Valence émotionnelle des mots dans la schizophrénie

Abstract

Emotion recognition is a domain in which deficits have been reported in schizophrenia. A number of emotion classification studies have indicated that emotion processing deficits in schizophrenia are more pronounced for negative affects. Given the difficulty of developing material suitable for the study of these emotional deficits, it would be interesting to examine whether patients suffering from schizophrenia are responsive to positively and negatively charged emotion-related words that could be used within the context of remediation strategies. The emotional perception of words was examined in a clinical experiment involving schizophrenia patients. This emotional perception was expressed by the patients in terms of the valence associated with the words. In the present study, we investigated whether schizophrenia patients would assign the same negative and positive valences to words as healthy individuals. Twenty volunteer, clinically stable, outpatients from the Psychiatric Service of the University Hospital of Clermont-Ferrand were recruited. Diagnoses were based on DSM-IV criteria. Global psychiatric symptoms were assessed using the Positive and Negative Symptoms Scale (PANSS). The patients had to evaluate the emotional valence of a set of 300 words on a 5-point scale ranging from "very unpleasant" to "very pleasant". . The collected results were compared with those obtained by Bonin et al. (2003) [13] from 97 University students. Correlational analyses of the two studies revealed that the emotional valences were highly correlated, i.e. the schizophrenia patients estimated very similar emotional valences. More precisely, it was possible to examine three separate sets of 100 words each (positive words, neutral words and negative words). The positive words that were evaluated were the more positive words from the norms collected by Bonin et al. (2003) [13], and the negative words were the more negative examples taken from these norms. The neutral words reflected the more neutral emotional valences collected. The results suggested that the emotional valences for the negative words were significantly greater in the patients than is observed in the collected norms (P<0.001). Moreover, the emotional valence associated with positive words was significantly lower in the patients than in the collected norms (P<0.001). Nevertheless, only 16 words differed significantly in terms of evaluated emotional valence between patients and young adults. Despite their overall emotional impairments, the patients with schizophrenia had a very similar perception of word valence as controls. This result suggests that the emotional perception of most emotional words is preserved. This research provides data important to consider during rehabilitation. Moreover, this study will make it possible to select stimuli for use in future studies of emotion in patients.