Аңдатпа
Learning analytical chemistry requires not only mastery of analytical and calculation methods but also the ability to make consistent experimental decisions. This study tested a process-oriented assessment framework based on the experimental decision-making pathway (EDMP) construct and the analysis of student-specific decision-making trajectories. A quasi-experimental study involved 60 second-year students, 30 each in a project-based learning group and a control group with a traditional laboratory format. During an eight-week module on determining total iron in water, 2317 individual decisions were recorded; independence, justification, use of data, and revision patterns were coded for each event. Dynamics were analyzed using generalized linear mixed models and sequence analysis, and trajectory characteristics were compared with educational and analytical outcomes. In the experimental group, the proportion of independent decisions increased from 35.7 to 61.5%, and the proportion of data-driven decisions increased from 32.0 to 58.2%. In the control group, changes were less pronounced. The between-group difference was greater for analytical problem solving (d = 1.26) than for subject knowledge (d = 0.84). The proportion of data-driven solutions correlated positively with the final problem-solving score (ρ = 0.52) and negatively with the relative error in iron determination (ρ = −0.44). Overall, the results supported the hypotheses and demonstrated that process analysis complements the final indicators, allowing us to trace the transition from dependent procedure execution to more independent and adaptive experimental reasoning.