Fig. 1: A Question Development Cycle describes the general process of developing assessment items and associated predictive machine learning models. | Humanities and Social Sciences Communications

Fig. 1: A Question Development Cycle describes the general process of developing assessment items and associated predictive machine learning models.

From: FEW questions, many answers: using machine learning to assess how students connect food–energy–water (FEW) concepts

Fig. 1

Adapted from Urban—Lurain et al. (2015), each box represents a stage of the process beginning with Question Design with outputs from one stage being used in subsequent stages, as indicated by solid arrows. A predictive model in the top, right corner is the ultimate goal of the cycle, in which a machine learning model can accurately predict classifications of new responses. A dashed arrow represents possible iteration(s) of the cycle depending on the outcomes of previous stages.

Back to article page