Abstract

Three different models of burnout evolution (Leiter, 1990; 1993; Leiter and Maslach, 1988) have been examined (Lisrel VII) among a panel of 140 trained nurses, questioned four times in the first two years of working activity. The findings suggest that none of the patterns analysed properly explain the covariation structure observed in the sample. A new model is then proposed (T = 37.84, p = .431), for testing in future research, in which the level of professional self-efficacy reached at the beginning of working experience appears to be a critical element for understanding the burning-out process.