M. Lee Van Horn

Jeffery M. Bellis

Scott W. Snyder

Civitan International Research Center

The University of Alabama at Birmingham

Paper presented at the biannual conference of the Society for Research in Child Development, April 15^th 1999, Albuquerque, New Mexico. Correspondence should be addressed to M. Lee Van Horn, Civitan International Research Center – UAB, 1719 6^th Ave. S. Rm 235c, Birmingham AL 35294. Or by E-mail to Vanhorn@uab.edu.

Availability of resources is an important construct within the context of child and family outcome research. An alternative to using income is to have family members rate the adequacy of available resources. In contrast to measures such as income, perception of resources has been shown to be more likely to drive behavior and perceived well being (Dunst, Leet, & Trivette, 1988). The current study refines an existing subjective measure of family resources.

The Family Resource Scale (FRS) was created as a clinical tool to assist practitioners developing treatment and intervention plans (Dunst & Leet, 1986). The measure consists of thirty items rated using a five point Likert-type scale anchored by "not at all adequate" to "almost always adequate." Measurement properties of the FRS were originally reported on a sample of 45 mothers (Dunst, & Leet, 1986), however, these analyses have been called into question based on sample size and methodology (McGrew, Gilman, & Johnson, 1992).

Sample: FRS data was collected within the context of a thirty-site, longitudinal study of former Head Start children and their families. The study design included two cohorts of families of kindergartners that were followed through the third grade. The FRS was completed by a parent of the Head Start child. Cohort I kindergarten (N = 3418) and third grade (N = 2902) data comprised exploratory samples that were analyzed separately to avoid the potential impact of dependency. Cohort II kindergarten (N = 3867) and third grade (N = 3318) data were used as confirmatory samples. Listwise deletion was used to define membership for each sample (Cohort I: kindergarten (N = 2321), third grade (N = 1883); Cohort II: kindergarten (N = 2688), third grade (N = 2101).

Exploratory Analyses: Exploratory principal component analysis was used to establish a theoretically relevant measurement model. Two procedures (Kaiser’s eigenvalue > 1, Cattel’s Scree Plot) suggested that three components be retained. A parallel series of principal component analyses (using polychoric correlation matrices and varimax rotation) were performed on the Cohort I kindergarten and third grade samples. Non-salient and complex items were sequentially dropped resulting in a three component solution consisting of 22 items and accounting for 56% and 65% of the variance in the kindergarten and third grade samples, respectively (See Table 1). The three components corresponded closely to three hypothesized theoretical constructs: Basic Needs, Time, and Money.

Confirmatory Analyses: Using LISREL 8 with a CSM estimation procedure (weighted least squares estimation of the polychoric matrices (Kaplan, 1990)), the three correlated factor model was imposed on the two Cohort II samples. Fit indices were adequate (see Table 2), suggesting that the three correlated factor model fit these data reasonably well. Additionally, these fit indices reflected superior model fit in contrast to competing models. As an additional test of criterion validity the predictive value of the subscales was examined, and they were found to differentially predict key child and family variables.

Construct Validity. Construct validity was assessed in two ways. First of all the correlations between the FRS subscales and income (measured by percent of federal poverty guidelines) were assessed (See Table 3). Percent of poverty was moderately correlated with the Money subscale, but only weekly correlated with the Time or Basic Needs subscales.

Next, the relationship between the FRS subscales and child’s scores on cognitive and social outcomes were assessed. Regression models controlling for Site, Ethnicity, and Poverty found that the Money subscale was significantly related to child’s cognitive assessments (Woodcock Johnson Reading and Math scores). The Time subscale was significantly related to child’s ratings on the Social Skills Rating System.

In these analyses the original FRS was reduced to 22 items and 3 highly intuitive subscales that differentially relate to key outcome variables. Researchers are provided with a tool to assess familys’ perceptions of their needs that has been validated on a large sample of at-risk children. Replication of this measurement model and further research assessing the relation of these subscales to important outcome variables is warranted, however, work to date suggests that the FRS will prove a useful research instrument.

Citations

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

Dunst, C. J., Leet, H. E., & Trivette, C. M. (1988). Family resources, personal well-being, and early intervention. The Journal of Special Education, 22, 108-116.

Kaplan, D. (1990). Evaluationg and modifying covariance structure models: A review and recommendation. Multivariate Behavioral Research, 25, 137-155.

McGrew, K. S., Gilman, C. J., & Johnson, S. (1992). A review of scales to assess family needs. Journal of Psychoeducational Assessment, 10, 4-26.

Table 1. FRS items and factors on which they load.

Table 2. Results of the Confirmatory Factor Analyses.

* RESEA is Root Mean Square Error of Approximation, GFI is Goodness of Fit Index, and CFI is Comparative Fit Index

Table 3. Correlations between FRS subscales and poverty level in third grade.

All correlations are significant p<.001