Linda Collins
Linda Collins
Scroll
Professor of Social and Behavioral Sciences
-
Professional overview
-
Linda M. Collins is Professor of Global Public Health in the Department of Social and Behavioral Sciences, with a secondary appointment in the Department of Biostatistics. She earned her B.A. in Psychology at the University of Connecticut and her Ph.D. in Quantitative Psychology at the University of Southern California.
Collins’ research interests are focused on the development, dissemination, and application of the multiphase optimization strategy (MOST), a framework for the optimization of behavioral, biobehavioral, and social-structural interventions. The objective of MOST is to improve intervention effectiveness, efficiency, economy, and scalability. She is currently collaborating on research applying MOST in the areas of smoking cessation, the prevention of excessive drinking and risky sex in college students, and HIV services.
Collins’ research has been funded by the National Institute on Drug Abuse, the National Institute on Alcohol Abuse and Alcoholism, and the National Science Foundation, among others. She has given more than 150 presentations on MOST around the world, and her publications have appeared in journals in the fields of behavioral science, quantitative methodology, medicine, and engineering.
Collins has held tenured faculty positions at the University of Southern California and at Penn State University, where she was Distinguished Professor of Human Development and Family Studies and Director of The Methodology Center. She is a Fellow of the American Psychological Association, the Association for Psychological Science, the Society of Behavioral Medicine, and is a past president of the Society of Multivariate Experimental Psychology and the Society for Prevention Research.
-
Education
-
BA, Psychology, University of Connecticut, Storrs, CTPhD, Quantitative Psychology, University of Southern California, Los Angeles, CA
-
Honors and awards
-
Fulbright Specialist, National University of Ireland Galway (2018)Pauline Schmitt Russell Distinguished Career Award, Pennsylvania State University’s College of Health and Human Development (2017)Evan G. and Helen G. Pattishall Outstanding Research Achievement Award, Pennsylvania State University’s College of Health and Human Development (2011)President’s Award, Society for Prevention Research (2004)Faculty Scholar Medal for the Social and Behavioral Sciences, Pennsylvania State University (2000)Psychology Department Teacher of the Year, University of Southern California (1992)Psychology Department Mentorship Award, University of Southern California (1991)Society of Multivariate Experimental Psychology Award for Distinguished Early Career Contributions to Multivariate Behavioral Research (1991)
-
Areas of research and study
-
Behavioral ScienceCost EffectivenessCost-effective Health Programs and PoliciesDissemination and Implementation of Evidence-based Programs
-
Publications
Publications
Customizing treatment to the patient: Adaptive treatment strategies
Murphy, S. A., Collins, L. M., & Rush, A. J. (n.d.).Publication year
2007Journal title
Drug and alcohol dependenceVolume
88Page(s)
S1-S3Patterns of substance use onset among Hispanics in Puerto Rico and the United States
AbstractMaldonado-Molina, M. M., Collins, L. M., Lanza, S. T., Prado, G., Ramírez, R., & Canino, G. (n.d.).Publication year
2007Journal title
Addictive BehaviorsVolume
32Issue
10Page(s)
2432-2437AbstractPurpose: Examine patterns of progression in substance use among Hispanic youth 13 to 17 years of age from two longitudinally representative studies. Method: Patterns of substance use among youth in Puerto Rico were examined using a longitudinal study (n = 663) of adolescents living on the island. The National Longitudinal Study of Youth was used to examine patterns of substance use among Hispanics living in the United States (n = 1445). Latent transition analysis was used to estimate the probability of membership in each stage of substance use and incidence of transitions between different substance use stages over time. Results: Six stages best described the heterogeneity in substance use among youth in Puerto Rico. Five stages were sufficient to describe patterns of substance use among youth in the United States. Youth living in Puerto Rico reported lower rates of smoking and illicit drug use, but higher rates of alcohol use, when compared with rates among Hispanics in the United States. Discussion: Similar patterns of substance use were identified for Hispanic youth living in the United States and youth living in Puerto Rico.PROC LCA: A SAS procedure for latent class analysis
AbstractLanza, S. T., Collins, L. M., Lemmon, D. R., & Schafer, J. L. (n.d.).Publication year
2007Journal title
Structural Equation ModelingVolume
14Issue
4Page(s)
