The intervention with EW on participants followed the design of Pennebaker and Beall (1986) with adjusted writing tasks based on Tolcher et al. (2024). Lasting for 40 days, it included four rounds of writing with a detached vantage point-one month, three months, and six months after the intervention Each of the four rounds consisted of three consecutive days of writing followed by a gap week. On each day, they were asked to write continuously for 20-25 minutes about her life experiences that had a significant impact on her, either positively or negatively. After writing on the third day of each round, they responded to the BDI-II and answered interview questions about their emotional state and self-cognition.

Each of the four rounds consisted of three consecutive days of writing followed by a gap week.

Additionally, to assess changes at cognitive levels, we separately recorded and compared the items related to self-cognition (7 items) within the BDI-II. The specific determination rules were as follows:the direction of score change at the item level served as the basis for judgement. Within a given assessment interval, we compared the number of items that decreased (indicating improvement) with those that increased (indicating deterioration). If decreases outnumbered increases, we classified the interval as improvement-predominant; otherwise, deterioration predominated. If the two were comparable, the change was recorded as insignificant. Accordingly, we sequentially compared items across the self cognition dimensions in the nine assessments, thereby revealing the immediate and sustained manifestations of expressive writing at both cognitive levels. Besides, the cognitive words and self cognition related sentences in interview were also analysis as the evidences for self cognition.

we separately recorded and compared the items related to emotional state (5 items) The specific determination rules were as follows: the direction of score change at the item level served as the basis for judgement. Within a given assessment interval, we compared the number of items that decreased (indicating improvement) with those that increased (indicating deterioration). If decreases outnumbered increases, we classified the interval as improvement-predominant; otherwise, deterioration predominated. If the two were comparable, the change was recorded as insignificant. Accordingly, we sequentially compared items across the negtaive emotion and positive emotion dimensions in the nine assessments, thereby revealing the immediate and sustained manifestations of expressive writing at both emotional levels. Besides, the emotional words used in the writing texts and emotional related senteneces in semi-strcutured interview also analysis as the evidences for positive and negtaive emotion.