Manipulations checks are postexperimental measures widely used to verify that subjects understood the treatment. Some researchers drop subjects who failed manipulation checks in order to limit the analyses to attentive subjects. This short report offers a novel illustration on how this practice may bias experimental results: in the present case, through confirming a hypothesis that is likely false. In a survey experiment, subjects were primed with a fictional news story depicting an economic decline versus prosperity. Subjects were then asked whether the news story depicted an economic decline or prosperity. Results indicate that responses to this manipulation check captured subjects’ preexisting beliefs about the economic situation. As a consequence, dropping subjects who failed the manipulation check mixes the effects of preexisting and induced beliefs, increasing the risk of false positive findings. Researchers should avoid dropping subjects based on posttreatment measures and rely on pretreatment measures of attentiveness.

Dropping subjects based on the results of a manipulation check following treatment assignment is common practice across the social sciences, presumably to restrict estimates to a subpopulation of subjects who understand the experimental prompt. We show that this practice can lead to serious bias and argue for a focus on what is revealed without discarding subjects. Generalizing results developed in Zhang and Rubin (2003) and Lee (2009) to the case of multiple treatments, we provide sharp bounds for potential outcomes among those who would pass a manipulation check regardless of treatment assignment. These bounds may have large or infinite width, implying that this inferential target is often out of reach. As an application, we replicate Press, Sagan, and Valentino (2013) with a design that does not drop subjects that failed the manipulation check and show that the findings are likely stronger than originally reported. We conclude with suggestions for practice, namely alterations to the experimental design.

Researchers investigating causal mechanisms in survey experiments often rely on nonrandomized quantities to isolate the indirect effect of treatment through these variables. Such an approach, however, requires a “selection-on-observables” assumption, which undermines the advantages of a randomized experiment. In this paper, we show what can be learned about casual mechanisms through experimental design alone. We propose a factorial design that provides or withholds information on mediating variables and allows for the identification of the overall average treatment effect and the controlled direct effect of treatment fixing a potential mediator. While this design cannot identify indirect effects on its own, it avoids making the selection-on-observable assumption of the standard mediation approach while providing evidence for a broader understanding of causal mechanisms that encompasses both indirect effects and interactions. We illustrate these approaches via two examples: one on evaluations of US Supreme Court nominees and the other on perceptions of the democratic peace.

This article reviews some of the most effective programmes for saving children from a life of crime, and also presents the results of cost-benefit analyses of some of these programmes. The best programmes include general parent education in home visiting programmes, parent management training, pre-school intellectual enrichment programmes, child skills training, Functional Family Therapy, Multidimensional Treatment Foster Care and Multisystemic Therapy. Communities That Care is a useful overarching programme. Most of these programmes have been shown to reduce crime and save money. The time is ripe to establish national agencies in all countries which will advance knowledge about early risk factors (from longitudinal studies) and about effective developmental interventions (from randomized experiments and cost-benefit analyses).