High-Stakes Testing Case Study: A Latent Variable Approach for Assessing Measurement and Prediction Invariance.

The existence of differences in prediction systems involving test scores across demographic groups continues to be a thorny and unresolved scientific, professional, and societal concern. Our case study uses a two-stage least squares (2SLS) estimator to jointly assess measurement invariance and prediction invariance in high-stakes testing. So, we examined differences across groups based on latent as opposed to observed scores with data for 176 colleges and universities from The College Board. Results showed that evidence regarding measurement invariance was rejected for the SAT mathematics (SAT-M) subtest at the 0.01 level for 74.5% and 29.9% of cohorts for Black versus White and Hispanic versus White comparisons, respectively. Also, on average, Black students with the same standing on a common factor had observed SAT-M scores that were nearly a third of a standard deviation lower than for comparable Whites. We also found evidence that group differences in SAT-M measurement intercepts may partly explain the well-known finding of observed differences in prediction intercepts. Additionally, results provided evidence that nearly a quarter of the statistically significant observed intercept differences were not statistically significant at the 0.05 level once predictor measurement error was accounted for using the 2SLS procedure. Our joint measurement and prediction invariance approach based on latent scores opens the door to a new high-stakes testing research agenda whose goal is to not simply assess whether observed group-based differences exist and the size and direction of such differences. Rather, the goal of this research agenda is to assess the causal chain starting with underlying theoretical mechanisms (e.g., contextual factors, differences in latent predictor scores) that affect the size and direction of any observed differences.

Attendance Patterns and Links to Non-Response on Child Report of Internalizing among Mexican-Americans Randomized to a Universal Preventive Intervention.

We examined attendance trajectory profiles among 335 Mexican-American families participating in an 11-week universal intervention to explore if heterogeneity in attendance and thus dosage was associated with intervention response, defined as pre-to-2-year post (T2) reductions in child report of internalizing symptoms. We estimated trajectories accounting for the influence of baseline covariates, selected based on the Health Belief Model (HBM) and Latino family research, to understand covariate associations with trajectories. Results supported six attendance trajectory groups: non-attenders (NA), early dropouts-low internalizing (EDO-LI), early dropouts-high internalizing (EDO-HI), mid-program dropouts (MPDO), sustained attenders-low internalizing (SA-LI), and sustained attenders-high internalizing (SA-HI). All groups except EDO-HI showed significant pre-to-post change on child report of internalizing; however, trajectory groups reflecting more attendance did not have greater pre-to-post change. Nonetheless, child report of internalizing differentiated two subgroups of sustained attenders and two subgroups of early dropouts. These results suggest heterogeneity among families with similar patterns of attendance and highlight the importance of modeling this heterogeneity. Although life stress was a barrier to participation, there was minimal support for the HBM. Cultural influences, acculturation, and familism, played a more prominent role in distinguishing trajectories. As expected, the EDO-HI group was less acculturated than both sustained attender groups and reported weaker familism values than the SA-HI group. However, unexpectedly, the SA-LI group had lower familism than the EDO-LI group. The results suggest that the influence of culture on participation is nuanced and may depend on child symptomatology.

Testing measurement invariance in longitudinal data with ordered-categorical measures.

A goal of developmental research is to examine individual changes in constructs over time. The accuracy of the models answering such research questions hinges on the assumption of longitudinal measurement invariance: The repeatedly measured variables need to represent the same construct in the same metric over time. Measurement invariance can be studied through factor models examining the relations between the observed indicators and the latent constructs. In longitudinal research, ordered-categorical indicators such as self- or observer-report Likert scales are commonly used, and these measures often do not approximate continuous normal distributions. The present didactic article extends previous work on measurement invariance to the longitudinal case for ordered-categorical indicators. We address a number of problems that commonly arise in testing measurement invariance with longitudinal data, including model identification and interpretation, sparse data, missing data, and estimation issues. We also develop a procedure and associated R program for gauging the practical significance of the violations of invariance. We illustrate these issues with an empirical example using a subscale from the Mexican American Cultural Values scale. Finally, we provide comparisons of the current capabilities of 3 major latent variable programs (lavaan, Mplus, OpenMx) and computer scripts for addressing longitudinal measurement invariance. (PsycINFO Database Record

Family Influences on Mexican American Adolescents' Romantic Relationships: Moderation by Gender and Culture.

