Flora et al. (1992, 2003) found that aversive stimulation in the form of intense noise or cold water increased the number of impulsive responses when presented concurrently to a choice task. Hurtado-Parrado et al. (2023) systematically replicated Flora et al.’s (1992) study and found that adding a visual matching-to-sample task with images from the International Affective Picture System (IAPS; Lang et al., 2008) produced a higher number of responses on the impulsive option (Sooner-Smaller - SS) near indifference, as compared to a No-images group who showed preference for the self-controlled option (Larger-Later - LL). They also found that the overall rate of correct matching across all conditions was low. Hurtado-Parrado et al. proposed that the potential aversive or distractive effect of frequent failure during the matching task, and thus frequent point loss, could have had a similar effect of the cold pressor and aversive noise on impulsive responses (SS). The present study systematically replicated Hurtado-Parrado et al.’s (2023) experiment to evaluate the effects on SS responses of different rates of correct responses during the matching-to-sample component of the experimental task across four conditions: No-images, 0%, 60%, and 100% correct matching. Fourteen college students were exposed to all conditions in a within-subject design. The prediction that the number of SS responses would increase as the percentage of correct responses in the matching task decreased was not confirmed. Participants developed one of three choice patterns during the first no matching condition (No Images), namely preference for the self-control or impulsive option, or indifference. Subsequent changes in percentages of correct matching did not affect indifference or self-control preference, but contrary to the predicted effect, decreased impulsive responses towards indifference or self-control preference. These findings are discussed in terms of possible effects of (a) motivating operations (e.g., monetary compensation in the present study differed from previous studies by Flora et al., 1992, 2003 and Hurtado-Parrado et al., 2023); (b) extra-experimental contingencies (e.g., several students participated during the last days of the semester trying to complete course credits in small amount of time); (c) rule-construction and rule-following behavior (e.g., rule-construction during the No images condition could explain insensitivity to subsequent changes in percentage of correct matching), and (d) local and molar reinforcement rate variations across conditions (e.g., an increase in rates of point gain, and thus increase in their reinforcement value, could explain the reduction in initial impulsive preference of some participants when correct matching was later manipulated).