2.4. Statistical analysis
The effects of demographic characteristics and lifestyle factors on perceived musculoskeletal pain symptoms and psychological distress were explored through a series of analysis of variance (ANOVA) and regression models conducted using SPSS (SPSS version 21; SPSS Inc., Chicago, IL, USA). All regression analyses controlled for age, gender, job category, and experience, with categorical variables were recoded to have two levels (i.e., job type was coded as 1 = operator and -1 = other occupations; gender was coded as 1 = men and -1 = women). The models were grouped as follows: (1) one-way ANOVAs that examined the effects of job category, age, experience, and gender on both psychological distress and pain regions; (2) a linear regression exploring the impact of job category and number of pain regions on psychological distress scores; (3) one-way ANOVAs that investigated the impact of interference with work and absenteeism due to either neck, shoulder, lower back, or wrist/hand bodily pain on psychological distress scores; (4) a linear regression exploring the possible moderating effects of exercise or BMI on the relationship between musculoskeletal pain symptoms and psychological distress; and (5) a linear regression examining whether perceived sleep quality mediated the relationship between musculoskeletal pain symptoms and psychological distress.