Jiang X W, Liu J N, Liu F J, Zheng Z J, Chang C
School of Public Health, Peking University, Beijing 100191, China.
Zhonghua Lao Dong Wei Sheng Zhi Ye Bing Za Zhi. 2020 Dec 20;38(12):898-902. doi: 10.3760/cma.j.cn121094-20191114-00529.
To investigate the reliability and validity of Stanford attendance scale (sps-6) in the study of attendance among professional groups. In August, 2018, the 1455 employees from 81 workplaces in Beijing, Shanghai, Jiangsu and Guangdong were randomly investigated as the subjects. The reliability and validity of sps-6 were analyzed by using the internal consistency reliability (Cronbach's coefficient) , half split half coefficient, content validity, integration validity, discrimination validity, cluster analysis and structural validity analysis. Cronbach's coefficients of sps-6 scale, working process and work results were 0.692, 0.918 and 0.907, respectively; Guttman of scales and dimensions The split half coefficients were 0.792, 0.803 and 0.794, respectively; Pearson correlation coefficients of the total score of each item and scale were 0.526-0.673 (<0.01) ; the qualification rate of set validity and differentiation validity were 100%; the results of cluster analysis supported the theoretical basis for the formation of the scale. The general non-standard fitting index (TLI) =0.982, approximate error mean square root mean square (RMSEA) =0.071, comparative fit index (CFI) =0.990, fit goodness index (GFI) =0.987, modified fit goodness index (AGFI) =0.965, Norm fit index (NFI) =0.990. The results showed that the scale had higher structural validity, and the results of sps-6 in the occupational population were (21.36±4.04) , and the distribution was normal (deviation was 0.053, peak was 0.023) . The scores of sps-6 scale were statistically different in various charactoristics of gender, age, education level, marital status, annual income, position, position level and industry (< 0.01) . Stanford attendance scale has high reliability and validity, and can be applied to the study of attendance in professional groups.
探讨斯坦福出勤量表（sps - 6）在职业群体出勤情况研究中的信效度。2018年8月，随机抽取北京、上海、江苏和广东81个工作场所的1455名员工作为研究对象。采用内部一致性信度（克朗巴哈系数）、分半系数、内容效度、整合效度、区分效度、聚类分析及结构效度分析等方法对sps - 6的信效度进行分析。sps - 6量表、工作过程及工作结果的克朗巴哈系数分别为0.692、0.918和0.907；量表及维度的古特曼分半系数分别为0.792、0.803和0.794；各条目总分与量表的Pearson相关系数为0.526 - 0.673（<0.01）；设定效度和区分效度的合格率均为100%；聚类分析结果支持量表形成的理论依据。一般非标准拟合指数（TLI） = 0.982，近似误差均方根（RMSEA） = 0.071，比较拟合指数（CFI） = 0.990，拟合优度指数（GFI） = 0.987，修正拟合优度指数（AGFI） = 0.965，规范拟合指数（NFI） = 0.990。结果表明该量表具有较高的结构效度，职业人群中sps - 6的得分情况为（21.36±4.04），且分布呈正态（偏度为0.053，峰度为0.023）。sps - 6量表得分在性别、年龄、文化程度、婚姻状况、年收入、职位、职位层级及行业等各特征方面差异有统计学意义（<0.01）。斯坦福出勤量表具有较高的信效度，可应用于职业群体出勤情况的研究。