671-694AbstractLatent class analysis (LCA) is a statistical method used to identify a set of discrete, mutually exclusive latent classes of individuals based on their responses to a set of observed categorical variables. In multiple-group LCA, both the measurement part and structural part of the model can vary across groups, and measurement invariance across groups can be empirically tested. LCA with covariates extends the model to include predictors of class membership. In this article, we introduce PROC LCA, a new SAS procedure for conducting LCA, multiple-group LCA, and LCA with covariates. The procedure is demonstrated using data on alcohol use behavior in a national sample of high school seniors.Research Design and Methods
Collins, L. M. (n.d.). In Encyclopedia of Gerontology (1–).Publication year
2007Page(s)
V2-433-V2-442The Multiphase Optimization Strategy (MOST) and the Sequential Multiple Assignment Randomized Trial (SMART). New Methods for More Potent eHealth Interventions
AbstractCollins, L. M., Murphy, S. A., & Strecher, V. (n.d.).Publication year
2007Journal title
American journal of preventive medicineVolume
32Issue
5Page(s)
S112-S118AbstractAbstract: In this article two new methods for building and evaluating eHealth interventions are described. The first is the Multiphase Optimization Strategy (MOST). It consists of a screening phase, in which intervention components are efficiently identified for inclusion in an intervention or for rejection, based on their performance; a refining phase, in which the selected components are fine tuned and issues such as optimal levels of each component are investigated; and a confirming phase, in which the optimized intervention, consisting of the selected components delivered at optimal levels, is evaluated in a standard randomized controlled trial. The second is the Sequential Multiple Assignment Randomized Trial (SMART), which is an innovative research design especially suited for building time-varying adaptive interventions. A SMART trial can be used to identify the best tailoring variables and decision rules for an adaptive intervention empirically. Both the MOST and SMART approaches use randomized experimentation to enable valid inferences. When properly implemented, these approaches will lead to the development of more potent eHealth interventions.Using engineering control principles to inform the design of adaptive interventions: A conceptual introduction
AbstractRivera, D. E., Pew, M. D., & Collins, L. M. (n.d.).Publication year
2007Journal title
Drug and alcohol dependenceVolume
88Page(s)
S31-S40AbstractThe goal of this paper is to describe the role that control engineering principles can play in developing and improving the efficacy of adaptive, time-varying interventions. It is demonstrated that adaptive interventions constitute a form of feedback control system in the context of behavioral health. Consequently, drawing from ideas in control engineering has the potential to significantly inform the analysis, design, and implementation of adaptive interventions, leading to improved adherence, better management of limited resources, a reduction of negative effects, and overall more effective interventions. This article illustrates how to express an adaptive intervention in control engineering terms, and how to use this framework in a computer simulation to investigate the anticipated impact of intervention design choices on efficacy. The potential benefits of operationalizing decision rules based on control engineering principles are particularly significant for adaptive interventions that involve multiple components or address co-morbidities, situations that pose significant challenges to conventional clinical practice.A mixture model of discontinuous development in heavy drinking from ages 18 to 30: The role of college enrollment
AbstractLanza, S. T., & Collins, L. M. (n.d.).Publication year
2006Journal title
Journal of Studies on AlcoholVolume
67Issue
4Page(s)
552-561AbstractObjective: The purpose of this study was to illustrate the use of latent class analysis to examine change in behavior over time. Patterns of heavy drinking from ages 18 to 30 were explored in a national sample; the relationship between college enrollment and pathways of heavy drinking, particularly those leading to adult heavy drinking, was explored. Method: Latent class analysis for repeated measures is used to estimate common pathways through a stage-sequential process. Common patterns of development in a categorical variable (presence or absence of heavy drinking) are estimated and college enrollment is a grouping variable. Data were from the National Longitudinal Survey of Youth (N = 1,265). Results: Eight patterns of heavy drinking were identified: no heavy drinking (53.7%); young adulthood only (3.7%); young adulthood and adulthood (3.7%); college age only (2.6%); college age, young adulthood, and adulthood (8.7%); high school and college age (4.4%); high school, college age, and young adulthood (6.3%); and persistent heavy drinking (16.9%). Conclusions: We found no evidence that prevalence of heavy drinking for those enrolled in college exceeds the prevalence for those not enrolled at any of the four developmental periods studied. In fact, there is some evidence that being enrolled in college appears to be a protective factor for young adult and adult heavy drinking. College-enrolled individuals more often show a pattern characterized by heavy drinking during college ages only, with no heavy drinking prior to and after the college years, whereas nonenrolled individuals not drinking heavily during high school or college ages are at increased risk for adult heavy drinking.A multidimensional developmental model of alcohol use during emerging adulthood
