Knowledge and Questionnaire Defined Trust in Herbal Medicine Among Residents of Toyado Village in a Cross Sectional Study

Background: Herbal medicine remains part of community health practice, but factual knowledge and endorsement of statements about safety, use, and social experience are distinct constructs. Examining how the recorded scores co-occur may inform cautious, culturally appropriate communication. Objective: To examine the association between knowledge and questionnaire-defined trust in herbal medicine among surveyed residents of Toyado Village. Methods: This analytic cross-sectional survey was conducted in Toyado Village, Lage District, Poso Regency, on June 13, 2026. Ninety-three participants were recruited by purposive, nonprobability sampling. Knowledge and questionnaire-defined trust were recorded using eight-item and six-item dichotomous scores, respectively. The primary analysis used a two-tailed Spearman rank correlation of total scores; exploratory partial rank correlations adjusted for age, education, and ethnicity. Precision was summarized with 95% bias-corrected and accelerated bootstrap confidence intervals. Results: The median knowledge score was 7 (Q1–Q3: 6–8) and the median questionnaire-defined trust score was 4 (Q1–Q3: 2–5). Knowledge was categorized as good in 54 participants (58.1%), moderately good in 33 (35.5%), and poor in six (6.5%); corresponding trust categories were 45 (48.4%), 10 (10.8%), and 38 (40.9%). The total scores showed a positive association (Spearman’s ρ = 0.456; 95% CI: 0.267–0.614; p < 0.001). The exploratory adjusted partial rank correlation was 0.360 (95% CI: 0.160–0.571; p < 0.001). Forty participants (43.0%) reached the maximum knowledge score, indicating a potential ceiling effect. Conclusion: Knowledge and questionnaire-defined trust scores were positively associated within this nonprobability sample, but the cross-sectional design does not establish temporal order or causality. Purposive recruitment, a ceiling effect for knowledge, limited documented reliability, and the heterogeneous trust items constrain population inference and construct interpretation