主观负荷评估技术和NASA任务负荷指数量表的信度与效度评价

目的对2个脑力负荷评价量表即主观负荷评估技术(SWAT)和NASA任务负荷指数(NASATLX)量表的信度和效度进行验证。方法采用整群随机抽样方法抽取科研、教育、管理、医疗卫生等行业的脑力劳动者共1268例组成样本,采用重测信度、分半信度、Cronbach’sα系数、条目分与总分之间的相关系数对此两量表进行信度验证,采用结构效度评价两量表的效度。结果两量表总分及各条目的重测信度在0.516～0.753之间,差异有统计学意义(P<0.01),具有较好的重测信度;SWAT量表的分半信度为0.645,Cronbach’sα系数>0.80,各条目分与总分之间的相关系数均>0.70,差异有统计学意义(P<0.01);NASATLX量表的分半信度、Cronbach’sα系数均>0.80,各条目(业绩水平除外)与总分之间的相关系数均>0.60,差异亦有统计学意义(P<0.01),两量表的内部一致性较好;两量表总分的Pearson相关系数为0.492,两量表测定结果一致性较好。因子分析结果表明两量表具有较好的结构效度。结论SWAT和NASATLX量表均具有较好的信度与效度,经过适当修订,可作为我国脑力负荷研究的有效工具。

The appraisal of reliability and validity of subjective workload assessment technique and NASA-task load index

OBJECTIVE: To test the reliability and validity of two mental workload assessment scales, i.e. subjective workload assessment technique (SWAT) and NASA task load index (NASA-TLX). METHODS: One thousand two hundred and sixty-eight mental workers were sampled from various kinds of occupations, such as scientific research, education, administration and medicine, etc, with randomized cluster sampling. The re-test reliability, split-half reliability, Cronbach's alpha coefficient and correlation coefficients between item score and total score were adopted to test the reliability. The test of validity included structure validity. RESULTS: The re-test reliability coefficients of these two scales and their items were ranged from 0.516 to 0.753 (P < 0.01), indicating the two scales had good re-test reliability; the split-half reliability of SWAT was 0.645, and its Cronbach's alpha coefficient was more than 0.80, all the correlation coefficients between its items score and total score were more than 0.70; as for NASA-TLX, both the split-half reliability and Cronbach's alpha coefficient were more than 0.80, the correlation coefficients between its items score and total score were all more than 0.60 (P < 0.01) except the item of performance. Both scales had good inner consistency. The Pearson correlation coefficient between the two scales was 0.492 (P < 0.01), implying the results of the two scales had good consistency. Factor analysis showed that the two scales had good structure validity. CONCLUSION: Both SWAT and NASA-TLX have good reliability and validity and may be used as a valid tool to assess mental workload in China after being revised properly.