Most existing studies of website characteristics and their impact on intentions have assumed a linear relationship between them. However, motivational theories suggest that while some characteristics have linear effects, others will act as either hygiene factors or motivator factors and have non-linear effects. Drawing on motivational theory and WebQual constructs, this study looks at different methods of testing for hygiene and motivator factors and proposes a different technique that could clarify misleading results. Linear regression, quadratic regression, and piecewise regression are used in the study. The results show the hygiene effect of trust and suggest that piecewise linear regression may be superior to quadratic regression in this context.