This study examined prospective associations between the family context and adolescents' romantic relationships as moderated by adolescents' gender and culture among Mexican American families (N = 189). Adolescents at Time 1 (early adolescence) were on average 12.29 years of age (SD = .50) and 54% female. Mothers and fathers reported on family structure and dynamics during early adolescence, and youth reported on their romantic relationship involvement and quality during middle and late adolescence. Results from path analyses indicated that family structure and dynamics (supportive parenting, consistent discipline, parent-adolescent, and interparental conflict) were associated with adolescents' romantic involvement and quality, with differences by adolescents' gender and culture. Findings highlight Mexican American family contexts that contribute uniquely to adolescents' romantic relationships.

Mexican-Origin Youth Participation in Extracurricular Activities: Predicting Trajectories of Involvement from 7th to 12th Grade.

The potential benefits of participation in extracurricular activities may be especially important for youth who are at risk for academic underachievement, such as low income Mexican-origin youth in the U.S. To advance understanding of factors that drive participation for this population, this study examined Mexican-origin youth's trajectories of participation in extracurricular activities across Grades 7-12 and tested theoretically-derived predictors of these trajectories. Participants were 178 adolescents (53.9 % Female, Mage = 12.28) and their mothers who separately completed in-home interviews. Youth reported the frequency of their participation across a range of extracurricular activities. Latent growth curve models of overall extracurricular activities participation, sports participation, and fine arts participation were individually estimated via structural equation modeling. The findings demonstrated developmental declines in overall participation and in sports participation. For fine arts, declines in participation in middle school were followed by subsequent increases during high school (a curvilinear pattern). Motivationally-salient predictors of participation trajectories included youth's traditional cultural values orientation (sports), the mothers' educational aspirations for the youth (sports, fine arts, overall activity), and youth gender (sports, fine arts). Overall, the results suggest variability in participation trajectories based on program type, and highlight the need for additional research to enhance our understanding of the impact of culturally-relevant predictors on participation over time.

The Role of Social and Personal Identities in Self-Esteem Among Ethnic Minority College Students.

This study examined the latent personal-social identity profiles that emerged from simultaneous consideration of ethnic, national (United States), and personal identities among ethnic minority college students (N = 3,009) as well as how personal and social identities are jointly associated with self-esteem. Results indicated that the structure of personal-social identity profiles significantly differed across ethnicity, but also indicated some commonalities. The study identified three profiles among Blacks, four among Asian Americans, and two among Latinos. Some personal-social identity profiles were common across multiple ethnic groups, but others were unique within one specific ethnic group. Overall, the profiles indicated important associations between ethnic identity, U.S. identity, and personal identity. These profiles were linked with self-esteem such that individuals who reported high levels of multiple social and personal identities had the highest self-esteem compared to other profiles.

Participation patterns among Mexican-American parents enrolled in a universal intervention and their association with child externalizing outcomes.

This study used growth mixture modeling to examine attendance trajectories among 292 Mexican-American primary female caregivers enrolled in a universal preventive intervention and the effects of health beliefs, participation intentions, cultural influences, and intervention group cohesion on trajectory group membership as well as trajectory group differences on a distal outcome, immediate posttest teacher report of child externalizing (T2). Results supported four trajectory groups-early terminators (ET), mid-program terminators (MPT), low-risk persistent attenders (LRPA), and high-risk persistent attenders (HRPA). Compared with LRPAs, caregivers classified as HRPAs had weaker familism values, less parenting efficacy, and higher externalizing children with lower GPAs. Caregivers in the two persistent attender groups reported strong group cohesion and providers rated these caregivers as having strong participation intentions. Children of caregivers in the LRPA group had the lowest T2 child externalizing. Children of caregivers in the MPT group had lower T2 externalizing than did those of the ET group, suggesting partial intervention dosage can benefit families. Despite high levels of attendance, children of caregivers in the HRPA had the highest T2 externalizing, suggesting this high-risk group needed either more intensive services or a longer period for parents to implement program skills to evidence change in child externalizing.

Parenting intervention effects on parental depressive symptoms: examining the role of parenting and child behavior.

