Response bias shows up in many fields of behavioural and healthcare research where self-reported data are used. We demonstrate how to use stochastic frontier estimation (SFE) to identify response bias and its covariates. In our application to a family intervention, we examine the effects of participant demographics on response bias before and after participation; gender and race/ethnicity are related to magnitude of bias and to changes in bias across time, and bias is lower at post-test than at pre-test. We discuss how SFE may be used to address the problem of 'response shift bias' – that is, a shift in metric from before to after an intervention which is caused by the intervention itself and may lead to underestimates of programme effects.

The purpose of the present study was to examine a common practice in some areas of program evaluation, the retrospective pretest, and to present recommendations regarding its use. The authors review literature to emphasize first, that bias is likely in both prospective and retrospective ratings, and second, that under some circumstances, retrospective pretests may introduce greater bias than traditional pretests. The authors examined data from 100 participants who attended a family program at 15 sites. Results supported the hypothesis that items on which parents were asked to endorse socially desirable parenting behaviors resulted in greater discrepancies between prospective and retrospective ratings. Effect sizes calculated from prospective ratings were comparable with published effect sizes from research trials of the same program. The authors conclude that replacing traditional with retrospective pretests does not eliminate bias. They recommend traditional pretests for examination of program effects and retrospective pretests for examination of subjective experiences of program-related change.

OBJECTIVE. Use of daily controller medications is a critical task in management of persistent asthma. Study aims were to examine (1) the association between child age and extent of daily controller-medication responsibility in a sample aged 4 to 19 years, (2) parent, child, and disease predictors of child daily controller-medication responsibility and overall daily controller-medication adherence, and (3) the association between child daily controller-medication responsibility and overall daily controller-medication adherence. METHODS. We conducted a cross-sectional telephone survey of 351 parents of children who were prescribed daily controller medication. Children's mean age was 10.4 years; 61.5% were male, and 88.1% were white. Parents provided all data, including an estimate of the percentage of child and parent daily controller-medication responsibility. Daily controller-medication adherence was measured as parents' report of percentage of daily doses taken per doses prescribed in a typical week. We used multivariate linear regression to determine associations between parent race/ethnicity, education, income, number of dependents, child age, gender, years since diagnosis, parent perception of symptom severity and control, and dependent variables (child daily controller-medication responsibility and daily controller-medication adherence). We also examined associations between child daily controller-medication responsibility and daily controller-medication adherence. RESULTS. Child daily controller-medication responsibility increased with age. By age 7, children had assumed, on average, almost 20% of daily controller-medication responsibility; by age 11, ∼50%; by age 15, 75%; and by age 19, 100%. In multivariate models, child age and male gender remained significantly associated with child daily controller-medication responsibility, and child's age and parents' race/ethnicity remained significantly associated with daily controller-medication adherence. CONCLUSIONS. Clinicians may need to screen for child daily controller-medication management and include even young children when educating families on the use of asthma medications and other key asthma-management tasks.

Accurate, early screening is a prerequisite for indicated interventions intended to prevent development of externalizing disorders and delinquent behaviors. Using the Fast Track longitudinal sample of 396 children drawn from high-risk environments, the authors varied assumptions about base rates and examined effects of multiple-time-point and multiple-rater screening procedures. The authors also considered the practical import of various levels of screening accuracy in terms of true and false positive rates and their potential costs and benefits. Additional research is needed to determine true costs and benefits of early screening. However, the results indicate that 1st grade single- and multiple-rater screening models effectively predicted externalizing behavior and delinquent outcomes in 4th and 5th grades and that early screening is justified.