AbstractAuerbach, K. J., & Collins, L. M. (n.d.).Publication year
2006Journal title
Journal of Studies on AlcoholVolume
67Issue
6Page(s)
917-925AbstractObjective: Longitudinal analyses identified unique multidimensional classes of alcohol use and examined individuals' movement among these classes during emerging adulthood. Method: Latent transition analysis was used to identify a developmental model of alcohol use incorporating four aspects of use: use in the past year, frequency of use, quantity of use, and heavy episodic drinking. Participants were drawn from the Reducing Risk in Young Adult Transitions study (N = 1,143). Participants' alcohol use was assessed at mean ages of 18.5, 20.5, and 22.5 years. Results: Through exploratory analysis, a five-class developmental model was identified as the best description of participants' alcohol use between ages 18.5 and 22.5 years. This model consisted of five multidimensional alcohol-use latent variables: no use, occasional low use, occasional high use, frequent high use, and frequent high use with heavy episodic drinking. Analyses provided information regarding the proportion of participants in each latent class in the model at each measurement occasion and patterns of participants' movement among latent classes during the observed age period. Conclusions: Although alcohol use increased overall for study participants between ages 18.5 and 22.5, participants in lower-level alcohol-use latent classes were more likely to remain in low-level latent classes over time, and participants in moderate- and high-level latent classes were more likely to be in the frequent high use with heavy episodic drinking latent class over time. Implications for the prevention of heavy episodic drinking are discussed.Analysis of longitudinal data: The integration of theoretical model, temporal design, and statistical model
AbstractCollins, L. M. (n.d.).Publication year
2006Journal title
Annual review of psychologyVolume
57Page(s)
505-528AbstractThis article argues that ideal longitudinal research is characterized by the seamless integration of three elements: (a) a well-articulated theoretical model of change observed using (b) a temporal design that affords a clear and detailed view of the process, with the resulting data analyzed by means of (c) a statistical model that is an operationalization of the theoretical model. Two general varieties of theoretical models are considered: models in which the time-related change of primary interest is continuous, and those in which it is characterized by movement between discrete states. In addition, two general types of temporal designs are considered: the longitudinal panel design and the intensive longitudinal design. For each general category of theoretical models, some of the analytic possibilities available for longitudinal panel designs and for intensive longitudinal designs are discussed. The article concludes with brief discussions of two issues particularly relevant to longitudinal research-missing data and measurement-and a few words about exploratory research.Methodological Considerations in Prevention Research
AbstractCollins, L. M., & Flaherty, B. P. (n.d.). In Handbooks of Sociology and Social Research (1–).Publication year
2006Page(s)
557-573AbstractThis chapter discussed a number of methodological considerations that face prevention research. It examined the central importance of theory in design and analysis of prevention studies. It considered the role of factorial invariance in developing culture-specific measures. It discussed the importance of statistical power and how it is dependent on factors other than sample size. It considered growth curve models, survival analysis, and lta, all relatively new procedures for dealing with change over time. It also discussed two approaches to the trait-state distinction and looked at missing data procedures and how important they are for prevention research. Finally, it discussed two general types of models that frequently arise in prevention research, mediation models and models of reciprocal causation, and how the customary ways of testing these models should perhaps be reconsidered.The proximal association between smoking and alcohol use among first year college students
AbstractDierker, L., Lloyd-Richardson, E., Stolar, M., Flay, B., Tiffany, S., Collins, L., Bailey, S., Nichter, M., Nichter, M., Clayton, R., Abrams, D., Balster, R., Dahl, R., Giovino, G., Henningfield, J., Koob, G., McMahon, R., Merikangas, K., Shiffman, S., … Stroud, L. (n.d.).Publication year
2006Journal title
Drug and alcohol dependenceVolume
81Issue
1Page(s)
