Use for measurement rigor in 《南开管理评论》 (Nankai Business Review) survey studies — reliability (Cronbach's α, composite reliability), validity (convergent via AVE, discriminant via Fornell-Larcker / HTMT), and common-method-bias diagnosis (Harman single-factor, marker variable, common latent factor). Use whenever a paper uses self-report scales, especially if measures were borrowed or single-source.
Scanned 6/6/2026
Install to Claude Code
npx -y skills add brycewang-stanford/Awesome-Journal-Skills --skill nbr-measurement --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Nbr Measurement?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/brycewang-stanford-nbr-measurement)More formats (shields.io, HTML) on the badges page.
---
name: nbr-measurement
description: Use for measurement rigor in 《南开管理评论》 (Nankai Business Review) survey studies — reliability (Cronbach's α, composite reliability), validity (convergent via AVE, discriminant via Fornell-Larcker / HTMT), and common-method-bias diagnosis (Harman single-factor, marker variable, common latent factor). Use whenever a paper uses self-report scales, especially if measures were borrowed or single-source.
---
# 测量规范(nbr-measurement)
## 触发时机
- 用了量表(个体或企业层级自报)
- 量表照搬国外/他文,未做本土化与心理测量
- 单一来源、同一时点测自变量与因变量(CMV 高风险)
## 信度(Reliability)
- **Cronbach's α** ≥ 0.70(探索性 0.60 可商榷,需说明)
- **组合信度 CR** ≥ 0.70
- 报告每个构念的题项数、均值/标准差、来源出处
## 效度(Validity)
| 类型 | 指标 | 经验门槛 |
|------|------|----------|
| 收敛效度 | 标准化载荷显著且较高;**AVE** | 载荷 > 0.50(宜 > 0.70);AVE > 0.50 |
| 区分效度 | Fornell-Larcker:√AVE > 构念间相关 | 或 **HTMT** < 0.85(严格 0.90) |
| 内容效度 | 量表来源、专家评议、预测试 | 翻译-回译,本土化适配 |
| 结构效度 | 验证性因子分析(CFA)拟合 | 见 nbr-survey-sem 拟合标准 |
## 共同方法偏差(CMV / CMB)
自报、单源、同时点测量必须**事前防范 + 事后检验**:
- **事前**:自变量/因变量分时点或分来源;匿名;反向题;不同量尺
- **事后(任选并报告)**:
- **Harman 单因子**:未旋转首因子方差 < 50%(最弱证据,别只用它)
- **标记变量(marker variable)**:引入理论无关变量做局部调整
- **共同潜因子(CLF / ULMC)**:加方法因子,比较载荷变化
- 结论要写明"CMV 不构成实质威胁"的依据,而非一句"做了 Harman"
## 自检清单
- [ ] 每个构念报告 α 与 CR(≥ 0.70)
- [ ] 报告 AVE(> 0.50)证收敛效度
- [ ] 用 Fornell-Larcker 或 HTMT 证区分效度
- [ ] 量表注明来源、是否本土化、预测试
- [ ] 自报数据做了 CMV 事前设计 + 至少一种事后检验(勿只靠 Harman)
- [ ] 控制变量的测量同样交代清楚
## 反模式
- 只报 α,不报 CR/AVE
- 借用量表不验证、不本土化,直接套
- CMV 只做 Harman 单因子就宣称"无偏差"
- 区分效度只看相关系数,不与 √AVE 比较
## 输出格式
```
【信度】构念 α / CR 列表(最低值 <…>)
【收敛】AVE 列表(最低 <…> 是否 > 0.50)
【区分】Fornell-Larcker / HTMT:通过□ 风险□
【量表来源】本土化□ 预测试□
【CMV】事前设计:<…>|事后:Harman <%> + <标记/CLF> |结论:<…>
【下一步】nbr-survey-sem
```
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!