Design and study population
This cross-sectional study was conducted using employees working in an industrial estate during the summer and fall of 2021. All workers with at least 1 year of work experience were initially invited and a total of 286 out of the 462 workers (response rate: 61.9%) participated in the study. In the initial stage, 16 questionnaires were removed from the study due to missing data, misleading responses and authentication responses. Therefore, 270 questionnaires were statistically analyzed. The participants completed the survey anonymously after providing informed consent. Participants were aware of the purposes and hypotheses of the study. The study was conducted in 2021, according to the principles of the Declaration of Helsinki, and was approved by the Research Ethics Committee of Shiraz University of Medical Sciences (IR.SUMS.REC.1398.537).
The purpose of the study, how to answer the questions, and the ethical obligations of the researchers regarding the completed questionnaires was explained to the employees. Then, a questionnaire was distributed to each consenting participant with information on how to return it to the researchers after completing it. Participation in this research was entirely voluntary, the questionnaires were anonymous, and the results were analyzed in terms of the information of all respondents.
Translation and cross-cultural adaptation
To respect the intellectual property rights of the instrument, first, the approval of the original authors was obtained for translation into Persian. Based on the process proposed by Beaton et al. [21] for translation and cross-cultural adaptation, forward translation was performed by two ergonomics experts fluent in the English language. The research team then compared the translated Persian versions, and differences and ambiguities were discussed. Afterward, translated items were reviewed in terms of semantics, cultural adaptation, and terminology. Eventually, a single Persian provisional version was developed. The next step was the backward translation of the Persian provisional version by two language experts fluent in English unaware of the original English version. The research team revised the translated versions, and a single English provisional version was obtained by combining them. Afterward, this version, along with ambiguities and inconsistencies, was sent to the developers of the WDQ for clarification and conditions of acceptability. The proposed revisions were made, and the final version was prepared for psychometric evaluations. The Persian version is available on the developer Prof. Morgeson’s questionnaire website (http://www.morgeson.com/wdq.html).
Measuring validity and reliability
Face and content validity
Face validity is the extent to which a test can cover the concept it intends to measure [22]. In this study, to achieve face validity we sought the opinions of 10 university professors (ergonomics and industrial and organizational psychology). That is, the professors were asked to evaluate each of the items in terms of comprehensibility, wording, interpretations, cultural issues, and clarification. In addition, a sample of 15 employees from the participant pool was selected to assess the questionnaire in terms of potential ambiguities and understandability of items. After minor revisions, content validity was evaluated using the content validity index (CVI). Ten experts in ergonomics, occupational health, and industrial and organizational psychology were asked to comment on each item separately [22]. For CVI, a score higher than 0.79 is favorable, 0.7–0.79 indicates the need for revision, and less than 0.7 is unacceptable [23].
Construct validity
Construct validity of the questionnaire was assessed using confirmatory factor analysis (CFA) based the maximum likelihood estimation (MLE) method. The model fit was assessed using the root mean square error of approximation (RMSEA), the comparative fit index (CFI), the incremental fit index (IFI) and the chi-square/degrees of freedom ratio (χ2/df). If the RMSEA is less than 0.08, CFI value is 0.9 or higher, the IFI value is 0.8 or 0.9, and the χ2/df is less than 3, then the fit of the model is appropriate [24].
Based on previous studies, five models of 4, 18, 19, 20, and 21 factors were examined in this study. The 4-factor model contains 4 wide categories of work characteristics, including those related to task, knowledge, social, and work context. The 18-factor model covers the work characteristics without any divisions, including autonomy and interdependence, each as a unique factor. The 19-factor model divides interdependence into 2 components of received and initiated. The 20-factor model contains 3 elements of autonomy, including work scheduling, decision making, and work methods. The 21-factor model also includes identified components containing interdependence and autonomy.
Incremental validity
Incremental validity of the WDQ was assessed using Perceived Stress Scale (PSS-10) and Michigan Organizational Assessment Questionnaire Job Satisfaction Subscale (MOAQ-JSS). The PSS measure the degree to which situations are appraised as stressful during the past month [25]. It uses a 5-point Likert scale, ranging from 0 (never) to 4 (always). Three versions of PSS are available that comprise 4, 10, and 14 items. Most prior studies have used the 10-item version, as it has fewer items than the original 14-item version, with similar psychological properties [26, 27]. The MOAQ-JSS measures job satisfaction using three items using a 7-point Likert scale [28].
Convergent/discriminant validity
The average variance extracted (AVE) and the maximum shared squared variance (MSV) were used to assess convergent validity and discriminant validity, respectively. The AVE is “the average variance extracted is calculated as the mean variance extracted for the item loading on a [factor] and is a summary indicator of convergence” and values ≥ 0.5 confirms the convergent validity [29]. On the other hand, discriminant validity is confirmed when the AVE is greater than the MSV [29].
Reliability
Cronbach’s alpha coefficient and composite reliability (CR) were used to assess the internal consistency of the WDQ. A Cronbach’s alpha and CR value of 0.70 or higher is considered acceptable [30]. In addition, the item-to-total correlation and Cronbach’s alpha if item deleted were also calculated separately. For each item, the item-to-total correlation should be higher than 0.3 to be considered acceptable [31].
Statistical analysis
Data analysis was conducted using SPSS version 23 and AMOS version 23 (SPSS Inc, Chicago, Ill, USA). The Kolmogorov–Smirnov test was applied to test for normal distribution. Also, Mardia’s coefficient (based on critical ratio) was used to determine multivariate normality. Critical ratio values were significantly < 5, which indicated that the data met assumptions of normal distribution [32]. A P value < 0.05 was considered statistically significant.