Evaluasi Efektivitas dan Efisiensi Dashboard Usaha Jamur Tiram dengan Uji Tugas dan Keystroke-Level Model

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Analytics dashboards are increasingly adopted by agricultural MSMEs to record production and finances, yet usability evaluations in this setting mostly rely on questionnaires and rarely compare task performance with a reference point. This study evaluated the effectiveness and efficiency of a deployed web-based dashboard for an oyster mushroom micro-enterprise by combining task-based usability testing following ISO 9241-11 with the Keystroke-Level Model, a predictive performance model that estimates error-free expert time. Seven participants, namely two owners, two employees, and three external growers, completed six tasks on a laptop covering harvest recording, baglog cycle updates, expense recording, and the retrieval of production and profit information. Task success, task completion time, and errors were analysed with the Clopper–Pearson interval, the Friedman test, and the one-sample Wilcoxon signed-rank test against the predicted times. All 42 attempts succeeded (success rate 100%; 95% CI 91.6–100%), with a mean of 13.0 seconds per task and 11 errors, mostly in data-entry tasks. Completion time differed significantly across tasks (χ²(5) = 17.20, p = 0.004). Observed times did not differ from the predictions for harvest and expense recording (p ≥ 0.875) but were significantly longer for identifying the rack due for baglog replacement (ratio 3.97, p = 0.016) and updating the baglog date (ratio 1.48, p = 0.031). The dashboard is effective, while its efficiency is limited by visual search in the rack status display and by report periods that are not directly available. Combining task data with a predictive model identifies redesign priorities without additional participants.