Duquesne University, School of Nursing, United States.
Res Social Adm Pharm. 2021 May;17(5):1004-1011. doi: 10.1016/j.sapharm.2020.07.027. Epub 2020 Aug 15.
This commentary provides a brief mathematical review of exploratory factor analysis, the common factor model, and principal components analysis. Details and recommendations related to the goals, measurement scales, estimation technique, factor retention, item retention, and rotation of factors. For researchers interested in attempting to identify latent factors, exploratory factor analysis, the common factor model, is the appropriate analysis. For surveys with Likert-type scales weighted least squares with robust standard errors is recommended along with oblique rotation. Alternative techniques for analyzing the data, e.g., item response theory and machine learning, are briefly discussed. Finally, a basic check list for researchers and reviewers is provided.
本评论简要回顾了探索性因素分析、共同因素模型和主成分分析的数学原理。评论还就目标、度量尺度、估计技术、因子保留、项目保留和因子旋转等方面的细节和建议进行了讨论。对于希望尝试识别潜在因素的研究人员来说，探索性因素分析和共同因素模型是合适的分析方法。对于使用李克特量表的调查，建议使用加权最小二乘法和稳健标准误差，并进行斜交旋转。还简要讨论了其他数据分析技术，例如项目反应理论和机器学习。最后，为研究人员和评论员提供了一份基本的检查表。
