            2.01   .05

Table 5.
Results of Multiple Regression Equations for Prediction of Bayley
Scales of Infant Development-Revised Version (BSID II) Psychomotor
Developmental Index Scores From the Mother's Age, Socioeconomic
Status (SES), Nursing Child Assessment Teaching Scale (NCATS), and
Infants' Gestational Age at Birth

                          Standardized
                          Regression
                          Coefficient
Predictor Variables       (Beta)         t       P

Gestational age (R=.61,    .52            4.28   .0001
  P=.001, adjusted
  [R.sup.2]=.032)
SES                        .32            2.30   .02
Mother's age              -.03           -0.26   .79
NCATS                     -.71           -0.49   .62

Table 6.
Results of t Tests Comparing the Parent Behavior Checklist (PBC),
Home Observation for Measurement of the Environment (HOME) Inventory,
and Nursing Child Assessment Teaching Scale (NCATS) Scores Based on
Maternal Socioeconomic Status (SES), Level of Education, and
Diagnosis of Developmental Delay (a)

Demographic            Childrearing
Variables              Measures         t        df   P (2-tailed)

SES                    PBC-E            -1.871   59   .07
                       PBC-D             0.157   59   .88
                       PBC-N            -3.752   59   .001
                       HOME Inventory   -4.317   51   .001
                       NCATS            -2.482   49   .02

Level of education     PBC-E            -1.591   60   .117
                       PBC-D             0.845   60   .40
                       PBC-N            -3.407   60   .001
                       HOME Inventory   -2.846   51   .006
                       NCATS            -2.929   49   .005

Developmental status   PBC-E             0.444   59   .66
                       PBC-D            -0.078   59   .94
                       PBC-N            -0.343   59   .73
                       HOME Inventory    1.089   51   .28
                       NCATS             0.925   49   .36

(a) PBC-E = Parent Behavior Checklist expectation subscale,
PBC-D = Parent Behavior Checklist discipline subscale, PBC-N = Parent
Behavior Checklist nurturing subscale.

References

(1) Individuals With Disabilities Education Act amendments of 1997, 20 USC 1401.

(2) Shelton TL, Stepanek JS. Family-Centered Care for Children Needing Specialized Health and Developmental Health and Developmental Services. 3rd ed. Bethesda, Md: Association for the Care of Children's Health; 1994.

(3) Farran DC. Another decade of intervention for children who are low income or disabled: what do we know now? In: Shonkoff JP, Meisels SJ, eds. Handbook of Early Childhood Intervention. 2nd ed. New York, NY: Cambridge University Press; 2000:510-548.

(4) McCormick MC, McCarton C, Brooks-Gunn J, Gross RT. The infant health and development program: interim summary. J Dev Behav Pediatr. 1998;19:359-370.

(5) Brooks-Gunn J, Berlin LJ, Fuligni AS. Early childhood intervention programs: what about the family? In: Shonkoff JP, Meisels SJ, eds. Handbook of Early Childhood Intervention. 2nd ed. New York, NY: Cambridge University Press; 2000:549-577.

(6) Bradley RH, Whiteside L, Mundfrom DJ, et al. Early indications of resilience and their relation to experiences in the home environments of low birthweight, premature children living in poverty. Child Dev. 1994;65:346-360.

(7) Garcia-Coll CT. Developmental outcome of minority infants: a process-oriented look into our beginnings. Child Dev. 1990;61:270-289.

(8) Brooks-Gunn J, Klebanov PK, Duncan GJ. Ethnic differences in children's intelligence test scores: role of economic deprivation, home environment, and maternal characteristics. Child Dev. 1996;67: 396-408.

(9) Moreno RP. Everyday instruction: a comparison of Mexican American and Anglo mothers and their preschool children. Hisp J Behav Sci. 1997;19:527-539.

(10) Benasich AA, Brooks-Gunn J. Maternal attitudes and knowledge of child-rearing: associations with family and child outcomes. Child Dev. 1996;67:1186-1205.

(11) Shonkoff JP, Houser-Cram P, Krauss M, Upshur C. Development of infants with disabilities and their families: implications for theory and service delivery. Monogr Soc Res Child Dev. 1992;57:v-vi, 1-153.

(12) Caldwell BM, Bradley RH. Home Observation for Measurement of the Environment. Little Rock, Ark: Center for Research on Teaching and Learning at the University of Arkansas; 1984.

(13) Laosa LM. Maternal teaching strategies in Chicano and Anglo-American families: the influence of culture and education on maternal behavior. Child Dev. 1980;51:759-765.

(14) Garcia-Coll C, Magnuson K. Cultural differences as sources of developmental vulnerabilities and resources. In: Shonkoff JP, Meisels SJ, eds. Handbook of Early Childhood Intervention. 2nd ed. New York, NY: Cambridge University Press; 2000:549-577.

(15) McLoyd VC. Cultural influences in a multicultural society: conceptual and methodological issues. In: Masten AS, ed. Cultural Processes in Child Development: The Minnesota Symposia on Child Psychology. Mahwah, NJ: Lawrence Erlbaum Associates Inc; 1999:23-135.

