ENA mapped an empathy gap between real geriatric caregiving and training narratives

Kiafar and colleagues asked whether the experience of caring for older adults is represented adequately in nursing assistants' training. Their ICQE 2024 conference paper uses Epistemic Network Analysis to compare how ten nursing assistants described real caregiving with how they described previous training. The purpose was exploratory and practical: identify patterns that could inform more realistic geriatric-care simulations.

The participants were nine women and one man, aged 19 to 28, with between five months and five years of caregiving experience. Each completed an approximately one-hour online interview covering daily work, interactions with older adults, challenging incidents, prior training, and recommendations for simulation design. The cleaned transcripts contained more than 810 utterances. Two researchers coded each utterance for the possible presence of communication, empathy, flexibility, and critical thinking. More than one code could be assigned to an utterance. The reported coding agreement was 93.7 percent, with Cohen's kappa of .863.

The researchers used ENA Web Tool version 1.7.0. Each interview question formed a stanza, and individual caregivers were the units used to calculate ENA scores. The analysis normalized networks for different numbers of coded lines, used singular value decomposition, and compared Real and Training response categories. The first ENA dimension explained 30.6 percent of variance and separated the two categories in the reported test. Real-life narratives were characterized by a strong connection between empathy and communication. Training narratives showed their strongest connection between communication and critical thinking, while empathy was small and weakly connected.

The pattern suggests a concrete design question. If empathy and communication are closely coupled in everyday geriatric care but are weakly coupled in remembered training, simulation designers may need scenarios in which learners must understand an older adult's perspective while making and communicating decisions. ENA made that gap visible as a relationship among codes rather than as isolated counts.

The evidence must remain bounded. This was an exploratory sample of ten caregivers, based on recalled accounts rather than observed behavior. Older adults were not interviewed, and the participants represented a narrow age and experience range. The same caregivers contributed Real and Training responses, yet the reported group comparison used an unequal-variance two-sample test; the paper does not establish that the patterns generalize to a wider workforce. The defensible contribution is therefore a preliminary network-based account and an actionable hypothesis for training design, not proof that existing geriatric education systematically lacks empathy.