Assessing parenting in racially and ethnically diverse families: A lack of measurement equivalence.

Abstract

The present study explored measurement invariance of the Multidimensional Assessment of Parenting Scale (MAPS; Parent & Forehand, 2017) across White, Hispanic, Black, and Asian American parents. Participants included 2,734 parents, 58% of whom were mothers. On average, parents were 36.32 years old (SD = 9.54); the parent sample was 66.9% White non-Hispanic, 10.1% Black, 5.3% Asian, and 17.7% Hispanic regardless of race. Child ages ranged from 3 to 17 years (M = 9.84, SD = 3.71), and 58% were identified as male. Parents completed a demographics questionnaire about themselves and their target child, and the 34-item MAPS. We explored measurement equivalence of the MAPS Broadband Positive and Negative parenting scales using item response theory to identify differential item functioning (DIF). Univariate analyses for Positive and Negative Parenting showed reliability was excellent. Twelve items assessing negative aspects of parenting exhibited bias by race/ethnicity. Specifically, when comparing racial and ethnic groups, three items had nonuniform DIF comparing Black and Asian participants, two items had nonuniform DIF comparing Black and Hispanic participants, and one item showed nonuniform DIF comparing Asian and Hispanic participants. When looking at Positive Parenting, no items showed evidence of DIF. Results from the present study suggest broadband Positive Parenting can be compared across ethnoracial groups, while findings raise concern about assessing Negative Parenting items when examining invariance across race and ethnicity. Findings from the present study imply that racial and ethnic comparisons are potentially invalid. These findings offer guidance for improving parenting assessment for racially/ethnically diverse populations.

Psychometrics of the Children’s Report of Parental Behavior Inventory among caregivers and youth with HIV.

“…Seven out of the 10 items were related to negative parenting practices. This is consistent with previous research showing psychometric challenges with measuring negative parenting practices (Rodriguez, Cadet, et al, 2023). It is worth noting that, compared with positive parenting practices, caregivers may have greater difficulty reflecting on their own negative parenting behaviors.…”

“…A decade ago, Hurley et al (2014) conducted a systematic review summarizing the state of the evidence of psychometric studies related to parenting; this review classified the state of the evidence as “dismal.” Since Hurley’s et al (2014) review, several studies on the psychometrics of parenting scales have emerged, particularly focusing on measurement invariance/equivalence of parenting scales when used in multicultural settings in the United States (Lindhiem et al, 2019; Rodriguez, Cadet, et al, 2023; Rodriguez, La Barrie, et al, 2023; Shaffer et al, 2022). Unfortunately, this growing evidence of the psychometrics of parenting scales has not extended to countries in the Global North, rendering the validity of the research findings derived from parenting measures largely unclear in these settings when applied cross-culturally.…”

Uncovering measurement bias: Differential item functioning by income in the Couples Satisfaction Index.

“…Detailed item-level analyses, such as those offered with IRT, would strengthen the measurement of family constructs and contribute to a more robust science. Moreover, qualitative methods would be useful at the item development stage and assist in pinpointing exact sources of measurement bias (Rodriguez et al, 2023). Further, quantitative analyses examining the conceptualization of each relationship satisfaction item could illuminate other useful items for measurement within diverse populations (e.g., Fehr, 1999).…”