Parental depression is a major risk factor in child development. Growing research suggests parenting programs can positively impact parental depressive symptoms, although the specific mechanisms that explain these effects are unknown. The current study examined parenting mediated effects of a parenting program on mothers' and fathers' depressive symptoms, as well as the role of child behavior in linking parenting to reductions in depressive symptoms. The study samples included 494 mothers and 288 fathers of Mexican origin adolescents who participated in a randomized trial of the Bridges to High School Program/Proyecto Puentes a la Secundaria, a universal prevention and promotion intervention that included parent training but did not directly target parental depressive symptoms. Parenting mediator models tested program effects on parental depressive symptoms through changes in harsh and supportive parenting. Results showed a significant indirect intervention effect on maternal depressive symptoms through changes in mothers' harsh parenting. Next, child behavior models revealed a partial mediation effect of harsh parenting and a full mediation effect of supportive parenting on maternal depressive symptoms through mothers' reports of child externalizing symptoms. Indirect effects of fathers' harsh and supportive parenting on paternal depressive symptoms were also found through fathers' reports of child behavior.

Long-term effects of a universal family intervention: mediation through parent-adolescent conflict.

This randomized trial of a family-focused preventive intervention for Mexican American middle schoolers examined internalizing, externalizing, and substance use outcomes in late adolescence, 5 years after completing the intervention. Parent-adolescent conflict was tested as a mediator of these effects. The role of parent and adolescent acculturation in these pathways was also examined. There were 494 seventh-grade adolescents and their primary female caregivers randomized to receive either a 9-week multicomponent intervention or a brief workshop control group. Assessments were conducted at pretest, 2-year follow-up (9th grade), and 5-year follow-up (when most participants were in the 12th grade). The Bridges program significantly reduced mother-adolescent conflict measured in the 9th grade, with conflict mediating program effects on internalizing and externalizing symptoms, adolescent substance use, and diagnosed internalizing disorder in late adolescence. Mother and child acculturation were both significantly predictive of late adolescence outcomes. Contrary to hypotheses, neither mother nor child acculturation emerged as a significant predictor of mother-adolescent conflict, and the interaction of mother and adolescent acculturation was similarly not related to mother-adolescent conflict. Intervention effects were largely consistent across different levels of acculturation. These findings provide support for the efficacy of family-focused intervention during early adolescence, both in reducing mental health problems and substance use in the long term and in impacting parent-adolescent conflict processes that appear to play an important role in the development of later adjustment problems.

School engagement mediates long-term prevention effects for Mexican American adolescents.

This 5-year follow-up of a randomized clinical trial evaluated the efficacy of a family-focused intervention delivered in middle school to increase school engagement following transition to high school (2 years post-test), and also evaluated mediated effects through school engagement on multiple problem outcomes in late adolescence (5 years post-test). The study sample included 516 Mexican American adolescents who participated in a randomized trial of the Bridges to High School Program (Bridges/Puentes). Path models representing the direct and indirect effects of the program on four outcome variables were evaluated using school engagement measured in the 9th grade as a mediator. The program significantly increased school engagement, with school engagement mediating intervention effects on internalizing symptoms, adolescent substance use, and school dropout in late adolescence when most adolescents were in the 12th grade. Effects on substance use were stronger for youth at higher risk based on pretest report of substance use initiation. There were no direct or indirect intervention effects on externalizing symptoms. Findings support that school engagement is an important prevention target for Mexican American adolescents.

Religion and intergroup conflict: findings from the Global Group Relations Project.

How might religion shape intergroup conflict? We tested whether religious infusion-the extent to which religious rituals and discourse permeate the everyday activities of groups and their members-moderated the effects of two factors known to increase intergroup conflict: competition for limited resources and incompatibility of values held by potentially conflicting groups. We used data from the Global Group Relations Project to investigate 194 groups (e.g., ethnic, religious, national) at 97 sites around the world. When religion was infused in group life, groups were especially prejudiced against those groups that held incompatible values, and they were likely to discriminate against such groups. Moreover, whereas disadvantaged groups with low levels of religious infusion typically avoided directing aggression against their resource-rich and powerful counterparts, disadvantaged groups with high levels of religious infusion directed significant aggression against them-despite the significant tangible costs to the disadvantaged groups potentially posed by enacting such aggression. This research suggests mechanisms through which religion may increase intergroup conflict and introduces an innovative method for performing nuanced, cross-societal research.

Using the Bollen-Stine Bootstrapping Method for Evaluating Approximate Fit Indices.