1-9AbstractObjective: This study was undertaken to evaluate the association between patterns of day-to-day smoking and drinking among first year college students. Method: Using 210 days of weekly time-line follow-back diary data, the authors examined the within-person relationships between smoking and drinking. Bivariate time series procedures were utilized. Results: Findings revealed a high degree of significant cross-correlations between smoking and drinking in which the amount of use of one substance could be predicted by current, as well as past and future use of the other. For the majority of participants, smoking and drinking were positively associated with the alternate behavior. The most common pattern of prediction for individuals was within day (i.e. synchronous correlations). When examining rates of individuals showing significant cross-correlations according to their level of either smoking or drinking, those smoking less than one cigarette on average per day were found to be less likely to demonstrate a synchronous cross-correlation between the two behaviors than those smoking at higher rates. No significant association was found between level of drinking and the rate of significant synchronous cross-correlations between smoking and drinking. Conclusions: Reports of daily behavior over long periods of time have the potential to provide insight into the more proximal influences of smoking and alcohol use on one another. Future research is needed to establish the specific factors (i.e. third variables) and related mechanisms that may drive both behaviors.A strategy for optimizing and evaluating behavioral interventions
AbstractCollins, L. M., Murphy, S. A., Nair, V. N., & Strecher, V. J. (n.d.).Publication year
2005Journal title
Annals of Behavioral MedicineVolume
30Issue
1Page(s)
65-73AbstractBackground: Although the optimization of behavioral interventions offers the potential of both public health and research benefits, currently there is no widely agreed-upon principled procedure for accomplishing this. Purpose: This article suggests a multiphase optimization strategy (MOST) for achieving the dual goals of program optimization and program evaluation in the behavioral intervention field. Methods: MOST consists of the following three phases: (a) screening, in which randomized experimentation closely guided by theory is used to assess an array of program and/or delivery components and select the components that merit further investigation; (b) refining, in which interactions among the identified set of components and their interrelationships with covariates are investigated in detail, again via randomized experiments, and optimal dosage levels and combinations of components are identified; and (c) confirming, in which the resulting optimized intervention is evaluated by means of a standard randomized intervention trial. To make the best use of available resources, MOST relies on design and analysis tools that help maximize efficiency, such as fractional factorials. Results: A slightly modified version of an actual application of MOST to develop a smoking cessation intervention is used to develop and present the ideas. Conclusions: MOST has the potential to husband program development resources while increasing our understanding of the individual program and delivery components that make up interventions. Considerations, challenges, open questions, and other potential benefits are discussed.Historical review of school-based randomized trials for evaluating problem behavior prevention programs
AbstractFlay, B. R., & Collins, L. M. (n.d.).Publication year
2005Journal title
Annals of the American Academy of Political and Social ScienceVolume
599Page(s)
115-146AbstractThe authors provide a historical review of developments in the methods of school-based evaluations of problem behavior prevention interventions. The design and statistical methodologies used in school-based intervention research have advanced tremendously over the past twenty years. Methods have improved for approaches to the randomization of whole schools, the choice of appropriate comparison or control groups, solutions when randomization breaks down, limiting and handling of variation in integrity of the intervention received, limiting biases introduced by data collection, awareness of the effects of intensive and long-term data collection, limiting and analysis of subject attrition and other missing data, approaches to obtaining parental consent for children to engage in research, design and analysis issues when only small numbers of schools are available or can be afforded, the choice of the unit of analysis, phases of research, optimizing and extending the reach of interventions, and differential effects in subpopulations. The authors conclude that sequential planning, timing, keeping up with methodological advances, publication of results, and accumulation of knowledge are all important in conducting high-quality school-based intervention research, and that the devil is in the details.Using data augmentation to obtain standard errors and conduct hypothesis tests in latent class and latent transition analysis
AbstractLanza, S. T., Collins, L. M., Schafer, J. L., & Flaherty, B. P. (n.d.).Publication year
2005Journal title
Psychological MethodsVolume
10Issue
1Page(s)
