The new research, published in the Journal of Child and Family Studies, evaluated the Catalan parent-rated version of the questionnaire in a community sample of 937 children aged 6 to 12, drawn from primary schools across the Spanish region. The team, led by Sebastián Fernández-Ortiz of the Universitat Autònoma de Barcelona, set out to answer a deceptively simple question: does the instrument actually measure the same thing in the same way for boys and girls, and for children of different ages? The answer, it turns out, is yes for gender, but no for age.

The Alabama Parenting Questionnaire was developed in the 1990s to capture the parenting practices most strongly linked to disruptive childhood behavior. Its 42-item parent version, built on decades of evidence from developmental and clinical psychology, targets five theoretically distinct dimensions: Involvement, Positive Parenting, Poor Monitoring/Supervision, Inconsistent Discipline, and Corporal Punishment. Higher scores on the first two indicate effective parenting, while higher scores on the last three flag practices that raise the risk of behavioral and emotional problems. Parents rate how often specific behaviors occur in the household on a five-point scale ranging from never to always.

Notably, the original questionnaire was never subjected to exploratory factor analysis because its initial sample was too small, leaving later researchers to verify its structure empirically. Decades of adaptations across dozens of countries, from Germany and Poland to Chile, Qatar, Russia and Uganda, have produced strikingly inconsistent findings. Some studies recovered the original five factors, others found four, three, or even two, and confirmatory analyses have reported solutions ranging from three to six factors. This heterogeneity has real consequences: when the structure is unstable, scores cannot be compared fairly across studies, cultures, or developmental stages.

To test the five-factor structure rigorously, the Catalan team used confirmatory factor analysis with weighted least squares estimation, treating the 35 core items as observed variables and the five parenting dimensions as latent, intercorrelated constructs. The model showed what the researchers describe as a split verdict. Absolute fit indices, the root mean square error of approximation and the standardized root mean square residual, indicated an acceptable fit, but incremental fit indices, the comparative fit index and Tucker-Lewis Index, fell below the conventional cutoff of 0.95, landing at 0.86 and 0.84 respectively. The authors explain this apparent contradiction with reference to a well-documented statistical phenomenon: incremental fit indices tend to deteriorate in models with many degrees of freedom, and this model carried 550 of them. In such cases, they argue, absolute indices are the more informative gauge, though the five-factor solution should still be interpreted with care.

Most of the items behaved as theory predicted. Thirty-two of the 35 loaded significantly on their assigned factors, with standardized loadings ranging from 0.19 to 0.86. Only three items, one each from the Involvement, Positive Parenting and Monitoring scales, fell below the 0.30 retention threshold, and the team kept them to preserve the original structure rather than massage the statistics. The correlations among the latent factors followed the expected pattern: the two positive parenting dimensions were positively associated with each other, while the three negative dimensions correlated positively with one another and negatively with the positive scales.

The measurement invariance analyses produced the study’s most consequential findings. Comparing boys and girls, the researchers found configural, partial metric, partial scalar and strict invariance, meaning the questionnaire’s parameters were comparable across gender, allowing legitimate comparisons of parenting scores between the sexes. That comparison turned out to be substantive: parents reported more poor monitoring, more inconsistent discipline and more corporal punishment directed at boys than at girls, with small effect sizes.

Age was a different story. The invariance analyses failed across the three grade groups, spanning second, fourth and sixth grade, meaning the questionnaire does not measure the constructs equivalently across this developmental window. In practical terms, a difference in, say, monitoring scores between a 7-year-old and an 11-year-old cannot be confidently interpreted as a real difference in parenting, because the items themselves may mean different things at different ages. The descriptive data hinted at why: monitoring scores rose progressively across grades, while involvement and positive parenting declined as children matured and gained autonomy, precisely the developmental shifts that make certain items, such as expecting a young child to leave a note about their whereabouts, developmentally inappropriate.

Reliability added a further caution. Internal consistency, assessed with both Cronbach’s alpha and McDonald’s omega, was marginal for Involvement, Positive Parenting and Inconsistent Discipline, and poor for Poor Monitoring/Supervision and Corporal Punishment. The corporal punishment scale is only three items long, and it addresses three distinct forms of physical punishment that parents tend not to use together, which naturally weakens the correlations among its items. For the monitoring scale, the problem appears to be low base rates: many monitoring behaviors simply do not occur often in middle childhood, when children have little independence, producing floor effects and restricted variability. The researchers stress that this reflects the developmental stage of the sample rather than a flaw in the monitoring construct itself, which becomes far more salient in adolescence.

Yet despite these psychometric caveats, the questionnaire did what it was built to do in one crucial respect: it predicted children’s adjustment. The team examined convergent and divergent validity by relating parenting scores to mental health and social outcomes reported by both parents and teachers. Parents filled out the Strengths and Difficulties Questionnaire, a screening tool covering emotional symptoms, conduct problems, hyperactivity, peer problems and prosocial behavior, while teachers completed both that instrument and the School Social Behavior Scales, which measures peer relations, self-management and academic behavior.

The pattern of associations was strikingly coherent. Ineffective parenting practices, particularly inconsistent discipline and corporal punishment, correlated positively with internalizing and externalizing difficulties, including conduct problems, hyperactivity and emotional symptoms, across informants. The same ineffective practices correlated negatively with prosocial behavior and, in teacher reports, with every dimension of school social competence, from peer relations to academic behavior. Effective parenting practices, by contrast, were associated with greater prosocial behavior. Regression analyses showed that parenting practices explained only modest proportions of variance, roughly 4 to 18 percent for parent-reported outcomes and less for teacher reports, but the direction and specificity of the links matched theory across the board.

For the authors, the takeaway is not that the Alabama Parenting Questionnaire should be abandoned, but that it should be used with developmental awareness. The partial gender invariance supports using it to study gender-related differences in parenting and their ties to child adjustment. The failure of age invariance, however, means that age-group comparisons are on shaky ground, because there is no guarantee that observed differences reflect anything substantive. Clinicians and school psychologists using the instrument for screening or program evaluation should weigh a child’s developmental stage when interpreting scale scores, particularly for younger children, where certain parenting behaviors may be poorly captured.

The study’s design lent it unusual strength. Rather than recruiting a convenience sample, the team employed a stratified multistage sampling scheme covering Catalonia’s four provinces, classifying territories by population size and schools by type to ensure territorial and sociodemographic representativeness. Nearly all recruited families, 99.5 percent, provided usable data, and the researchers verified reliability by telephoning 20 percent of participating households. Most respondents were mothers, at 77.2 percent, and the children were evenly split between boys and girls.

The authors also acknowledge limitations that temper the conclusions. The cross-sectional design cannot establish whether parenting practices cause children’s difficulties or vice versa, and the focus on school-age children says nothing about adolescence, where monitoring becomes a central construct. Only the parent report version was examined, leaving child and teacher perspectives on parenting untested. And the modest regression results suggest that parenting, however measured, is one influence among many on child adjustment.

The broader message resonates well beyond Catalonia. As the questionnaire continues to be adapted into new languages and applied in intervention programs worldwide, this study adds to a growing body of evidence that measurement instruments cannot simply be assumed to travel unchanged across ages, cultures and contexts. The Catalan findings align with earlier work in Portugal, Romania and Greece, all of which found partial or absent invariance across developmental stages. What remains clear, and what the validity data confirm forcefully, is the substantive science underneath the statistics: inconsistent discipline and corporal punishment go hand in hand with children’s emotional and behavioral problems, while engaged, positive parenting travels with prosocial, socially competent children. The tool that measures those truths, the study suggests, now needs refinement to keep pace with the children it is meant to describe.