ZJUTADAC的个人博客分享 http://blog.sciencenet.cn/u/ZJUTADAC

博文

常见问题三:未能预先定义假设及其作用形式

已有 138 次阅读 2026-2-17 08:34 |个人分类:论文分享|系统分类:科研笔记

Common Error 3: Failure to Predefine Hypotheses and Their Functional Forms

Moderation analyses using NHST require a priori hypothesis specification, as they test predetermined expectations (Becker et al., 2023; Dawson, 2014; Rasoolimanesh et al., 2021). Our review found that 34.67% of studies omitted or presented incomplete hypotheses, contrasting with Rasoolimanesh et al.’s (2021) finding of 5%. This oversight risks theoretical rigor. Researchers should specify how the moderator alters the X→Y relationship—including its functional form (e.g., strengthening, weakening, reversing) and scope—prior to testing. Vague statements (e.g., “M moderates the effect”) are insufficient, especially for categorical moderators. All hypothesized mechanisms require explicit theoretical justification before testing.

常见问题三:未能预先定义假设及其作用形式

调节分析若采用零假设显著性检验(NHST),其本质是检验一个预先设定的预期(Becker等,2023)。因此,在实证检验之前,明确陈述关于调节效应的具体假设是基本要求。然而,我们的系统综述发现,34.67%的研究要么完全省略了假设,要么假设陈述不完整。

一个常见的、但不够充分的表述是:“M(正向或负向)调节XY的影响。这种表述过于模糊,尤其是在调节变量(M)是分类变量(如性别)时。它没有说明调节的具体作用形式”——M是如何改变X→Y关系的(是增强强度、减弱强度,还是逆转方向?),以及其影响范围(适用于整个模型还是部分情境?)。

问题实质:

模糊的假设等于没有假设。事后从显著的结果中反推出一个解释,违背了假设检验的逻辑,降低了研究的理论严谨性,并使结果难以被准确解释和复制。

解决建议:

  • 提出先验的、具体的假设:在数据分析前,就应基于理论,明确假设调节变量将如何改变核心关系。

  • 完整定义函数形式与范围:一个规范的假设应像示例中那样清晰:在完整模型中,财务支持对消费者情感承诺的影响,在女性中比在男性中更强。这明确指出了调节变量(性别)、影响方向(更强)及适用范围(完整模型)。

  • 区分探索性与验证性分析:若研究确实是探索性的,应明确声明,并避免使用验证性分析的框架和语言来报告结果。

Reference

  • Becker, J. M., Cheah, J. H., Gholamzade, R., Ringle, C. M.,      & Sarstedt, M. (2023). PLS-SEM's most wanted guidance. International      Journal of Contemporary Hospitality Management, 35(1), 321-346.

  • Dawson, J. F. (2014). Moderation in management research: What,      why, when, and how. Journal of Business and Psychology, 29(1),      1-19.

  • Mertens, W., & Recker, J. (2020). New guidelines for null      hypothesis significance testing in hypothetico-deductive IS research. Journal      of the Association for Information Systems, 21(4), 1072-1102.

  • Rasoolimanesh, S. M., Wang, M., Mikulić, J., &      Kunasekaran, P. (2021). A critical review of moderation analysis in      tourism and hospitality research toward robust guidelines. International      Journal of Contemporary Hospitality Management, 33(12), 4311-4333.

  • Xu, Y., & Shiau, W. L. (2026). Moderation analysis in      business and management research: Common issues, solutions, and guidelines      for future research. International Journal of Information Management86,      102995.



https://blog.sciencenet.cn/blog-3444471-1522539.html

上一篇:常见问题二:调节假设缺乏充分的理论论证
收藏 IP: 36.230.14.*| 热度|

0

该博文允许注册用户评论 请点击登录 评论 (0 个评论)

数据加载中...
扫一扫,分享此博文

Archiver|手机版|科学网 ( 京ICP备07017567号-12 )

GMT+8, 2026-2-17 17:20

Powered by ScienceNet.cn

Copyright © 2007- 中国科学报社

返回顶部