Accepting that a model will not exactly fit any empirical data, global approximate fit indices quantify the degree of misfit. Recent research (Chen et al., 2008) has shown that using fixed conventional cut-points for approximate fit indices can lead to decision errors. Instead of using fixed cut-points for evaluating approximate fit indices, this study focuses on the meaning of approximate fit and introduces a new method to evaluate approximate fit indices. Millsap (2012) introduced a simulation-based method to evaluate approximate fit indices. A limitation of Millsap's work was that a rather strong assumption of multivariate normality was implied in generating simulation data. In this study, the Bollen-Stine bootstrapping procedure (Bollen & Stine, 1993) is proposed to supplement the former study. When data are non-normal, the conclusions derived from Millsap's (2012) simulation method and the Bollen-Stine method can differ. Examples are given to illustrate the use of the Bollen-Stine bootstrapping procedure for evaluating RMSEA. Comparisons are made with the simulation method. The results are discussed, and suggestions are given for the use of proposed method.

Father enrollment and participation in a parenting intervention: personal and contextual predictors.

Fathers are an important, though often underrepresented, population in family interventions. Notably, the inclusion of ethnic minority fathers is particularly scarce. An understanding of factors that promote and hinder father participation may suggest strategies by which to increase fathers' presence in studies designed to engage the family unit. The current research examined Mexican origin (MO) fathers' involvement in a family-focused intervention study. Participants included 495 fathers from eligible two-parent MO families with an adolescent child. Individual, familial, and culturally relevant predictors based on father, mother, and/or child report data were collected through pretest interviews and included in two separate logistic regression analyses that predicted the following: (1) father enrollment in the study and (2) father participation in the intervention. Results indicated that higher levels of maternal education and lower levels of economic stress and interparental conflict were associated with increased father enrollment in the study. Rates of father participation in the intervention were higher among families characterized by lower levels of interparental conflict, economic stress, and Spanish language use. Results highlight the relevancy of the familial and environmental context to MO fathers' research participation decisions. These findings as well as their implications for future research and practice are discussed.

Maternal Warmth Moderates the Link between Harsh Discipline and Later Externalizing Behaviors for Mexican American Adolescents.

OBJECTIVE: This study examined maternal warmth as a moderator of the relation between harsh discipline practices and adolescent externalizing problems 1year later in low-income, Mexican American families. DESIGN: Participants were 189 adolescents and their mothers who comprised the control group of a longitudinal intervention program. RESULTS: Maternal warmth protected adolescents from the negative effects of harsh discipline such that, at higher levels of maternal warmth, there was no relation between harsh discipline and externalizing problems after controlling for baseline levels of externalizing problems and other covariates. At lower levels of maternal warmth, there was a positive relation between harsh discipline practices and later externalizing problems. CONCLUSIONS: To understand the role of harsh discipline in the development of Mexican American youth outcomes, researchers must consider contextual variables that may affect youths' perceptions of their parents' behavior such as maternal warmth.

Fifteen-year follow-up of a randomized trial of a preventive intervention for divorced families: effects on mental health and substance use outcomes in young adulthood.

OBJECTIVE: This 15-year follow-up assessed the effects of a preventive intervention for divorced families, the New Beginnings Program (NBP), versus a literature control condition (LC). METHOD: Mothers and their 9- to 12-year-olds (N = 240 families) participated in the trial. Young adults (YAs) reported on their mental health and substance-related disorders, mental health and substance use problems, and substance use. Mothers reported on YA's mental health and substance use problems. Disorders were assessed over the past 9 years (since previous follow-up) and 15 years (since program entry). Alcohol and marijuana use, other substance use and polydrug use, and mental health problems and substance use problems were assessed over the past month, past year, and past 6 months, respectively. RESULTS: YAs in NBP had a lower incidence of internalizing disorders in the past 9 years (7.55% vs. 24.4%; odds ratio [OR] = .26) and 15 years (15.52% vs. 34.62%; OR = .34) and had a slower rate of onset of internalizing symptoms associated with disorder in the past 9 years (hazard ratio [HR] = .28) and 15 years (HR = .46). NBP males had a lower number of substance-related disorders in the past 9 years (d = 0.40), less polydrug (d = 0.55) and other drug use (d = 0.61) in the past year, and fewer substance use problems (d = 0.50) in the past 6 months than LC males. NBP females used more alcohol in the past month (d = 0.44) than LC females. CONCLUSIONS: NBP reduced the incidence of internalizing disorders for females and males and substance-related disorders and substance use for males.

