4.1. Identifying Relationships among Factors: ISM Approach
This study aimed to explore the use of ISM to recognize the associations among the factors that affect the ergonomic assessment of an industry.
ISM is used as a tool in decision making, in which a group of factors affecting the application of a system are identified in an organized manner. ISM utilizes the practical experience of professionals currently working in the area to develop a multilevel organizational model that emphasizes the key aspects involved in the application of a specific system [42
]. These features make this a more suitable tool for the present study.
The working mechanism of the ISM technique is as follows [42
A circumstantial relationship between the factors and the structural self-interaction matrix (SSIM) of factors is developed. This matrix indicates the pairwise association among the factors analyzed, as presented in Table 3
. To indicate the direction of the association between two factors (i and j), the following symbols are defined:
V denotes that factor i affects factor j;
A denotes that factor j impacts factor I;
X denotes that factors i and j affect each other;
X denotes that factors i and j are unrelated.
The SSIM is developed using expert opinions and is improved to create a reachability matrix by changing each entry in the SSIM into binary digits 1 and 0.
The initial reachability matrix is derived from the SSIM matrix. After checking the transitivity, the final reachability matrix is obtained, and the final reachability matrix is partitioned into different levels by listing the factors in the antecedent set, intersection set, reachability set, and partition level, as presented in Table 4
A directed graph of the interactions listed in Table 4
is then drawn. The ISM model thus constructed is shown in Figure 2
The model shown in Figure 2
is subsequently used for SEM analysis.
4.2. SEM Validation of the Relationships among Factors
SEM analyzes the hypotheses for unmeasurable variables by considering independent and dependent variables using statistical analysis [42
]. The theoretical model obtained from the ISM was validated using PLS-SEM. After a careful literature review, a questionnaire was used in discussions with experts working in the automotive industry. The questionnaire was administered via email to 150 experts from different firms. At the end of the survey, 101 complete and usable responses were received from 40 automotive manufacturing firms. This represents a response rate of 67.30%, which is sufficient for this empirical study [43
]. This sample size is adequate for evaluating the hypotheses developed in this study [43
]. Our respondents consisted of an ergonomics engineer, head supply chain managers, maintenance engineers, operations managers, and risk assessment specialists. Of the respondents, 67% were between 27 and 50 years of age, and only 12% were over 50 years of age. The model representation of the factors, along with the tests of the hypotheses considered in this study, are shown in Figure 3
The hypotheses tested in this study are described below:
Hypothesis 1 (H1).
Psychological factors and the ergonomic assessment are correlated.
Hypothesis 2 (H2).
The ergonomic assessment and physiological factors are correlated.
Hypothesis 3 (H3).
The ergonomic assessment and safety factors are correlated.
Hypothesis 4 (H4).
Psychological factors and physiological factors are correlated.
Hypothesis 5 (H5).
Environmental factors and physiological factors are correlated.
Hypothesis 6 (H6).
Safety factors and physiological factors are correlated.
Hypothesis 7 (H7).
Environmental factors and safety factors are correlated.
The convergent validity was evaluated by assessing Cronbach’s α, the average variance extracted (AVE), and the composite reliability, which is used to analyze the level of correlation/convergence of different variables of the same construct. The composite reliability is greater than the acceptable limit of 0.7 for all constructs, and the convergent validity [44
] of the constructs is confirmed if the AVE is greater than or equal to 0.50. Moreover, the Cronbach’s α value for the ergonomic factors is above 0.5, and hence, the factors are found to be reliable, as indicated in Table 5
In this study, discriminant validity measures were evaluated using a confirmatory factor analysis. It can be found that correlation values of factors such as physiological, psychological, environmental, and safety were less than the square root of the AVE [45
]. Hence, it was found that the factors considered in this model have discriminant validity, as indicated in Table 6
The bootstrapping test (Table 7
) was performed using the PLS-SEM to determine the t
-test values. Based on the bootstrapping results presented in the Table 7
. Materials and the hypothesis testing, all of the hypotheses were supported by the survey results (p
The standard errors and t
-test values are listed in Table 8
Based on the “t” statistic, it can be inferred that there is a significant correlation between each of the verified factors and the ergonomic practices. Among these, the physiological factors have very high correlation with the ergonomic practices in the manufacturing industry, with a “t-stat” result of 49.652. It is also observed that there is a significant correlation between psychological and physiological factors, with a “t-stat” value of 107.55.
