“…The moderated regression model of facet satisfaction was proposed in response to the flaws identified in the simple discrepancy (have-want) model of job satisfaction, which despite its "seductive face validity" (Johns, 1981, p. 443) has been discredited on both theoretical and statistical grounds (Cronbach & Furby, 1970;Evans & Ondrack, 1991;Johns, 1981;Wall & Payne, 1973). This moderated regression approach is now widely accepted as a valid model of facet satisfaction (e. g., Butler, 1983;Rice, Gentile, & McFarlin, 1991;McFarlin & Rice, 1992;Rice, McFarlin, & Bennett, 1989), despite some limited criticism on statistical grounds (Morris, Sherman, & Mansfield, 1986).…”