(16) Chen C, Stevenson HW. Homework: a cross-cultural examination. Child Dev. 1989;60:551-561.

(17) Harwood RL, Schoelmerich A, Ventura-Cook E, et al. Culture and class influences on Anglo and Puerto Rican mothers' beliefs regarding long-term socialization and child behavior. Child Dev. 1996;67: 2446-2461.

(18) Harkness S, Super C. Introduction. In: Harkness S, Super C, eds. Parents' Cultural Belief Systems: Their Origins, Expressions, and Consequences. New York, NY: Guilford Press; 1996:1-23.

(19) Darling N, Steinberg L. Parenting style as context: an integrative model. Psychol Bull. 1993;113:487-496.

(20) Fox RA. Parent Behavior Checklist. Brandon, Vt: Clinical Psychology Publishing Co; 1994.

(21) Barnard KE, Eyres SJ. Child Health Assessment, Part 2: The First Year of Life. Washington, DC: US Government Printing Office; 1979. Dept of Health, Education, and Welfare publication HRA 79-25.

(22) Sumner F, Spietz A. NCATS/Caregiver/Parent-Child Interaction Teaching Manual. Seattle, Wash: NCATS Publications; 1994.

(23) Fox RA. Development of an instrument to measure the behaviors and expectations of parents of young children. J Pediatr Psychol. 1992;17:231-239.

(24) Fox RA, Platz DL, Bentley KS. Maternal factors related to parenting practices, developmental expectations, and perceptions of child behavior problems. J Genet Psychol. 1995;156:431-441.

(25) Gutierrez J, Sameroff A. Determinant of complexity in Mexican American and Anglo American mothers' conception of child development. Child Dev. 1990;61:384-394.

(26) Bartz KW, Levine ES. Child rearing by black parents: a description and comparison to Anglo and Chicano parents. J Marriage Fam. 1978;40:709-719.

(27) Vega WA. Hispanic families in the 1980s: a decade of research. J Marriage Fam. 1990;52:1015-1024.

(28) Ehling MB. The Mexican American (El Chicano). In: Clark AL, ed. Culture and Childrearing. Philadelphia, Pa: FA Davis Co; 1981:192-209.

(29) Zuniga M. Families with Latino roots. In: Lynch EW, Hanson MJ, eds. Developing Cross-Cultural Competence: A Guide for Working With Young Children and Their Families. Baltimore, Md: Paul H Brookes Publishing Co; 1992:51-179.

(30) Kochanska G, Clark LA, Goldman MS. Implications of mothers' personality for their parenting and their young children's developmental outcomes. J Personality. 1997;65:387-420.

(31) Ramey C, Bryant D, Wasik B, et al. Early intervention for high-risk children: the Carolina early intervention program. Prevention in Human Services. 1990;7:33-57.

(32) Brooks-Gunn J, McCarton C, Casey P, et al. Early intervention in low-birth-weight premature infants: results through age 5 years from the Infant Health and Development Program. JAMA. 1994;272: 1257-1262.

(33) Gutierrez J, Sameroff AJ, Karrer BM. Acculturation and SES effects on Mexican-American parents' concepts of development. Child Dev. 1988;59:250-255.

(34) Garcia-Coll C, Vazquez-Garcia HA. Hispanic children and their families: on a different track from the very beginning. In: Fitzerald HE, Lester BM, Zuckerman B, eds. Children and Policy of Poverty: Research, Health Issues. New York, NY: Garland; 1995:57-83.

(35) Cuellar H, Harris LC, Jasso R. An acculturation scale for Mexican American normal and clinical populations. Hisp J Behav Sci. 1980;2: 199-217.

(36) Zambrana RE, Scrimshaw SCM, Collins N, Dunkel-Schetter C. Prenatal health behaviors and psychosocial risk factors in pregnant women of Mexican origin: the role of acculturation. Am J Public Health. 1997;87:1022-1026.

(37) Anderson LM, Wood DL, Sherbourne CD. Maternal acculturation and childhood immunization levels among children in Latino families in Los Angeles. Am J Public Health. 1997;87:2018-2021.

(38) Rauh VA, Wasserman GA, Brunelli SA. Determinants of maternal child-rearing attitudes. J Am Acad Child Adolesc Psychiatry. 1990;29: 375-381.

(39) Chavez JM, Buriel R. Mother-child interactions involving a child with epilepsy: a comparison of immigrant and Native-born Mexican Americans. J Pediatr Psychol. 1988;13:349-361.

(40) Chapa J, Valencia RR. Latino population growth, demographic characteristics, and educational stagnation: an examination of recent trends. Hisp J Behav Sci. 1993;5:165-187.

(41) Does Chicago's Latino Population Mirror the National Latino Population? Chicago, Ill: Latino Institute Research; October 1993.

(42) Frank-Stromborg M. Changing Demographics in the United States: Implications for Health Professionals. Paper presented at the American Cancer Society National Conference, San Francisco, Calif. 1990:1-5.