84-100AbstractLatent class analysis (LCA) provides a means of identifying a mixture of subgroups in a population measured by multiple categorical indicators. Latent transition analysis (LTA) is a type of LCA that facilitates addressing research questions concerning stage-sequential change over time in longitudinal data. Both approaches have been used with increasing frequency in the social sciences. The objective of this article is to illustrate data augmentation (DA), a Markov chain Monte Carlo procedure that can be used to obtain parameter estimates and standard errors for LCA and LTA models. By use of DA it is possible to construct hypothesis tests concerning not only standard model parameters but also combinations of parameters, affording tremendous flexibility. DA is demonstrated with an example involving tests of ethnic differences, gender differences, and an Ethnicity X Gender interaction in the development of adolescent problem behavior.A conceptual framework for adaptive preventive interventions
AbstractCollins, L. M., Murphy, S. A., & Bierman, K. L. (n.d.).Publication year
2004Journal title
Prevention ScienceVolume
5Issue
3Page(s)
185-196AbstractRecently, adaptive interventions have emerged as a new perspective on prevention and treatment. Adaptive interventions resemble clinical practice in that different dosages of certain prevention or treatment components are assigned to different individuals, and/or within individuals across time, with dosage varying in response to the intervention needs of individuals. To determine intervention need and thus assign dosage, adaptive interventions use prespecified decision rules based on each participant's values on key characteristics, called tailoring variables. In this paper, we offer a conceptual framework for adaptive interventions, discuss principles underlying the design and evaluation of such interventions, and review some areas where additional research is needed.Analyzing the acquisition of drug self-administration using growth curve models
AbstractLanza, S. T., Donny, E. C., Collins, L. M., & Balster, R. L. (n.d.).Publication year
2004Journal title
Drug and alcohol dependenceVolume
75Issue
1Page(s)
11-21AbstractCurrent approaches to studying acquisition of drug self-administration have modest power to detect individual differences in the pattern of acquisition or to efficiently and accurately describe trajectories of behavior change. Methodological advances in human research have elucidated approaches to describing repeated measure data that focus on modeling the behavior of individual subjects. In this article, we re-analyzed data published in [Psychopharmacology 136 (1998) 83] using growth curve modeling to characterize the acquisition of nicotine-taking in rats. Change over time in the infusion rate was examined, revealing that the acquisition process could be described with a quadratic equation represented by intercept, slope, and acceleration parameters. Unit dose of nicotine, sex and fixed ratio (FR) schedule of reinforcement had significant effects on the acquisition curves. Dose altered the absolute rate of infusions, but not the slope or acceleration, indicating that, when an effective dose was available, the shape of acquisition trajectories was not affected by dose. In addition, dose impacted acquisition by moderating the disruption in infusion rates after an increase in the response requirement. Thus, the role of a higher dose may not be to accelerate the acquisition process but to lead to behavior that is more resistant to change. Trajectories differed between males and females at the smallest dose, but these differences dissipated by the end of acquisition. Growth curve modeling captures the process of acquisition of drug self-administration and facilitates a greater understanding of the individual differences in change in drug-taking behavior over time.Using growth models to relate acquisition of nicotine self-administration to break point and nicotinic receptor binding
AbstractDonny, E. C., Lanza, S. T., Balster, R. L., Collins, L. M., Caggiula, A., & Rowell, P. P. (n.d.).Publication year
2004Journal title
Drug and alcohol dependenceVolume
75Issue
1Page(s)
23-35AbstractGrowth modeling can be used to characterize individual and mean acquisition trajectories for drug self-administration. Individual characteristics can also be incorporated into the growth model, providing a powerful tool for investigating the relationship between acquisition and other behavioral and biological measures. We illustrate the utility of this method by examining the relationship between acquisition of nicotine self-administration and (1) break point on a progressive ratio schedule of reinforcement, and (2) the density of brain nicotinic receptors (Bmax). Daily infusion rates from male and female Sprague-Dawley rats were modeled with break point or Bmax as time-invariant covariates. Use of this model led to two novel findings regarding individual differences in acquisition. First, greater rates of change in infusions early in acquisition were related to higher break points; this relationship was mediated by a similar effect of increasing the number of responses required to obtain nicotine. Second, animals displaying more resistance to increases in the response requirement during acquisition, as indicated by a smaller drop in the rate of nicotine self-administration, generally had fewer nicotinic receptors at the end of the experiment. The relationships revealed demonstrate the usefulness of growth models in the quantitative analysis of individual differences in drug self-administration behavior.Adaptive sampling in research on risk-related behaviors
AbstractThompson, S. K., & Collins, L. M. (n.d.).Publication year
2002Journal title
Drug and alcohol dependenceVolume
68Page(s)
57-67AbstractThis article introduces adaptive sampling designs to substance use researchers. Adaptive sampling is particularly useful when the population of interest is rare, unevenly distributed, hidden, or hard to reach. Examples of such populations are injection drug users, individuals at high risk for HIV/AIDS, and young adolescents who are nicotine dependent. In conventional sampling, the sampling design is based entirely on a priori information, and is fixed before the study begins. By contrast, in adaptive sampling, the sampling design adapts based on observations made during the survey; for example, drug users may be asked to refer other drug users to the researcher. In the present article several adaptive sampling designs are discussed. Link-tracing designs such as snowball sampling, random walk methods, and network sampling are described, along with adaptive allocation and adaptive cluster sampling. It is stressed that special estimation procedures taking the sampling design into account are needed when adaptive sampling has been used. These procedures yield estimates that are considerably better than conventional estimates. For rare and clustered populations adaptive designs can give substantial gains in efficiency over conventional designs, and for hidden populations link-tracing and other adaptive procedures may provide the only practical way to obtain a sample large enough for the study objectives.Measurement and design issues in tobacco and drug use research
Collins, L. M., Flaherty, B. P., & Colby, S. M. (n.d.).Publication year
2002Journal title
Drug and alcohol dependenceVolume
68Page(s)
S1-S2Pubertal timing and the onset of substance use in females during early adolescence
AbstractLanza, S. T., & Collins, L. M. (n.d.).Publication year
2002Journal title
Prevention ScienceVolume
3Issue
1Page(s)
69-82AbstractThe goal of this study is to examine in detail the relationship between pubertal timing and substance use onset using a sample of females from The National Longitudinal Study of Adolescent Health. The sample includes 966 females who were in 7th grade at Wave 1 and 8th grade at Wave 2. Participants in the sample are approximately 69% White, 20% African American, 4% Asian or Pacific Islander, 2% American Indian, 4% other, of Hispanic origin, and 1% other, not of Hispanic origin. Twenty percent of the females were identified as early maturers based on self-reports of body changes (increased breast size and body curviness) measured in 7th grade. These participants are hypothesized to be at increased risk for substance use onset. Important differences in substance use onset were found between early maturers and their on-time and late-maturing counterparts. During 7th grade, females in the early-maturing group are three times more likely to be in the most advanced stage of substance use (involving alcohol use, drunkenness, cigarette use, and marijuana use) than are those in the on-time/late group. Prevalence rates indicate that early maturers are more likely to have tried alcohol, tried cigarettes, been drunk, and tried marijuana. Prospective findings show that early developers are significantly more likely to transition out of the "No Substance Use" stage between 7th and 8th grade (47% for early developers vs. 22% for on-time and late developers). In addition, early developers are more likely to advance in substance use in general, regardless of their level of use at Grade 7.The effect of the timing and spacing of observations in longitudinal studies of tobacco and other drug use: Temporal design considerations
AbstractCollins, L. M., & Graham, J. W. (n.d.).Publication year
2002Journal title
Drug and alcohol dependenceVolume
68Page(s)
85-96AbstractThis article explores the impact of the temporal design, i.e. the sampling of times of measurement, on the statistical and substantive conclusions drawn from longitudinal biomedical and social science research. It is shown that for a study of a given duration, if observations are spaced too far apart the resulting data can support misleading conclusions, whereas if observations are spaced relatively close together, a much more veridical picture of the process of interest is provided. The application of these ideas in several areas is discussed, including correlation and regression analysis where a variable measured at one time is used to predict a variable measured at a later time; growth curve analyses; and analyses involving stage-sequential processes. We argue that longitudinal designs should relate the choice of timing and spacing of observations in longitudinal studies to characteristics of the processes being measured. In addition, consideration of the possible effects of measurement design on results of statistical analyses may aid in their interpretation. New approaches involving intensive data collection with much shorter measurement intervals, such as Ecological Momentary Assessment, are promising, but are costly and are not suitable for every research question. More information is needed to help guide researchers in their choice of temporal design.A comparison of inclusive and restrictive strategies in modern missing data procedures
