Nina C. Martin

The design and analysis of longitudinal studies of development and psychopathology in context: Statistical models and methodological recommendations

The utility and flexibility of recent advances in statistical methods for the quantitative analysis of developmental data--in particular, the methods of individual growth modeling and survival analysis--are unquestioned by methodologists, but have yet to have a major impact on empirical research within the field of developmental psychopathology and elsewhere. In this paper, we show how these new methods provide developmental psychopathologists with powerful ways of answering their research questions about systematic changes over time in individual behavior and about the occurrence and timing of life events. In the first section, we present a descriptive overview of each method by illustrating the types of research questions that each method can address, introducing the statistical models, and commenting on methods of model fitting, estimation, and interpretation. In the following three sections, we offer six concrete recommendations for developmental psychopathologists hoping to use these methods. First, we recommend that when designing studies, investigators should increase the number of waves of data they collect and consider the use of accelerated longitudinal designs. Second, we recommend that when selecting measurement strategies, investigators should strive to collect equatable data prospectively on all time-varying measures and should never standardize their measures before analysis. Third, we recommend that when specifying statistical models, researchers should consider a variety of alternative specifications for the time predictor and should test for interactions among predictors, particularly interactions between substantive predictors and time. Our goal throughout is to show that these methods are essential tools for answering questions about life-span developmental processes in both normal and atypical populations and that their proper use will help developmental psychopathologists and others illuminate how important contextual variables contribute to various pathways of development.

Empirical and conceptual problems with longitudinal trait-state models: introducing a trait-state-occasion model.

The latent trait-state-error model (TSE) and the latent state-trait model with autoregression (LST-AR) represent creative structural equation methods for examining the longitudinal structure of psychological constructs. Application of these models has been somewhat limited by empirical or conceptual problems. In the present study, Monte Carlo analysis revealed that TSE models tend to generate improper solutions when N is too small, when waves are too few, and when occasion factor stability is either too large or too small. Mathematical analysis of the LST-AR model revealed its limitation to constructs that become more highly auto-correlated over time. The trait-state-occasion model has fewer empirical problems than does the TSE model and is more broadly applicable than is the LST-AR model.

Variation in predictors of primary care career choice by year and stage of training

CONTEXT: It is not known whether factors associated with primary care career choice affect trainees differently at different times or stages of medical education.OBJECTIVE: To examine how role models, encouragement, and personal characteristics affect career choice at different stages (medical school vs residency) and periods (1994 vs 1997) of training.DESIGN: A split-panel design with 2 cross-sectional telephone surveys and a panel survey in 1994 and 1997.PARTICIPANTS: A national probability sample of fourth-year students (307 in 1994, 219 in 1997), 645 second-year residents in 1994, and 494 third-year residents in 1997. Of the fourth-year students interviewed in 1994, 241 (78.5%) were reinterviewed as third-year residents in 1997.MAIN OUTCOME MEASURE: Primary care (general internal medicine, general pediatrics, or family medicine) career choice.RESULTS: Having a primary care role model was a stronger predictor of primary care career choice for residents (odds ratio [OR], 18.0; 95% confidence interval [95% CI], 11.2 to 28.8 in 1994; OR, 43.7; 95% CI, 24.4 to 78.3 in 1997) than for students (OR, 6.5; 95% CI, 4.3 to 10.2; no variation by year). Likewise, peer encouragement was more predictive for residents (OR, 5.4; 95% CI, 3.3 to 8.9 in 1994; OR, 16.6; 95% CI; 9.7 to 28.4 in 1997) than for students (OR, 2.1; 95% CI, 1.3 to 3.2; no variation by year). Orientation to the emotional aspects of care was consistently associated with primary care career choice across stages and years of training.CONCLUSIONS: The effect of peer encouragement and role models on career choice differed for students and residents and, in the case of residents, by year of training, suggesting that interventions to increase the primary care workforce should be tailored to stage of training.

The Temporal Relation between Depression and Comorbid Psychopathology in Adolescents at Varied Risk for Depression.

BACKGROUND This study examined the temporal comorbidity of depressive disorders with anxiety, externalizing, and substance use disorders in adolescents who varied in risk for depression. METHODS Participants were 240 adolescents and their mothers who had either a history of depression (high-risk, n = 185) or were lifetime-free of psychiatric disorders (low-risk, n = 55). Children (54.2% females) were first evaluated in 6th grade (mean age = 11.86, SD = .57) with the K-SADS-PL to assess current and lifetime diagnoses, and then annually through 12th grade with the A-LIFE to assess diagnoses since the previous evaluation. RESULTS For girls, the rate of depression was high regardless of prior anxiety, whereas for boys, the odds that those with prior subthreshold anxiety would have subsequent subthreshold depression were 1.5 times those of boys with no prior subthreshold anxiety, controlling for risk. In addition, the odds that girls with prior substance use disorders would have a threshold depressive disorder subsequently were three times those of girls with no prior substance use disorders, controlling for risk. CONCLUSIONS These results highlight the importance of early detection of various forms of psychopathology in youth who then can be targeted for intervention. The prospective paths to comorbidity differed by sex, thus suggesting that interventions need to be constructed with sensitivity to these distinct diagnostic trajectories.

