International Gambling Studies | 2019
The need for open science practices and well-conducted replications in the field of gambling studies
Abstract
A field of study is only as strong as the research it produces. Although catchy and surprising results may generate acclaim, the research has actual value to the extent that the findings are true – they must be replicable (Popper, 1959; Schmidt, 2009). In the field of gambling studies, the scientific merit of published research has added importance given its ability to influence the regulation of gambling as well as efforts to prevent and treat disordered gambling. Unfortunately, there is reason to believe that many fields in the social sciences have a weak rate of replicability (Camerer et al., 2016, 2018; Open Science Collaboration, 2015), a revelation that has led to the so-called ‘replication crisis’ or ‘credibility revolution’. As a result of this crisis, attention has been directed at why many published findings fail to replicate. Central to these discussions has been: 1) low statistical power – the probability that a study will reject the null hypothesis (H0) when a specific alternative hypothesis (H1) is true – in published empirical research, and 2) the lack of open science practices (e.g. being explicit about how p-value(s) were obtained, posting data and materials to an online public repository). Low statistical power and the lack of open science practices decrease the replicability and credibility of reported findings. The net result has been decreased confidence in the results of published research and reforms aimed at improving new reports’ credibility. Regrettably, the field of gambling studies has lagged behind other fields in recognizing this replication crisis and taking appropriate action. To our knowledge, as of 2019, not a single session of any academic gambling-oriented conference has been dedicated to the replication crisis, and not a single paper (until this issue of International Gambling Studies) has been published on the relevance of the replication crisis for the field of gambling studies. A positive sign is InternationalGambling Studies’move toward encouraging contributors to fully disclose their research practices (i.e. open science). We laud the Journal for this change. We also laud International Gambling Studies for opening its pages to replication-based reports. This is a much needed change – a change that will help all stakeholders (e.g. researchers, policymakers, industry-related professionals) assess the replicability of past work in gambling studies. That said, we caution stakeholders that not all replications are created equal. It is important that all interested parties come to understand what constitutes a convincing replication, and why there is a need for a credibility revolution. In the current commentary, we discuss why, regardless of journal mandates, it is imperative that contributors and reviewers consider a priori statistical power analyses, effect sizes, and ensure open disclosures about research practices. Sound research practices are essential to the vitality, credibility, and replicability of any field of research. However, a deep dive into all the aforementioned issues is beyond the scope of the current commentary; our commentary has two goals. First, we hope to initiate a conversation that focuses researchers’ attention on the replicability of previously published research in the field of gambling studies. Second, we want to shine a light on how (as a field) we can do better to INTERNATIONAL GAMBLING STUDIES 2019, VOL. 19, NO. 3, 369–376 https://doi.org/10.1080/14459795.2019.1672769