Methods for Modeling Autocorrelation and Handling Missing Data in Mediation Analysis in Single Case Experimental Designs (SCEDs).

Single-Case Experimental Designs (SCEDs) are increasingly recognized as a valuable alternative to group designs. Mediation analysis is useful in SCEDs contexts because it informs researchers about the underlying mechanism through which an intervention influences the outcome. However, methods for conducting mediation analysis in SCEDs have only recently been proposed. Furthermore, repeated measures of a target behavior present the challenges of autocorrelation and missing data. This paper aims to extend methods for estimating indirect effects in piecewise regression analysis in SCEDs by (1) evaluating three methods for modeling autocorrelation, namely, Newey-West (NW) estimation, feasible generalized least squares (FGLS) estimation, and explicit modeling of an autoregressive structure of order one (AR(1)) in the error terms and (2) evaluating multiple imputation in the presence of data that are missing completely at random. FGLS and AR(1) outperformed NW and OLS estimation in terms of efficiency, Type I error rates, and coverage, while OLS was superior to the methods in terms of power for larger samples. The performance of all methods is consistent across 0% and 20% missing data conditions. 50% missing data led to unsatisfactory power and biased estimates. In light of these findings, we provide recommendations for applied researchers.

Beyond the Mean: A Flexible Framework for Studying Causal Effects Using Linear Models.

Graph-based causal models are a flexible tool for causal inference from observational data. In this paper, we develop a comprehensive framework to define, identify, and estimate a broad class of causal quantities in linearly parametrized graph-based models. The proposed method extends the literature, which mainly focuses on causal effects on the mean level and the variance of an outcome variable. For example, we show how to compute the probability that an outcome variable realizes within a target range of values given an intervention, a causal quantity we refer to as the probability of treatment success. We link graph-based causal quantities defined via the do-operator to parameters of the model implied distribution of the observed variables using so-called causal effect functions. Based on these causal effect functions, we propose estimators for causal quantities and show that these estimators are consistent and converge at a rate of [Formula: see text] under standard assumptions. Thus, causal quantities can be estimated based on sample sizes that are typically available in the social and behavioral sciences. In case of maximum likelihood estimation, the estimators are asymptotically efficient. We illustrate the proposed method with an example based on empirical data, placing special emphasis on the difference between the interventional and conditional distribution.

Forecasting Causal Effects of Interventions versus Predicting Future Outcomes.

The present article provides a didactic presentation and extension of selected features of Pearl's DAG-based approach to causal inference for researchers familiar with structural equation modeling. We illustrate key concepts using a cross-lagged panel design. We distinguish between (a) forecasts of the value of an outcome variable after an intervention and (b) predictions of future values of an outcome variable. We consider the mean level and variance of the outcome variable as well as the probability that the outcome will fall within an acceptable range. We extend this basic approach to include additive random effects, allowing us to distinguish between average effects of interventions and person-specific effects of interventions. We derive optimal person-specific treatment levels and show that optimal treatment levels may differ across individuals. We present worked examples using simulated data based on the results of a prior empirical study of the relationship between blood insulin and glucose levels.

The Role of Time in the Quest for Understanding Psychological Mechanisms.

The lead-lag structure of multivariate time-ordered observations and the possibility to disentangle between-person (BP) from within-person (WP) sources of variance are major assets of longitudinal (panel) data. Hence, psychologists are making increasing use of such data, often with the intent to delineate the dynamic properties of psychological mechanisms, understood as a sequence of causal effects that govern psychological functioning. However, even with longitudinal data, psychological mechanisms are not easily identified. In this article, we show how an adequate representation of time may enhance the tenability of causal interpretations in the context of multivariate longitudinal data analysis. We anchor our considerations with an example that illustrates some of the main problems and questions faced by applied researchers and practitioners. We distinguish between static versus dynamic and discrete versus continuous time modeling approaches and discuss their advantages and disadvantages. We place particular emphasis on different ways of addressing BP differences and stress their dual role as potential confounds versus valuable sources of information for improving estimation and aiding causal inference. We conclude by outlining an approach that offers the potential of better integration of information on BP differences and WP changes in the search for causal mechanisms along with a discussion of current problems and limitations.