Does supportive parenting mitigate the longitudinal effects of peer victimization on depressive thoughts and symptoms in children

Cohen and Wills (Cohen, S., & Wills, T. A., 1985, Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310-357) described two broad models whereby social support could mitigate the deleterious effects of stress on health: a main effect model and stress-buffering model. A specific application of these models was tested in a three-wave, multimethod study of 1888 children to assess ways parental support (social support) mitigates the effects of peer victimization (stress) on childrens depressive symptoms and depression-related cognitions (health-related outcomes). Results revealed that (a) both supportive parenting and peer victimization had main effects on depressive symptoms and cognitions; (b) supportive parenting and peer victimization did not interact in the prediction of depressive thoughts and symptoms; (c) these results generalized across age and gender; and (d) increases in depressive symptoms were related to later reduction of supportive parenting and later increase in peer victimization. Although supportive parenting did not moderate the adverse outcomes associated with peer victimization, results show that its main effect can counterbalance or offset these effects to some degree. Implications for practice and future research are discussed.

Structure and Measurement of Depression in Youths: Applying Item Response Theory to Clinical Data

Our goals in this article were to use item response theory (IRT) to assess the relation of depressive symptoms to the underlying dimension of depression and to demonstrate how IRT-based measurement strategies can yield more reliable data about depression severity than conventional symptom counts. Participants were 3,403 children and adolescents from 12 contributing clinical and nonclinical samples; all participants had received the Kiddie Schedule of Affective Disorders and Schizophrenia for School-Aged Children. Results revealed that some symptoms reflected higher levels of depression and were more discriminating than others. Furthermore, use of IRT-based information about symptom severity and discriminability in the measurement of depression severity was shown to reduce measurement error and increase measurement fidelity.

Linking peer victimization to the development of depressive self-schemas in children and adolescents

Previous theory and research suggest that childhood experiences are more likely to generate depressive self-schemas when they focus attention on negative information about oneself, generate strong negative affect, and are repetitive or chronic. Persistent peer victimization meets these criteria. In the current study, 214 youths (112 females) with empirically-validated histories of high or low peer victimization completed self-report measures of negative and positive self-cognitions as well as incidental recall and recognition tests following a self-referent encoding task. Results supported the hypothesis that depressive self-schemas are associated with peer victimization. Specifically, peer victimization was associated with stronger negative self-cognitions, weaker positive self-cognitions, and an elimination of the normative memorial bias for recall of positive self-referential words. Effects were stronger for relational and verbal victimization compared to physical victimization. Support accrues to a model about the social-developmental origins of cognitive diatheses for depression.

A Randomized Controlled Trial of a Cognitive-Behavioral Program for the Prevention of Depression in Adolescents Compared With Nonspecific and No-Intervention Control Conditions

Adolescent depression is a common and recurrent disorder associated with significant impairment and other forms of psychopathology. Finding an effective intervention that prevents depression in adolescents is an important public health priority. Participants were 518 high school students (mean age = 15.09; SD = 0.76) from the mid-south of the United States. Participants were randomly assigned to 1 of 3 conditions: a cognitive-behavioral program (CB; n = 166), nonspecific control (NSp; n = 175), or a no-intervention control condition (NIC; n = 177). Both the CB and NSp conditions consisted of 90-min sessions administered once a week over a 10-week period during regular school hours. Depressive symptoms were assessed with the Childrens Depression Inventory (CDI) at baseline; postintervention; and at 4-, 8-, and 12-month follow-ups. The Time × Condition interaction was significant, F(8, 478.57) = 3.32, p = .001, indicating that at the 4-month follow-up, youth in the CB condition had significantly lower CDI scores compared with those in the NSp (p = .047, g = 0.29; CI [0.06, 0.52]) and the NIC conditions (p = .003, g = 0.30; CI [0.07, 0.53]). Future studies need to examine the importance of theory-driven change mechanisms, interpersonal relationships, and structural circumstances in schools as factors impacting the long-term effects of CB prevention programs.

Survival Analysis in Family Research

Within the field of family psychology, questions regarding the risk of event occurrence may be common. Such questions, about whether and when events occur and what predicts these occurrences, pose particular methodological challenges and are often best addressed via a statistical method known as survival analysis. This article provides a brief overview of that method, explicating through a data example the major components of a discrete-time survival analysis. Readers not familiar with this method are encouraged to use this article as an introduction to survival analysis and recognize its potential usefulness within the field of family psychology.

Effects of rumination on child and adolescent depressive reactions to a natural disaster: the 2010 Nashville flood

The current longitudinal study tested hypotheses about Nolen-Hoeksemas (1987, 1991) response styles theory (RST) of depression in a sample of child and adolescent public school students. Wave 1 measures of rumination, distraction, and depression were obtained 6 months prior to the 2010 Nashville flood. Similar measures plus a measure of flood-related stressors were administered at Wave 2, approximately ten days after students returned to school after the flood. Results revealed an indirect effect of preflood rumination on postflood depressive symptoms via the intervening variable of postflood rumination, and partial mediation of the effect of preflood depression on postflood depression. Further, the interaction of rumination with flood-related stressors was moderated by age, suggesting that rumination may not become a strong cognitive diathesis for depression until adolescence. Developmental implications emerged for the treatment of child and adolescent victims of natural disasters and for the application of RST to children and adolescents.