Earlier research on gender effects with software features intended to help problem-solvers in end-user debugging environments has shown that females are less likely to use unfamiliar software features. This poses a serious problem because these features may be key to helping them with debugging problems. Contrasting this with research documenting males' inclination for tinkering in unfamiliar environments, the question arises as to whether encouraging tinkering with new features would help females overcome the factors, such as low self-efficacy, that led to the earlier results. In this paper, we present an experiment with males and females in an end-user debugging setting, and investigate how tinkering behavior impacts several measures of their debugging success. Our results show that the factors of tinkering, reflection, and self-efficacy, can combine in multiple ways to impact debugging effectiveness differently for males than for females.