“…In past studies on SF failure, the use of the correlation heuristic has always been the main part of errors (Cronin et al ., 2008; Brunstein et al ., ; Gonzalez and Wong, ; Baghaei Lakeh and Ghaffarzadegan, ; Fischer and Gonzalez, ; Qi and Gonzalez, ; Gonzalez et al ., ), and the proportion of it in all errors has been hardly affected by factors. The logistic regression models, which are used to estimate the probability of participants' following a correlation heuristic when they cannot answer Q3 and Q4 correctly using mathematical score as one of the predictors, have been reported only by Qi and Gonzalez (), yet no such effect has been found.…”