Biomedical subjects
John W Graham
Publications and source records attributed to John W Graham.
Planned missing data designs in psychological research.
The authors describe 2 efficiency (planned missing data) designs for measurement: the 3-form design and the 2-method measurement design. The 3-form design, a kind of matrix sampling, allows researchers to leverage limited resources to collect data for 33% more survey questions than can be answered by any 1 respondent. Power tables for estimating correlation effects illustrate the benefit of this design. The 2-method measurement design involves a relatively cheap, less valid measure of a construct and an expensive, more valid measure of the same construct. The cost effectiveness of this design stems from the fact that few cases have both measures, and many cases have just the cheap measure. With 3 brief simulations involving structural equation models, the authors show that compared with the same-cost complete cases design, a 2-method measurement design yields lower standard errors and a higher effective sample size for testing important study parameters. With a large cost differential between cheap and expensive measures and small effect sizes, the benefits of the design can be enormous. Strategies for using these 2 designs are suggested.
The drug resistance strategies intervention: program effects on substance use.
This study evaluates the Drug Resistance Strategies (DRS) project, a culturally grounded, communication-based substance use prevention program implemented in 35 middle schools in Phoenix, Arizona. The intervention consisted of 10 lessons taught by the classroom teacher that imparted the knowledge, motivation, and skills needed to resist drug offers. The evaluation used growth modeling to analyze significant differences in average postintervention substance use (alcohol, cigarettes, and marijuana) and growth of use over the course of the study. The study involved 6,298 seventh graders (65% Mexican/Mexican American) who responded to at least 1 of 4 questionnaires (1 pretest and 3 follow-up measures). When compared to a control group, the DRS intervention appeared to significantly limit the increase in the number of students reporting recent substance use, especially alcohol and marijuana use. The multicultural version of the curriculum proved most broadly effective, followed by the version targeting Mexican American youth. The development of a culturally grounded prevention curriculum for Mexican American youth expands the population being served by interventions. Moreover, the success of the multicultural curriculum version, which has the broadest application, provides particular promise, and the article demonstrates how a growth modeling approach can be used to evaluate a communication-based intervention by analyzing changes over time rather than differences between the pretest and posttest scores.
Bisphosphonates and orthodontics: Clinical implications.
Explore the source record for details and available documents.
The effect of the timing and spacing of observations in longitudinal studies of tobacco and other drug use: temporal design considerations.
This 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.
Construct validity in health behavior research: interpreting latent variable models involving self-report and objective measures.
Latent variable models assess the common variance across multiple indicators of a specific construct and are often used when measurement error may bias parameter estimates. However, care must be taken when interpreting the meaning of the latent construct when using item indicators that come from different measurement domains (e.g., self-report and biochemical indicators of smoking). Utilizing simulated data, we demonstrate that even though a model may be considered to have a "good fit" based on conventional criteria, data interpretation may be misleading or erroneous if precautions are not taken when specifying residual covariances. These findings have important implications for health-related research. Whenever different kinds of data are used to define latent variables in a health domain, exactly what items are used, and what biases may be present can affect, sometimes dramatically, (a) the definition of the latent variables and (b) the effects of the latent variables on other variables of interest.
Data quality in evaluation of an alcohol-related harm prevention program.
The authors report the reliability and convergent validity in a sample of college students for 27 composite scales and two items covering alcohol use, cigarette smoking, marijuana use, and other drug use; beliefs relating to alcohol use; perceived norms for alcohol-related behavior; harm prevention skills; intentions to take prevention action; harm prevention action taken; risk taken; experienced harm; and other health-related behaviors and person characteristics. Data quality assessment strategies and missing data procedures were illustrated for large, multivariate, longitudinal data sets. Results indicate 23 of the 27 composite scales had at least acceptable reliability, and the remaining 4 composite scales had at least marginally acceptable reliability. At least moderate construct validity was demonstrated for 25 scales.
Missing data: our view of the state of the art.
Statistical procedures for missing data have vastly improved, yet misconception and unsound practice still abound. The authors frame the missing-data problem, review methods, offer advice, and raise issues that remain unresolved. They clear up common misunderstandings regarding the missing at random (MAR) concept. They summarize the evidence against older procedures and, with few exceptions, discourage their use. They present, in both technical and practical language, 2 general approaches that come highly recommended: maximum likelihood (ML) and Bayesian multiple imputation (MI). Newer developments are discussed, including some for dealing with missing data that are not MAR. Although not yet in the mainstream, these procedures may eventually extend the ML and MI methods that currently represent the state of the art.
The accuracy of artificial neural networks in predicting long-term outcome after traumatic brain injury.
OBJECTIVE: This study compared the accuracy of artificial neural networks to multiple regression and classification and regression trees in predicting outcomes of 1,644 patients in the Traumatic Brain Injury Model Systems database 1 year after injury. METHODS: Data from rehabilitation admission were used to predict discharge scores on the Functional Independence Measure, the Disability Rating Scale, and the Community Integration Questionnaire. RESULTS: Artificial neural networks did not demonstrate greater accuracy in predicting outcomes than did the more widely used method of multiple regression. Both of these methods outperformed classification and regression trees. CONCLUSION: Because of the sophisticated form of multiple regression with splines that was used, firm conclusions are limited about the relative accuracy of artificial neural networks compared to more widely used forms of multiple regression.