An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies

The propensity score is the probability of treatment assignment conditional on observed baseline characteristics. The propensity score allows one to design and analyze an observational (nonrandomized) study so that it mimics some of the particular characteristics of a randomized controlled trial. In particular, the propensity score is a balancing score: conditional on the propensity score, the distribution of observed baseline covariates will be similar between treated and untreated subjects. I describe 4 different propensity score methods: matching on the propensity score, stratification on the propensity score, inverse probability of treatment weighting using the propensity score, and covariate adjustment using the propensity score. I describe balance diagnostics for examining whether the propensity score model has been adequately specified. Furthermore, I discuss differences between regression-based methods and propensity score-based methods for the analysis of observational data. I describe different causal average treatment effects and their relationship with propensity score analyses.     

The evolution of alexia in two cases of posterior cortical atrophy

Posterior cortical atrophy (PCA) is an uncommon presentation of Alzheimer's disease (AD), characterised by prevalent anatomo-functional involvement of posterior cortical areas. Accordingly, the main clinical features at onset are disorders of high-order visual processing, such as alexia and impairments of visuo-spatial and visuo-constructional abilities. The clinical features in the early stages of disease are variable, and they have been suggested to stem from prevalent ventral or dorsal brain pathology, and/or asymmetric hemispheric involvement. With disease progression, these differences tend to blur with the increasing severity of neuropsychological dysfunction. We report two PCA patients showing different patterns of reading impairment (respectively, letter-by-letter reading and neglect dyslexia). A follow-up study suggested that the qualitative features of alexia remain distinctive with disease evolution. In addition, single photon emission tomography (SPECT) studies revealed different patterns of hypoperfusion, consistent with the alexia types. A careful reading assessment can provide important insights to the pattern of progression of the disease in patients with PCA up to the late stages of the pathology.     

An evaluation of interdependent and independent group contingencies during the good behavior game

The Good Behavior Game (GBG) uses an interdependent group contingency to improve classroom behavior. Despite the wealth of research on the effectiveness of the GBG, some teachers may have concerns about their students’ abilities to work in teams, particularly if they have a history of poor social skills. We used an alternating treatments design to compare the relative effectiveness of the GBG with interdependent and independent group contingencies in a classroom for children with emotional and behavioral disorders. Our results showed that both versions of the GBG reduced verbal disruptions, inappropriate sitting, and off‐task behaviors for all children. However, the majority of children preferred the interdependent arrangement. We discuss how these results may promote more widespread use of the GBG with children with substantial behavioral challenges.     

Predicting policy attitudes from general prejudice versus specific intergroup emotions

How are people’s social policy attitudes related to their affective reactions to the social groups affected by those policies? From a threat-based perspective, specific emotions toward a group—above and beyond general prejudice toward that group—should predict attitudes toward a policy affecting the group. To test this, 128 participants reported their support for four social policies, their general and specific affective reactions to each affected group, and the threats ostensibly posed by that group. Although general prejudice failed to predict each policy attitude after controlling for the specific emotions, specific emotions did indeed significantly predict each policy attitude after controlling for general prejudice. Moreover, these specific emotions tended to mediate predicted relationships between perceived threats and policy attitudes. In all, these results highlight the benefits of assessing specific intergroup emotions as they relate to political arenas.     

Severe pain predicts greater likelihood of subsequent suicide

Using data from the 1999 Large Health Survey of Veterans, Veterans Affairs' medical records, and the National Death Index (N = 260,254), the association between self-reported pain severity and suicide among veterans as examined, after accounting for demographic variables and psychiatric diagnoses. A Cox proportional hazards regression demonstrated that veterans with severe pain were more likely to die by suicide than patients experiencing none, mild, or moderate pain (HR: 1.33; 95% CI: 1.15, 1.54), after controlling for demographic and psychiatric characteristics. These results indicate that pain evaluations should be included in comprehensive suicide assessments and suicide prevention efforts.     

Measurement of ability emotional intelligence: Results for two new tests

Emotional intelligence (EI) has attracted considerable interest amongst both individual differences researchers and those in other areas of psychology who are interested in how EI relates to criteria such as well‐being and career success. Both trait (self‐report) and ability EI measures have been developed; the focus of this paper is on ability EI. The associations of two new ability EI tests with psychometric intelligence, emotion perception, and the Mayer–Salovey–Caruso EI test (MSCEIT) were examined. The new EI tests were the Situational Test of Emotion Management (STEM) and the Situational Test of Emotional Understanding (STEU). Only the STEU and the MSCEIT Understanding Emotions branch were significantly correlated with psychometric intelligence, suggesting that only understanding emotions can be regarded as a candidate new intelligence component. These understanding emotions tests were also positively correlated with emotion perception tests, and STEM and STEU scores were positively correlated with MSCEIT total score and most branch scores. Neither the STEM nor the STEU were significantly correlated with trait EI tests, confirming the distinctness of trait and ability EI. Taking the present results as a starting‐point, approaches to the development of new ability EI tests and models of EI are suggested.     

Re-conceptualizing Risk in Genetic Counseling: Implications for Clinical Practice

Risk communication is an important component of genetic counseling. However, many authors have noted that after genetic counseling, subjective risk frequently does not match the objective risk provided by the counselor. This inevitably leads to the conclusion that the risk communication process was not “effective”. There has been much discussion about how this problem can be better addressed, such that our clients recall numeric risks more accurately after genetic counseling. This article draws on the risk and probability literature from other fields (including psychology, economics, philosophy and climate change) to deconstruct the concepts of “risk” and risk perception to attempt to expand upon and develop thought and discussion about and investigation of the risk communication process in genetic counseling.     

Closing-in behaviour in preschool children

Many pre-school children show closing-in behaviour (CIB) in graphic copying tasks: a tendency to place their copy abnormally close to or even on top of the model. Similar phenomena have been studied in patients with dementia, though it is unclear whether the superficial similarities between CIB in development and dementia reflect common underlying cognitive mechanisms. The aim of the present study was to investigate the cognitive functions involved in CIB in pre-school children. Forty-one children (3–5 years) were assessed for CIB, and completed a neuropsychological battery targeting visuospatial abilities, short term memory (verbal and spatial) and attention (sustained attention, selective attention and attention switching). Binary logistic regression found that performance on the attention subtests was the best unique predictor of CIB. A second analysis, in which the three attention subtests were entered as separate predictors, suggested that attention switching ability was most strongly related to CIB. These results support the view that CIB in children reflects inadequate attentional control. The convergence of these results with similar observations in patients with dementia further suggests that similar cognitive factors underlie CIB in these two populations.     

Modeling collective choice: decision‐making on complex intellective tasks

Three studies examined group problem‐solving on complex intellective tasks. In Study 1, a decision model proposed by Laughlin and Hollingshead ( 1995 ) provided the best fit to actual group choices. This study also compared three‐person group versus individual performance with time constrained and number of problems unconstrained, with individuals solving non‐significantly more problems and groups obtaining significantly superior trials‐to‐solution scores. In Study 2, one member of each group was given additional information on how to perform the task and member extroversion was measured. Neither factor significantly impacted the decision‐making process. In Study 3, task expertise was assessed prior to the group interaction. Results indicate that group members were twice as likely to adopt an option proposed by an expert compared to other group members. Together these studies demonstrate that group problem solving is governed jointly by qualities of the task and qualities of the group members. Copyright © 2006 John Wiley & Sons, Ltd.