Reeping Paul M, Morrison Christopher N, Rudolph Kara E, Goyal Monika K, Branas Charles C
Department of Epidemiology, Columbia University, Mailman School of Public Health, 722 West 168th Street, New York, NY, 10032, USA.
Department of Pediatrics, Division of Emergency Medicine, Children's National Health System, Washington, D.C, USA.
Inj Epidemiol. 2021 Jan 18;8(1):2. doi: 10.1186/s40621-020-00296-5.
Due to the differences in the way gun law permissiveness scales were created and speculation about the politically motivated underpinnings of the various scales, there have been questions about their reliability.
We compared seven gun law permissiveness scales, varying by type and sources, for an enhanced understanding of the extent to which choice of a gun law permissiveness scale could affect studies related to gun violence outcomes in the United States. Specifically, we evaluated seven different scales: two rankings, two counts, and three scores, arising from a range of sources. We calculated Spearman correlation coefficients for each pair of scales compared. Cronbach's standardized alpha and Guttman's lambda were calculated to evaluate the relative reliability of the scales, and we re-calculated Cronbach's alpha after systematically omitting each scale to assess whether the omitted scale contributed to lower internal consistency between scales. Factor analysis was used to determine single factor loadings and estimates. We also assessed associations between permissiveness of gun laws and total firearm deaths and suicides in multivariable regression analyses.
All pairs of scales were highly correlated (average Spearman's correlation coefficient r = 0.77) and had high relative reliability (Cronbach's alpha = 0.968, Guttman's lambda = 0.975). All scales load onto a single factor. The choice of scale did not meaningfully change the parameter estimates for the associations between permissiveness of gun laws and gun deaths and suicides.
Gun law permissiveness scales are highly correlated despite any perceived political agenda, and the choice of gun law permissiveness scale has little effect on study conclusions related to gun violence outcomes.
由于枪支法律宽松程度量表的创建方式存在差异，以及对各种量表背后政治动机的猜测，人们对其可靠性提出了质疑。
我们比较了七种枪支法律宽松程度量表，这些量表在类型和来源上各不相同，以加深对选择枪支法律宽松程度量表可能影响美国枪支暴力结果相关研究程度的理解。具体而言，我们评估了七种不同的量表：两种排名、两种计数和三种得分，这些量表来自一系列来源。我们计算了每对比较量表的斯皮尔曼相关系数。计算了克朗巴哈标准化阿尔法系数和古特曼拉姆达系数以评估量表的相对可靠性，并且在系统地省略每个量表后重新计算克朗巴哈阿尔法系数，以评估被省略的量表是否导致量表之间的内部一致性降低。使用因子分析来确定单因子载荷和估计值。我们还在多变量回归分析中评估了枪支法律宽松程度与总枪支死亡和自杀之间的关联。
所有量表对之间都高度相关（平均斯皮尔曼相关系数r = 0.77），并且具有较高的相对可靠性（克朗巴哈阿尔法系数 = 0.968，古特曼拉姆达系数 = 0.975）。所有量表都加载到一个单一因子上。量表的选择并没有显著改变枪支法律宽松程度与枪支死亡和自杀之间关联的参数估计值。
尽管存在任何可感知的政治议程，枪支法律宽松程度量表之间高度相关，并且枪支法律宽松程度量表的选择对与枪支暴力结果相关的研究结论影响很小。
