Division on Addiction, Cambridge Health Alliance, a Harvard Medical School Teaching Hospital, Malden, MA, USA.
Harvard Medical School, Boston, MA, USA.
J Gambl Stud. 2023 Jun;39(2):987-1011. doi: 10.1007/s10899-022-10120-y. Epub 2022 Jun 9.
The replication crisis has stimulated researchers around the world to adopt open science research practices intended to reduce publication bias and improve research quality. Open science practices include study pre-registration, open data, open access, and avoiding methods that can lead to publication bias and low replication rates. Although gambling studies uses similar research methods as behavioral research fields that have struggled with replication, we know little about the uptake of open science research practices in gambling-focused research. We conducted a scoping review of 500 recent (1/1/2016-12/1/2019) studies focused on gambling and problem gambling to examine the use of open science and transparent research practices. Our results showed that a small percentage of studies used most practices: whereas 54.6% (95% CI: [50.2, 58.9]) of studies used at least one of nine open science practices, each practice's prevalence was: 1.6% for pre-registration (95% CI: [0.8, 3.1]), 3.2% for open data (95% CI: [2.0, 5.1]), 0% for open notebook, 35.2% for open access (95% CI: [31.1, 39.5]), 7.8% for open materials (95% CI: [5.8, 10.5]), 1.4% for open code (95% CI: [0.7, 2.9]), and 15.0% for preprint posting (95% CI: [12.1, 18.4]). In all, 6.4% (95% CI: [4.6, 8.9]) of the studies included a power analysis and 2.4% (95% CI: [1.4, 4.2]) were replication studies. Exploratory analyses showed that studies that used any open science practice, and open access in particular, had higher citation counts. We suggest several practical ways to enhance the uptake of open science principles and practices both within gambling studies and in science more generally.
复制危机促使世界各地的研究人员采用开放科学研究实践，旨在减少发表偏倚并提高研究质量。开放科学实践包括研究预注册、开放数据、开放获取以及避免可能导致发表偏倚和低复制率的方法。尽管赌博研究使用与行为研究领域类似的研究方法，但这些领域一直存在复制问题，我们对赌博相关研究中采用开放科学研究实践的情况知之甚少。我们对 500 项近期（2016 年 1 月 1 日至 2019 年 12 月 1 日）专注于赌博和赌博问题的研究进行了范围综述，以检查开放科学和透明研究实践的使用情况。我们的研究结果表明，只有一小部分研究使用了大多数实践：尽管有 54.6%（95%置信区间：[50.2，58.9]）的研究使用了九种开放科学实践中的至少一种，但每种实践的普遍性如下：1.6%的研究进行了预注册（95%置信区间：[0.8，3.1]），3.2%的研究开放了数据（95%置信区间：[2.0，5.1]），0%的研究开放了笔记本，35.2%的研究开放了获取（95%置信区间：[31.1，39.5]），7.8%的研究开放了材料（95%置信区间：[5.8，10.5]），1.4%的研究开放了代码（95%置信区间：[0.7，2.9]），15.0%的研究预印本发布（95%置信区间：[12.1，18.4]）。总体而言，6.4%（95%置信区间：[4.6，8.9]）的研究包括了功效分析，2.4%（95%置信区间：[1.4，4.2]）是复制研究。探索性分析表明，使用任何开放科学实践的研究，特别是开放获取，其引用次数更高。我们提出了一些实用的方法来增强赌博研究中以及更广泛的科学界对开放科学原则和实践的采用。
