Research on Social Work Practice

This special issue highlights the necessity of rigorous methodologies, transparency, and innovation in social work research. Emphasizing causal inference, the contributions in this volume address replication challenges, design complexities, and data transparency. Key advancements and recommendations we put forth include study preregistration, data sharing, and novel methodologies such as advanced mixed methods considerations, emerging recommendations in regression discontinuity designs, the use co-twin control designs, and how to handle complex nesting structures with partially nested designs. Papers also explore cost-effective trial planning, replication improvement, and fostering mentorship to enhance rigor and societal trust. These approaches aim to answer critical questions of efficacy and equity, providing a roadmap for impactful research that informs practice, shapes policy, and improves lives in an evolving scientific landscape.

In recent years, the veracity of scientific findings has come under intense scrutiny in what has been called the “replication crisis.” This crisis is marked by the propagation of scientific claims which were subsequently contested, found to be exaggerated, or deemed false. This article describes the replication crisis and identifies examples of unreproducible results and irreplicable findings from across the biomedical and social sciences. Purported causes and potential remedies to the crisis are examined. It is argued that social work research suffers from the many analytic and methodological vices described here and that the profession is likely in crisis itself. Consequences for the discipline, as both a research and practice-based profession, are explored and paths forward are proposed.

Randomized controlled trials (RCTs) are designed to answer causal questions with internal validity. However, threats to internal validity exist for even well-designed RCTs. In this article, we focus on how preregistration can help address some specific threats to internal validity related to the reporting of results. Preregistration involves researchers publicly posting critical decision points in a study prior to conducting it for the purpose of making researcher plans transparent, making deviations from those plans discoverable, and improving the validity of tests of significance. We provide a brief overview of null-hypothesis significance testing; consider how questionable research practices (e.g., p-hacking) and conducting data-dependent analysis threaten the validity of significance tests; discuss how preregistration can help address these threats and how preregistration works for RCTs; note limitations and challenges to preregistration; and provide recommendations for increasing the use of preregistration by researchers conducting RCTs in social work, education, and related fields.

Sharing data publicly can provide numerous benefits to the data owner, data user, as well as the social work research community as a whole. Given the time and resources required to collect data in randomized controlled trials, gleaning the maximum amount of information from this data is highly desirable. Data sets considered to be exhausted by the primary research team often have valuable information that can be used by researchers with different research interests or analytic skill sets. Sharing these data allows other researchers to use these data to answer their research questions without duplicating the data collection efforts. Sharing data can also increase attention to the work of the primary research team, with papers with open data receiving more citations than those without public data. Engaging in open science practices such as data sharing can lead research to be seen as more trustworthy and reliable.

Purpose: This article introduces the regression discontinuity design (RDD) as a valuable quasi-experimental approach for evaluating social work interventions. Methods: The RDD is outlined, providing guidance on identifying and framing research questions within its structure. Detailed explanations of analytical procedures for estimating intervention effects under both sharp (treatment assignment strictly determined by a cutoff) and fuzzy (treatment assignment is influenced but not strictly determined by a cutoff) RDD conditions are included, along with common supplemental analyses to address potential validity threats. Results: Concrete examples of social work scenarios where RDD can be effectively employed are presented, showcasing innovative approaches that can enhance the design's utility. This paper also includes code for implementing these techniques. Conclusions: By examining the applications and methodologies of RDD, this article equips social work practitioners and researchers with a versatile tool to rigorously assess the impact of interventions and advance evidence-based practice within the field.

This study introduces recent advances in statistical power analysis methods and tools for designing and analyzing randomized cost-effectiveness trials (RCETs) to evaluate the causal effects and costs of social work interventions. The article focuses on two-level designs, where, for example, students are nested within schools, with interventions applied either at the school level (cluster design) or student level (multisite design). We explore three statistical modeling strategies—random-effects, constant-effects, and fixed-effects models—to assess the cost-effectiveness of interventions, and we develop corresponding power analysis methods and tools. Power is influenced by effect size, sample sizes, and design parameters. We developed a user-friendly tool, PowerUp!-CEA, to aid researchers in planning RCETs. When designing RCETs, it is crucial to consider cost variance, its nested effects, and the covariance between effectiveness and cost data, as neglecting these factors may lead to underestimated power.

Randomized control trials are considered the pinnacle for causal inference. In many cases, however, randomization of participants in social work research studies is not feasible or ethical. This paper introduces the co-twin control design study as an alternative quasi-experimental design to provide evidence of causal mechanisms when randomization is not possible. This method maximizes the genetic and environmental sameness between twins who are discordant on an “exposure” to provide strong counterfactuals as approximations of causal effects. We describe how the co-twin control design can be used to infer causality and in what type of situations the design might be useful for social work researchers. Finally, we give advantages and limitations to the design, list a set of Twin Registries with data available after application, and provide an example code for data analysis.

Purpose: Group-administered and shared facilitator treatments can induce nested data in a treatment arm that is not present in the control arm. Failure to accommodate these partially nested data structures produces study design inefficiencies, biased parameter estimates, and inaccurate inferences. This work introduces partially nested data structures. Method: We began by describing the features of partially nested data then discuss best practices and guidelines for study planning and analysis through examples commonly found in social work research. Results: The totality of this work provides social work researchers with the knowledge and tools to accommodate partially nested data in study planning and analysis including integration of comprehensive effects (i.e., mediation and moderation). Discussion: Improved understanding of partially nested data structures help researchers avoid the detrimental effects associated with disregarding them. Broadly, these methodological advances increase the capacity and quality of research in the field of social work.