In the first part of our study, we aimed to i) explore the prevalence of over- and underestimation of language difficulties in a sample of aphasic patients and ii) address the potential statistical bias due to 'extreme' mild and severe cases of aphasia. The second part of our study aimed to further investigate possible underlying attentional (i.e., monitoring) and motivational (i.e., denial) factors that may contribute to the distortion of awareness.

Participants were 78 patients with unilateral left-brain injury and a diagnosis of aphasia (mean age= 60.2, 53.8% men). The study included a language assessment and a measure of explicit awareness of difficulties. A subsample of patients (N=10) participated in the second phase of the study in which they were asked to rate i) their own performance on a naming task while performing the task (online), ii) their recorded performance (offline-own) and iii) a manipulated recorded performance designed to convince them that they were listening to someone else's recording (offline-other). We used Fisher's exact probability tests and nonparametric means comparisons to examine trends in our sample and regression analyses to examine the relationship between aphasia level and awareness. Ratings across conditions of the second part were analysed by means of Crawford- Garthwaite t-tests for single cases.

We found that a significant proportion (34.6%) of patients had altered self- awareness, with a statistically significant higher frequency of awareness distortions in the direction of underestimation (i.e., anosognosia; 21.8%) compared to overestimation of language impairments. A Welch's t-test indicated that there was no significant difference in the extent of under- and overestimation when they did occur (t(71,4)=1.83, p=.07). Regression analyses showed no relationship between aphasia level and awareness (R²=.04, F(1,68)=3.33, p=.07), even when statistical biases due to ‘extreme’ cases of aphasia, were accounted for (R²=.04, F(1,52)=2.17, p=.14). We performed mean comparisons between conditions (i.e., “Online Own vs. Offline Own" for attentional deficits, and "Offline Own vs. Offline Others" for denial) on three anosognosic and seven aware patients. T-tests for single cases revealed a significant difference in performance in one anosognosic patient (t-value = 2.39, p =<.05), indicating the presence of attentional (i.e., monitoring) deficits.

These findings suggest that left-sided unilateral brain damage can lead to distortion of awareness for aphasia, which is mainly seen as a specific phenomenon of underestimation (anosognosia) of one's own deficits and cannot be easily explained as due to statistical biases. Evidence of attentional/monitoring difficulties as a cause of anosognosia was found in only one of three patients. Therefore, our results lack a consistent pattern that definitively links denial or attentional mechanisms to distorted awareness. This suggests that the nature of anosognosia is multifaceted and not easily defined by singular explanatory factors.