Study On Evaluation And Its Application Of Mental Worker's Mental Workload

Owing to technological advancements and the wide use of computers, people are facing to growing information quantity and complexity. Thus mental work is becoming the primary work type of most workers. Mental work is a process in which people receive information, then translate, edit, trim, analyze and at last make response to it. Demand of more mental work in task consequentially results to increase of mental workload. Moderate mental workload is necessary to fulfilling task and workers' health, but too high mental workload will influence their health and work performance. So correct evaluation on workers' mental workload and keeping it moderate are very important for protecting workers' health and increasing their work efficiency.Research on mental workload in West has been carried on for more than twenty years while in our country it is just underway, most of which was experimental or simulative study, but little epidemiological study was reported. The objectives of the study included the following aspects: the first was to translate, edit and revise sclales, including Subjective WorkloadAssessment Technique (SWAT) and National Aeronautics and Space Administration -Task Load Index (NASA-TLX), which were the two wide-used subjective rating scales of mental workload in West, and then test their reliability and validity. The second was to evaluate the mental workload of mental workers from various kinds of occupations and select the relatively preferable scale of mental workload to reveal influential factors of mental workers' mental workload. The last was to explore the relationships between occupational stress, work ability or mental capacity and mental workload.1 Evaluation of reliability and validity of SWAT and NASA-TLXIn order to test reliability and validity, 1268 mental workers were sampled from various kinds of occupations with the random cluster sampling method, 153 and 162 of whom were retested with SWAT and NASA-TLX respectively after 2 to 3 weeks. The re-test reliability, split-half reliability, Cronbachs alpha coefficient, theta coefficient, omega coefficient and correlation coefficients between item score and total score were adopted to assess the reliability. The test of validity included the assessment of structure validity and discriminant validity. According to the character of the value of its items, SWAT was divided into two types -Discrete SWAT (Dswat) and Continuous SWAT (Cswat).The reliability analysis showed that the re-test reliability of the total scores of the above scales and the scores of their items was good or excellent, raged from 0.509 to 0.775 (P<0.001). The split-half reliability ofDswat, Cswat and NASA-TLX was 0.708, 0.804 and 0.808 respectively. Their Cronbach's alpha coefficient (ranged from 0.693 to 0.784), theta coefficient and omega coefficient (ranged from 0.693 to 0.880) conformed to the requirements of psychometric study. The total scores of the above scales correlated significantly with the scores of their relevant items (P<0.001) and the correlation coefficients were higher than 0.55 except that between the total scores of NASA-TLX and the score of its item of performance. These results suggested that the above scales had good inner consistency. The correlation coefficients between the total scores of the above scales implied the results of these scales had good consistency (ranged from 0.459 to 1.000, P<0.001).The validity analysis revealed that the items of SWAT and NASA-TLX (except the item of performance) had good discriminant validity enough to distinguish different level of mental workload (P<0.001). Factor analysis showed that one common factor was drawn from Dswat and Cswat respectively, whose rate of variance was 61.931% and 73.736% respectively. Two common factors were drawn from NASA-TLX, their cumulative rate of variance was 67.214%. The communality of each item was over 0.50 (0.555-0.790), its factor loading attached to the relevant common factor was more than 0.60 (0.691-0.889). The factors with high rate of variance were consistent with the actual theoretic frame of scales on the whole. Therefore these scales had good structure validity.As we knowed from the above, the item of performance in NASA-TLX showed poor discriminant validity, its score was lowcorrelated with the total score and it couldn't have high factor loading attached to the same factor as the other items, suggesting that the item was not one of the constituent of mental workload and should be deleted from the scale. The analysis of the revised NASA-TLX in which the item of performance was deleted revealed that its split-half reliability, Cronbach's alpha coefficient and theta coefficient showed a little higher (0.829, 0.790 and 0.807 respectively), omega coefficient was 0.844, the correlation coefficients between the total scores and the scores of its items were ranged from 0.516 to 0.856 (P<0.001), each item could distinguish different level of mental workload (P<0.001). Factor analysis displayed that one common factor was drawn, whose rate of variance was 56.404%, higher than that of the first common factor of the un-revised scale. The factor loading of each item attached to this factor was more than 0.50 (ranged from 0.576 to 0.847).Generally speaking, NASA-TLX had relatively higher reliability than SWAT. KMO measure of sampling adequacy test showed that the KMO value of NASA-TLX was more great, implying that more commonness and less partial correlation exist among the items of this scale. Moreover, the reliability and validity of NASA-TLX based on Principal Component analysis (PCtlx) was relatively preferable.2 Study on mental workload and its influential factors of mental workersIn this cross sectional study, mental workload was measured with the revised PCtLx among 1285 mental workers employed in various kinds ofoccupations to investigate the distribution character and influential factors of their mental workload.Normality test revealed that the data of mental workers' mental workload displayed a profile approximate to normal distribution (P>0.05). Male mental workers had significantly higher mental workload level than female according to the results of one-way analysis of variance or analysis of covariance (P<0.01). Age and length of service showed evidently effects on mental workload. The level of mental workload had a trend to increasing with age and length of service increasing (P<0.01), and the increase of mental workload with age increasing was faster in female workers than male. Mental workload among shift workers was significantly higher than daytime workers (P<0.01). Work hours per day, educational degree and economical level showed significant effects on mental workload (P<0.05). Workers engaged in different occupations had different mental workload (f<0.01), in whom culture workers, teachers, science researchers and administrators had relatively higher mental workload.Results of linear regression analysis showed that age, length of service, educational degree, work hours per day and econo.mical level were the influential factors of mental workload (P<0.05). The partial regression coefficients of age, length of service, educational degree, work hours per day were positive while that of economical level was negative. According to the absolute value of standardized partial regression coefficient, work hours per day had the most effect on mental workload, especially whenlength of work per day exceeded 8 hours. Therefore, too many work hours per day may be one of the important risk factors that would bring to the increase of mental workload so that reducing work hours per day, especially keeping it in 8 hours, was the significant and valid measure of alleviating mental workers' mental workload.3 Relationships between occupational stress, work ability or mental function and mental workloadOccupational stress, work ability and mental workload of 245 mental workers were measured by Occupational Stress Inventory Revised Edition (OSI—R), questionnaire of Work Ability Index (WAI ) and the revised PCtlx to explore the effects of mental workload on occupational stress and work ability. Mental capacity was measured using 4 indices - short-time memory, decode, AYP index and mental functional index (MFI) in 99 workers at the same time their mental workload was measured to study the effect of mental workload on mental capacity in mental workers.Results demonstrated that the increase of mental workload was reflection of aggravation of occupational role. When mental workload was not more than certain level, occupational stress level was getting lower while personal coping resources and work ability getting higher with the increase of mental workload. But when mental workload exceed that level, occupational stress level was getting higher while personal coping resources and work ability getting lower with the increase of mental workload. Linear regression analysis and analysis of covariance suggested that mental workload had no significant effects on every index of mentalcapacity on the condition that the confounding factors that may effect mental capacity such as age, gender, and educational level etc were controlled.To sum up, either too high or too low mental workload may reduce personal coping ability and enhance occupational stress level so as to decline work ability of mental workers. Hence moderate mental workload was helpful for keeping high coping ability but lowering occupational stress level, consequently keeping good work ability. The influential factors of mental workload were various, therefore it is an important task that occupational researchers would be faced with how to take effectual interventional measures and optimize working conditions to protect workers keeping moderate mental workload.