2.4. Statistical analysisAll data were analyzed with SPSS 16.0 for Windows. We calculatedboth the person-based incidence rate of injury and theevent-based incidence rate of injury. The person-based incidencerate referred to the number of workers with injury among all participatingworkers and the event-based incidence rate referred tothe number of work-related injury events among all workers (oneworker could have several injury events in a year). The prevalencerates of MSD in different body parts were calculated for all participants,as well as by industry.Binary logistic analysis with backward stepwise method wasused to estimate odds ratios (OR) and 95% confidence interval (CI)of factors associated with injury cases and MSDs. Variables includedin the full model were gender, age, educational level, industry type,duration of employment, work hours per week, mental stress atwork, previous work experience, previous history of injury. Onlyvariables with p value <0.05 would be retained in the final models.Categorical variables were put into model as dummy variables,using the first or the lowest category as the reference group to estimatethe ORs of the other categories. For example, we used the“Primary school” as the reference group for the variable of educationlevel. The dummy variables were created: D1 = 1 if educationallevel is middle school, and equals zero otherwise, D2 = 1 if educationallevel is high school, and equals zero otherwise, D3 = 1 ifeducational level is university or above, and equals zero otherwise.The dummy variables were also created for age group, duration ofemployment, work hours per week, and mental stress at work.
การแปล กรุณารอสักครู่..