NEWS

  /  

April 10, 2007

Study: Race crucial in family research

A new report by University researchers found that measurements of parental competence did not work equally well for white families and ethnic minority families. The flawed measurements, based on correlations between parenting techniques not necessarily found among minority parents, sometimes miss minority parents’ child-rearing strengths.

The researchers found that in a popular model used to evaluate families, cultural background significantly biases measures of parental warmth, rules and routines, and the use of physical punishment.

“It’s not like there was higher or lower discipline or warmth between white and minority families,” said C. Cybele Raver, an associate professor in the Harris School and the study’s lead author, in a telephone interview. The differences are rooted in the relationships between certain variables. In white families, for example, high warmth usually correlates with low use of harsh punishments, but that relationship doesn’t necessarily hold for minority parents. This variation accounted for bias in some of the parenting measurements.

There has been concern that the parenting and child-competence measures normally used in this model may not work equally well across all groups of children, Raver said. “This paper says, ‘Let’s take a step back and check these measures carefully for evidence of cultural bias,’” she said. Published in the January/February 2007 issue of the journal Child Development, the study gathered data from over 21,000 kindergartners across the U.S.

The study also found that cultural background had no influence on a child’s skills and school readiness. “The measurements of competence were working well for all groups of kids. That’s good news,” Raver said.

“We’re capturing some of what everybody does right, but...we may have been missing important strengths in the way ethnic minority parents provide good parenting,” Raver added. “Competent parenting can take a whole bunch of different forms in different communities. We have to broaden our perspective on how to measure it.”

Concerns over bias extend beyond the scientific community. Some policymakers are concerned that measurements of the effectiveness of social programs do not accurately reflect the state of minority children. For school-readiness programs like Head Start, the implications are serious in an age of stringent program testing. “If researchers don’t do a good job, we could find erroneously that a policy doesn’t work because we observed little gain in children’s academic outcomes,” Raver said.

The study was coauthored by Elizabeth T. Gershoff of the University of Michigan and J. Lawrence Aber of New York University.