PhD defence Deniece Nazareth | Emotional Experiences and Expressions in Reminiscence | A Multimodal Analysis of Emotional Memories in Older Adults

Emotional Experiences and Expressions in Reminiscence | A Multimodal Analysis of Emotional Memories in Older Adults

Deniece Nazareth is a PhD student in the department Psychology, Health & Technology. (Co)Promotors are prof.dr. G.J. Westerhof & prof.dr. D.K.J. Heylen and dr. K.P. Truong from the faculty of Behavioural Management and Social Sciences (BMS) University of Twente.

As people live longer, it is important to focus on emotional well-being as key component of healthy ageing. This dissertation aims to contribute to a better understanding of the emotional well-being of older adults by combining the research fields of reminiscence, emotion, and affective computing to gain new insights and resources.

Reminiscence is effective to enhance the emotional well-being in older adults through autobiographical memory recall. Emotion research shows that older adults experience and express emotions differently due to more emotional complexity and different emotion regulation strategies. Affective computing focuses on recognizing emotional expressions in various modalities using automatic recognition technology and affective databases.

Based on these fields, challenges and gaps were identified and addressed. We present the Multimodal Emotional Memories of Older Adults (MEMOA) database, consisting of positive and negative memories through reminiscence by older adults. Following, the Valence of Emotional Memories (VEM) coding scheme is introduced, based on life events in autobiographical memories. Lexical and acoustic modalities were related to VEM valence scores, emphasizing the importance of multimodal analysis and the use of automatic recognition technology. Next, the role of reappraisal of emotional memories and self-reported valence is examined. A shift in the past and present valence was found as memory types were reappraised differently, due to a positivity and fading affect bias. Multimodalities were also slightly more related to past valence due to the reappraisal. Lastly, the relation between self-reported past and present valence, reminiscence functions and multimodalities is investigated. Different memory types are reminisced with different purposes. In positive memories, some modalities moderated the difference between past and present valence.

To contribute to a better understanding of emotional well-being in older adults, we have shown that reminiscence is a fitting elicitation method for older adults to evoke and investigate emotional experiences and expressions.  Further, the VEM coding scheme complements traditional self-reports to study valence in memories. We also highlight the importance of distinguishing between past and present valence in memories when studying emotions in older adults. Lastly, recommendations for future research and practical implications are provided.