Contextual interference is an established phenomenon in learning research; random practice schedules are associated with poorer performance, but superior learning, compared with blocked practice schedules. We present a secondary analysis of N = 84 healthy young adults, replicating the contextual interference effect in a time estimation task. We used the determinant of a correlation matrix to measure the amount of order in participant responses. We calculated this determinant in different phase spaces: trial space, the determinant of the previous five trials (lagged constant error 0-4); and target space, the determinant of the previous five trials of the same target. In trial space, there was no significant difference between groups (p = 0.98) and no Group × Lag interaction (p = 0.54), although there was an effect of Lag (p < 0.01). In target space, there were effects of Group (p = 0.02), Lag (p < 0.01), and a Group × Lag interaction (p = 0.03). Ultimately, randomly scheduled practice was associated with adaptive corrections but positive correlations between errors from trial to trial (e.g., overshoots followed by smaller overshoots). Blocked practice was associated with more adaptive corrections but uncorrelated responses. Our findings suggest that random practice leads to the retrieval and updating of the target from memory, facilitating long-term retention and transfer.