Do errors matter? Errorless and errorful learning in anomic picture naming

Abstract

Errorless training methods significantly improve learning in memory-impaired patients relative to errorful training procedures. However, the validity of this technique for acquiring linguistic information in aphasia has rarely been studied. This study contrasts three different treatment conditions over an 8 week period for rehabilitating picture naming in anomia: (1) errorless learning in which pictures are shown and the experimenter provides the name, (2) errorful learning with feedback in which the patient is required to generate a name but the correct name is then supplied by the experimenter, and (3) errorful learning in which no feedback is given. These conditions are compared to an untreated set of matched words. Both errorless and errorful learning with feedback conditions led to significant improvement at a 2-week and 12–14-week retest (errorful without feedback and untreated words were similar). The results suggest that it does not matter whether anomic patients are allowed to make errors in picture naming or not (unlike in memory impaired individuals). What does matter is that a correct response is given as feedback. The results also question the widely held assumption that it is beneficial for a patient to attempt to retrieve a word, given that our errorless condition involved no retrieval effort and had the greatest benefits.

Notes

¹ It is to be noted that the initial impetus for these studies was not to contrast errorful with errorless learning, but to address other questions relating to therapy. The studies did not use these terms or interpret their data in this way at any point in the studies.

² It is to be noted that Fridriksson et al. (2005) describe spaced retrieval as “essentially an errorless learning procedure” (p. 99). However, it does not fit well with our definition of errorless learning because each retrieval attempt could result in error, even if it is subsequently corrected by the experimenter.