AbstractCollins, L. M., Schafer, J. L., & Kam, C. M. (n.d.).Publication year
2001Journal title
Psychological MethodsVolume
6Issue
3Page(s)
330-351AbstractTwo classes of modern missing data procedures, maximum likelihood (ML) and multiple imputation (MI), tend to yield similar results when implemented in comparable ways. In either approach, it is possible to include auxiliary variables solely for the purpose of improving the missing data procedure. A simulation was presented to assess the potential costs and benefits of a restrictive strategy, which makes minimal use of auxiliary variables, versus an inclusive strategy, which makes liberal use of such variables. The simulation showed that the inclusive strategy is to be greatly preferred. With an inclusive strategy not only is there a reduced chance of inadvertently omitting an important cause of missingness, there is also the possibility of noticeable gains in terms of increased efficiency and reduced bias, with only minor costs. As implemented in currently available software, the ML approach tends to encourage the use of a restrictive strategy, whereas the MI approach makes it relatively simple to use an inclusive strategy.Developmental pathways to alcohol abuse and dependence in young adulthood
AbstractGuo, J., Collins, L. M., Hill, K. G., & Hawkins, J. D. (n.d.).Publication year
2000Journal title
Journal of Studies on AlcoholVolume
61Issue
6Page(s)
799-808AbstractObjective: To determine if people who were diagnosed with alcohol abuse or dependence (AAD) at age 21 had different developmental patterns of alcohol use in adolescence than non-AAD individuals. Method: An ethnically diverse urban sample of 808 children was surveyed at age 10 in 1985 and followed prospectively to age 21 in 1996. AAD at age 21 was assessed following DSM-IV criteria. Latent Transition Analysis (LTA) was used to identify four statuses of alcohol use (nonuse, initiation only, current use only, heavy episodic drinking), as well as transition probabilities between these four statuses from elementary school to middle school and from middle school to high school among the AAD and non-AAD group. Results: The prevalence of alcohol use statuses during elementary school was similar in the two groups. Differences in alcohol use emerged in middle school and became more pronounced in high school. In middle school, AAD individuals were more likely to have initiated or been current drinkers than non-AAD individuals. However, the two groups did not differ in the prevalence of heavy episodic drinking in middle school. In high school, most AAD individuals were in the heavy episodic drinking status (54%), while most non-AAD individuals were in the initiation only (33%) or current use only (34%) statuses. Conclusions: These findings suggest preventive intervention targets for different developmental periods. Alcohol abuse and dependence at age 21 may be prevented by delaying alcohol initiation, by reducing current use in middle school and by reducing heavy episodic drinking in high school.An alternative framework for defining mediation
AbstractCollins, L. M., Graham, J. W., & Flaherty, B. P. (n.d.).Publication year
1998Journal title
Multivariate Behavioral ResearchVolume
33Issue
2Page(s)
295-312AbstractThe present article provides an alternative framework for evaluating mediated relationships. From this perspective, a mediated process is a chain reaction, beginning with an independent variable that affects a mediator that in turn affects an outcome. The definition of mediation offered here, presented for stage sequences, states three conditions for establishing mediation: (a) the independent variable affects the probability of the sequence no mediator to mediator to outcome; (b) the independent variable affects the probability of a transition into the mediator stage; (c) the mediator affects the probability of a transition into the outcome stage at every level of the independent variable. This definition of mediation is compared and contrasted with the well-known definition of mediation for continuous variables discussed in Baron and Kenny (1986), Judd and Kenny (1981), and Kenny, Kashy, and Bolger (1997). The definition presented in this article emphasizes the intraindividual, time-ordered nature of mediation.Is reliability obsolete? A commentary on "are simple gain scores obsolete?"
AbstractCollins, L. M. (n.d.).Publication year
1996Journal title
Applied Psychological MeasurementVolume
20Issue
3Page(s)
289-292AbstractWilliams & Zimmerman (1996) provided much-needed clarification on the reliability of gain scores. This commentary translates these ideas into recognizable patterns of change that tend to produce reliable or unreliable gain scores. It also questions the relevance of the traditional idea of reliability to the measurement of change.