4.3. Determination of Human Factor Index Measurement
The ergonomic factors and their relationships, which were validated using the ISM-SEM approach, are considered to measure the human factor index using a multigrade fuzzy approach. The multigrade fuzzy approach overcomes the drawbacks associated with conventional crisp approaches. The major advantage of using the multigrade fuzzy method is that the average weights are multiplied by the responses given by experts for each variable. This method avoids fluctuations in the values, and extreme responses and their biases are neutralized as constant weights in the calculation. Fuzzy approaches provide a useful tool to deal with problems in which attribute phenomena are imprecise and vague. Most ergonomic measurements are characterized by ambiguity and multiple possibilities. The scoring of existing techniques is always criticized because the scales used to score ergonomics have limitations. In this context, a multigrade fuzzy approach was used to evaluate ergonomics in the automotive industry.
The application of the multigrade fuzzy approach was validated through a case study. The case study was conducted in an original equipment manufacturer (Sarang Auto Parts Pvt. Ltd., Chennai, India) located in Chennai, India. This organization produces a shaft lower link, nut-driving pinion, bearing lock, rod hydraulic lift connection, and support reverse cluster. They manufacture the following parts under the broad category of automotive components: electrical parts, drive transmission and steering parts, nut-driving pinion, bearing locks, rod hydraulic lift connections, shaft power take-off (PTO) drives, shaft lower links, support reverse clusters, and shaft front PTOs.
The human factor index (HFI) of a workplace is represented by the multiplication of the weight (W) and assessment factor (F).
The assessment has been divided into five scales because every ergonomic factor involves the fuzzy determination I = (10, 8, 6, 4, 2). “Excellent”, “good”, “fair”, “poor”, and “very poor” ergonomic levels correspond to scores of 8–10, 6–8, 4–6, 2–4, and 0–2, respectively. Four experts (L1
) were involved in the ergonomic evaluation discussion. The weightages assigned for each enabler is 0.4 for physiological factors (HFI1
), 0.2 for psychological factors (HFI2
), 0.2 for environmental factors (HFI3
), and 0.2 for safety factors (HFI4
). The physiological factors include the following criteria, including biomechanical aspects (HFI11
), and energy expenditure (HFI13
). The weightages assigned for each criterion are 0.5, 0.25, and 0.25, respectively. Similarly, the distribution of fuzzy weighting to the ergonomic attributes, criteria, and enablers is summarized in Table 9
4.3.1. Primary Assessment Calculation
The calculations pertaining to biomechanical aspects are shown below.
The index pertaining to biomechanical aspects is given by HFI11 = W11 × F1.
HFI11 = (7.6, 8.2, 8.4, 7.4)
Similarly, the index for each criterion is calculated and listed below.
HFI12 = (8.4, 8, 8.4, 7.8)
HFI13 = (7, 7, 6, 6.5)
HFI21 = (8.5, 7.25, 8.25, 8.25)
HFI22 = (8.5, 9, 9, 9)
HFI23 = (8.25, 7, 7.75, 7.75)
HFI24 = (9, 8.5, 9, 8)
HFI31 = (7.5, 8, 8, 6.5)
HFI32 = (7.5, 8, 8, 7.5)
HFI33 = (6, 6, 6.5, 5.5)
HFI34 = (8, 7.5, 8, 7.5)
HFI35 = (9, 8, 9, 8)
HFI41 = (7.5, 8, 6, 5.5)
HFI42 = (9, 8, 9, 8)
HFI43 = (8.5, 8, 8.5, 8.5)
4.3.2. Secondary Assessment Calculation
The index pertaining to physiological factors is calculated as shown below.
Similarly, the indexes for other enablers are calculated below.
HFI2 = (8.56, 7.93, 8.5, 8.25)
HFI3 = (7.2, 7.125, 7.55, 6.625)
HFI4 = (8.335, 8, 7.88, 7.345)
4.3.3. Tertiary Assessment Calculation
Finally, the total HFI is calculated as shown below.
The HFI is the average of (7.879, 7.791, 7.862, 7.612), which is equal to 7.786.
HUMAN FACTOR INDEX = 7.786 ϵ (6, 8)
A human factor index of 7.86 was determined using a multigrade fuzzy approach, which means that the organization is ERGONOMIC.