This study investigated relationships of age, gender, and supervisor level with job satisfaction, organizational commitment, stress perception, and turnover intention. The demographics were hypothesized to moderate the stress-satisfaction and commitment-turnover relationships. Hypotheses were tested using both parametric and non-parametric bootstrap methods. Subjects were taken from a national survey of 2,663 public sector IT workers. Missing data were imputed using NORM software. Ordinary least squares (OLS) regression indicated a significant direct effect from all main variables and covariates, except for age on turnover intent. No mediating effects were found. Age-Commitment was the only significant higher order modifier relationship, although Gender-Commitment explained substantial variance. LMG statistic results enabled the predictors to be rank ordered with confidence intervals. Best subset bootstrap regression explored all possible predictor orders to confirm which model explained the most variance. The original model and predictor sequence were confirmed. The bootstrap AIC statistic provided a model which maximized explained variance while optimizing parsimony. Since only age had a mediating effect, Hypotheses 1 and 2 were not supported. All other hypotheses were partially confirmed.