(43) US Bureau of the Census. Statistical Abstract of the United States: 2002. Washington, DC: US Government Printing Office; 2002.

(44) Marin G, Gamba RJ. A new measurement of acculturation for Hispanics: the bidimensional acculturation scale for Hispanics (BAS). Hisp J Behav Sci. 1996;18:297-316.

(45) Hollingshead AB. Four-Factor Index of Social Position. New Haven, Conn: Yale University; 1975.

(46) Hair JF, Anderson RE, Thatham RL, Black WC. Multivariate Data Analysis. Upper Saddle River, NJ: Prentice Hall; 1998.

(47) Kraemer HC, Thiemann S. How Many Subjects? Statistical Power Analysis in Research. Newbury Park, Calif: Sage Publications Inc; 1987: 59-65.

(48) Bayley N. Bayley Scales of Infant Development Manual, Second Edition. San Antonio, Tex: The Psychological Corporation; 1993.

(49) Solis-Camara RP, Fox RA. Parenting practices and expectations among Mexican mothers with young children. J Genet Psychol. 1996; 157:465-476.

(50) Bradley RH. Children's home environments, health, behavior, and intervention efforts: a review using the HOME inventory as a marker measure. Genet Soc Gen Psychol Monogr. 1993;119:437-490.

(51) Bradley RH, Mundfrom DJ, Whiteside L, et al. A factor analytic study of the infant-toddler and early childhood versions of the HOME Inventory administered to white, black, and Hispanic American parents of children born preterm. Child Dev. 1994;65:880-888.

(52) Barnard KE, Hammond MA, Booth CL, et al. Measurement and meaning of parent-child interaction. In: Morrison FJ, Lord CE, Keating DP, eds. Applied Development Psychology. Vol III. New York, NY: Academic Press; 1989:39-80.

(53) Tesh E M, Holditch-Davis D. HOME Inventory and NCATS: relation to mother and child behaviors during naturalistic observations. Res Nurs Health. 1997;20:295-307.

(54) Bayley N. Bayley Scales of Infant Development Manual. San Antonio, Tex: The Psychological Corporation; 1969.

(55) Daniel WW. Biostatistics: A Foundation for Analysis in the Health Sciences. New York, NY: John Wiley & Sons Inc; 1999.

(56) Pedhazur EJ. Multiple Regression in Behavioral Research: Explanation and Prediction. New York, NY: Hold, Rinehart and Winston; 1982.

(57) Klebanov PK, Brooks-Gunn J, McCarton C, McCormick MC. The contribution of neighborhood and family income to developmental test scores over the first three years of life. Child Dev. 1998;69: 1420-1436.

(58) Garrett P, Ferron J, Ngandu N, et al. A structural model for the developmental status of young children. J Marriage Fam. 1994;56: 147-163.

(59) Kline RB. Principles and Practice of Structural Equation Modeling. New York, NY: Guilford Press; 1998.

(60) Bradley RH, Caldwell BM, Rock SL, et al. Home environment and cognitive development in the first three years of life: a collaborative study involving six sites and three ethnic groups in North America. Dev Psychol. 1989;25:217-235.

(61) Bradley RH, Whiteside LA, Caldwell BM, et al. Maternal IQ, the home environment, and child IQ. Int J Behav Dev. 1993;16:61-74.

(62) Abbott AL, Bartlett DJ, Kneale Fanning JE, Kramer J. Infant motor development and aspects of the home environment. Pediatric Physical Therapy. 2000;12(2):62-67.

(63) Folio MR, Fewell RR. Peabody Developmental Motor Scales. Austin, Tex: Pro-ed; 2000.

(64) Negy C, Woods DJ. The importance of acculturation in understanding research with Hispanic-Americans. Hisp J Behav Sci. 1992;14: 224-247.

(65) McLain C, Provost B, Crowe TK. Motor development of two-year-old typically developing Native American children on the Bayley Scales of Infant Development II Motor Scale. Pediatric Physical Therapy. 2000; 12(3):108-113.

(66) Turner JC, Oakes PJ. The significance of the social identity concept for social psychology with reference to individualism, interactionism and social influence. Bri J Soc Psychol. 1986;25:237-252.

THA Kolobe, PT, PhD, is Associate Professor, Department of Rehabilitation Science, University of Oklahoma Health Sciences Center, 801 NE 13th St, Oklahoma City, OK 73104 (USA) ([e-mail]).

The author acknowledges Everett Smith for contributing to fund procurement, Vince Allocco for providing institutional liaisons, Alma Childers for recruiting subjects, Lupe Pasillas and Louisa Susman for helping with data collection, and Robert Palisano for consultation during preparation of the manuscript.

The study was approved by he University of Illinois at Chicago Institutional Review Board-Human Subjects Review Committee (IRB Research Protocol #199841404).

This work was supported by a grant from the University of Illinois at Chicago Campus Research Board.

This article was received October 17, 2002, and was accepted October 17, 2003.