These exciting results highlight the potential of LLM-powered linguistic analysis to detect people who are struggling with anxiety and depression. They also deepen our understanding of the interpersonal processes that support effective therapy.

LLMs can also expand beyond only analyzing client language and estimate how strongly therapists encouraged distance in their clients. Ratings also tracked changing symptoms when word-counting approaches could not!

These exciting results highlight the potential of LLM-powered linguistic analysis to detect people who are struggling with anxiety and depression. They also deepen our understanding of the interpersonal processes that support effective therapy.

LLMs can also expand beyond only analyzing client language and also estimate how strongly therapists encouraged distance in their clients. Ratings also tracked changing symptoms when word-counting approaches could not!

We discuss implications for affective and developmental science in the paper, but one that we're curious about is whether our results explain why folks with low emotion differentiation tend to think of their emotions as merely "happy" or "sad" - these words come to mind first!

The 1st emotion words that people tended to produce were those they learned earliest in life. Interestingly, the dimensions that shape emotion word accessibility were stable across age: Both children and adults tended to start their responses with words they learned early in life

We also tested whether affective and linguistic dimensions predicted the order in which participants thought of emotion words. Above and beyond the classic circumplex dimensions (i.e., valence and arousal), the strongest predictor of word order was actually age of acquisition.

We found that verbal fluency and emotion fluency followed similar age-related trajectories, increasing across childhood and adolescence before plateauing around 18-19 years of age. Thus, emotion fluency seems to be a subtype of verbal fluency, rather than a distinct capacity.

Here, we measured emotion fluency/verbal fluency by asking participants (4-25 year-olds) to produce as many emotion/fruit words as possible in 1min.

These studies have only been conducted in adults, leaving the relation between emotion fluency and verbal fluency uncharted in youth, when emotion concepts are still developing.

A handful of prior studies have examined emotion fluency in adults, seeking to chart the key predictors of emotion fluency and its potential relations with mental health, as facility with emotion words might provide greater efficacy managing one’s emotions