1: \begin{abstract}

2: We investigate and contrast, via the Wang-Landau (WL) algorithm,

3: the effects of quenched bond randomness on the self-averaging

4: properties of two Ising spin models in 2d. The random bond version

5: of the superantiferromagnetic (SAF) square model with nearest- and

6: next-nearest-neighbor competing interactions and the corresponding

7: version of the simple ferromagnetic Ising model are studied. We

8: find that, the random bond SAF model shows a strong violation of

9: self-averaging, much stronger than that observed in the case of

10: the random bond Ising model. Our analysis of the asymptotic

11: scaling behavior of the variance of the distribution of the

12: sample-dependent pseudocritical temperatures is found to be

13: consistent with the renormalization group prediction of Aharony

14: and Harris. Using this alternative approach, we find estimates of

15: the correlation length exponent $\nu$ in agreement with results

16: obtained from the usual finite-size scaling (FSS) methodology.

17: \end{abstract}

18: