Both researchers and practitioners have a common need to assess the resources families have available to them. Resource assessment is considered essential to understanding risks and strengths. Typically, family resources are measured via self-report of income, education, social support, and family demographics. Most likely as a result of difficulties with regard to measurement, researchers conceptually interested in the construct of family resources have frequently turned to single or combined measures of income and education. This approach fails to consider that a family's perception of their situation may differ from that intimated strictly by the financial resources and the educational attainment of its members. Furthermore, this restrictive approach does not consider resources that may have only limited relationships with available finances or educational achievement, such as adult family member having time available to be with the family.

As an alternative to relying solely on family income or educational level, other researchers have attempted to assess family resources by asking family members to rate perceived adequacy of available resources across an array of specific areas, effectively broadening the nature of the construct (Dunst & Leet, 1987). There is some evidence that, in contrast to approaches relying on measures of income and education, approaches to measuring resources that focus on family member perception of available resources are more likely to yield measures that predict behavior and correlate more highly with general well being and compliance with treatment programs that the family is participating in (Dunst, Leet, & Trivette, 1988).

The Family Resource Scale (FRS) is an assessment of family members perceptions of available resources across a range of specific areas. The FRS was originally created as a clinical tool to assist practitioners in developing treatment and intervention plans for families (Dunst & Leet, 1987). The measure consists of thirty items which respondents rate using a five-point Likert-type scale anchored by "not at all adequate" to "almost always adequate."

Research Questions

Sample

As part of a 4 year longitudinal intervention study (Ramey, Ramey, & Phillips, 1996) approximately 7500 families of former Head Start children were followed from kindergarten through third grade. Data collected include annual family interviews (98% of respondents to the family interview were a parent or grandparent) and assessments of child outcomes. Data were collected at 30 sites across the U.S. For present purposes Caucasian, African American, and Latino families (representing 88% of the total sample) are included in analyses.

Data Analyses

Analyses of these data were done using the general linear model (GLM). Because of the large sample size effect sizes are reported rather than significance levels and F tests. An effect must account for at least 1% of the variance in the outcome variable to be considered practically significant. All effects reported are statistically significant at our criteria of p<.001. Listwise deletion was used resulting in differing sample sizes, sometimes very different since data may come from different sources and different time periods.

Measures

Family Resource Scale - Revised (FRS - R): In previous psychometric work the Family Resource Scale (Dunst & Leet, 1987) was revised. The number of items was reduced from 30 to 21, and factor analyses were used to demonstrate the existence of 4 subscales, Basic Needs, Money, Time for Self, and Time for Family. The scale was found to be psychometrically sound in a large national sample. Multi-group confirmatory factor analyses were used to show that the factor structure was valid for the three ethnic groups in these analyses.

Ethnicity of Child: The child's ethnicity was asked as part of the family interview.

Percent of Poverty: Family income was calculated as a percent of the federal poverty level (which takes into account family size and changes by year).

Child Cognitive Outcomes: Peabody Picture Vocabulary Test (PPVT) scores are used as a measure of receptive language ability. Tests were administered by trained assessors according to standard protocol.

Child Social Outcomes: The Social Skills Rating System (SSRS) questionnaire was completed by parents of each child. Standardized scores are used in these analyses.

Family Variables: Other family variables used in the analyses came from the family interview and included parent's education level, chronic health problems, and family size (# children + # adults).

FRS - R Items and Subscales

Describing Family Resources

Descriptive statistics and histograms were run on all four FRS - R subscales for each ethnic group. Results showed that two of the subscales had high skewness and kurtosis, Basic Needs and Time for Family. Basic Needs was the most skewed for the Caucasian group, indicating that fewer of the Caucasians lacked basic needs than either other groups. The reverse was true for Time for Family, the African American and Hispanic groups scored higher and were more highly skewed than the Caucasian group. Overall, although some of the variables have significant skew and kurtosis, there is enough variability for the FRS - R to be a useful measurement tool.

Table 1. Descriptive Statistics for the FRS - R

Ethnic Variation in Family Resources

Ethnic variations in the FRS - R kindergarten scores were assessed using a multivariate GLM. The four FRS - R subscale scores were used as outcome measures. The subscale scores were regressed on ethnicity controlling for percent of poverty, family size, and parent's education level. These variables were controlled for because there was reason to expect ethnic differences in them.

Results showed that ethnicity accounted for 2.6% of the variance in the FRS - R after controlling for the covariates (See Table 2). However, ethnicity only accounted for a significant amount of variance in the Basic Needs subscale score. Caucasian's average Basic Needs score was higher than that of African American respondents which in turn was higher than the Hispanic respondents. Not surprisingly percent of poverty accounted for 19% of the variance in the Money score. Parent's education level was negatively related to Time for Family, accounting for 2% of the variance.

Table 2. Variance Accounted for in FRS - R by Ethnicity and Covariates

Note: - indicates less than .5%

The Relationship of the FRS - R and Ethnicity to Child Outcomes

Children's kindergarten PPVT scores were regressed on FRS - R scores and child ethnicity. FRS - R scores were entered in model one and child ethnicity was added in model two (see Table 3). The FRS - R subscale scores accounted for 4.6% of the variability in child PPVT scores, Basic Needs was positively related PPVT scores and Time for Family was negatively related the outcome. The addition of child ethnicity to the model increased the amount of variability accounted for by the model to 21.7% and decreased the contribution of the FRS - R. Caucasian children scored the highest, followed by African American children, and then Hispanic children. Although the contribution of the individual FRS - R scores were reduced, note that together they still account for nearly 4% of the variance in PPVT scores.

Next parents ratings of children's social skills were regressed on the FRS - R and child ethnicity (See Table 4). Results of the first model show that FRS - R subscale scores account for 5.6% of the variance in the SSRS, with Time for Self being the single unique predictor of social skills. The addition of child ethnicity to the model increased the variance accounted for by nearly 2%, with Hispanic children scoring the highest, followed by African American and then Caucasian children.

Table 3. PPVT Scores Regressed on FRS - R and Ethnicity

Table 4. SSRS Scores Regressed on FRS - R and Ethnicity

Conclusions

These analyses indicate that family resources can be assessed from a subjective perspective in ways that reveal important cross-cultural differences, as well as predict aspects of children’s academic and social competency. Four primary subscales were validated for the Family Resource Scale: the adequacy of Basic Needs, Time for Self, Time for Family, and Money. The results further indicate that subjective appraisal yields a somewhat different, and perhaps more sensitive portrayal of important influences on children than the traditional reliance on income per se. Clear ethnic differences in report of the adequacy of family resources emerged, even within a predominately poverty sample. The use of this revised version of the FRS allows researchers and practitioners to evaluate some of the different needs and strengths of families that may be overlooked by relying on traditional demographic marker variables.

Analyses of child outcomes provide additional support for this conclusion. The four FRS subscales predicted child outcomes differentially, and added information above and beyond that contained in the income and ethnicity variables. Particularly salient was the finding that children’s social skills were strongly related to parental ratings of the adequacy of their Time.

Dunst, C. J., & Leet, H. E. (1987). Measuring the adequacy of resources in households with young children. Child: Care, Health, and Development, 13, 111-125.

Ramey, S. L., Ramey, C. T., & Phillips, M. M. (1996). Head start children’s entry into public school: An interim report on the National Head Start-Public School Early Childhood Transition Demonstration Study. Prepared for: U.S. Department of Health & Human Services. Washington DC.