Caregiver responsiveness to the family bereavement program: what predicts responsiveness? What does responsiveness predict?

The study developed a multidimensional measure to assess participant responsiveness to a preventive intervention and applied this measure to study how participant baseline characteristics predict responsiveness and how responsiveness predicts program outcomes. The study was conducted with caregivers who participated in the parenting-focused component of the Family Bereavement Program (FBP), a prevention program for families that have experienced parental death. The sample consisted of 89 caregivers assigned to the intervention condition in the efficacy trial of the FBP. Positive parenting, caregiver depression, and child externalizing problems at baseline were found to predict caregivers' use of program skills outside the group, and more child internalizing problems predicted more positive perceptions of the group environment. Higher levels of skill use during the program predicted increased positive parenting at the 11-month follow-up, whereas positive perceptions of the group environment predicted decreased caregiver depressive symptoms at follow-up. Caregiver skill use mediated the relation between baseline positive parenting and improvements in positive parenting at 11-month follow-up, and skill use and perceived group environment mediated changes in caregiver depression from baseline to 11-month follow-up.

Statistical Power for a Simultaneous Test of Factorial and Predictive Invariance.

A common finding in studies of differential prediction across groups is that although regression slopes are the same or similar across groups, group differences exist in regression intercepts. Building on earlier work by Birnbaum (1979) , Millsap (1998) presented an invariant factor model that would explain such intercept differences as arising due to unique variance or measurement error in the predictors and criterion. The model can be tested even when only a single predictor (p = 1) and criterion are present, but Millsap (1998) only illustrated the use of the test in very large samples. The present study investigates the Type I error and power performance of the test of fit across a range of sample sizes and effect sizes in simulated data, using an alternative model that includes group differences in measurement intercepts. Surprising results showed little power for the chi-square test in the p = 1 case but good power for a diagnostic indicator based on negative unique variance estimates. The p = 2 case showed better power in modest samples. A real example from an application on reading disabilities in Share and Silva (2003) is given. The implications of the results for studies of differential prediction involving group differences in regression intercepts are discussed.

Randomized trial of a broad preventive intervention for Mexican American adolescents.

OBJECTIVE: This randomized trial of a family-focused preventive intervention for Mexican American (MA) adolescents evaluated intervention effects on adolescent substance use, internalizing and externalizing symptoms, and school discipline and grade records in 8th grade, 1 year after completion of the intervention. The study also examined hypothesized mediators and moderators of intervention effects. METHOD: Stratified by language of program delivery (English vs. Spanish), the trial included a sample of 516 MA adolescents (50.8% female; M = 12.3 years, SD = 0.54) and at least one caregiver that were randomized to receive a low-dosage control group workshop or the 9-week group intervention that included parenting, adolescent coping, and conjoint family sessions. RESULTS: Positive program effects were found on all 5 outcomes at 1-year posttest but varied depending on whether adolescents, parents, or teachers reported on the outcome. Intervention effects were mediated by posttest changes in effective parenting, adolescent coping efficacy, adolescent school engagement, and family cohesion. The majority of intervention effects were moderated by language, with a larger number of significant effects for families who participated in Spanish. Intervention effects also were moderated by baseline levels of mediators and outcomes, with the majority showing stronger effects for families with poorer functioning at baseline. CONCLUSION: Findings not only support the efficacy of the intervention to decrease multiple problem outcomes for MA adolescents but also demonstrate differential effects for parents and adolescents receiving the intervention in Spanish vs. English, and depending on their baseline levels of functioning.

Family stress and coping for Mexican origin adolescents.

Family-related stressors pose special challenges for adolescents of Mexican origin, given traditional cultural norms that compel youths to get involved with family problems despite their limited ability to effect change. The current study examines the prospective effects of coping strategies (i.e., active, distraction, avoidance, support-seeking, and religious coping) on psychological symptoms in the context of family stress with a sample (N = 189) of Mexican Origin adolescents (11-14). Hypotheses on the limits of coping were partially supported. Stress-coping interaction effects were further moderated by gender. Stress-buffering effect of active coping for internalizing symptoms was only found for girls and only at low levels of family stress for boys. Support-seeking and distraction coping both increased internalizing symptoms for boys at high levels of family stress.