Generalized non-constant α model framing learning as an error-reduction process¶

Theory · All theories
Evidence · 1 claim (1 for) · 1 study (1 theoretical), q2 · 1 of 1 report an effect size · 1 claim rests on one study

Description¶

In Section VI the article generalizes the γ-process as a special case of the α-process with α linearly dependent on score, allowing α to be a function of the score including constant, linear, and higher-order terms (Eqs. 24–25). The article states the meaning of Eq. (24) reflects 'a measurement based probability model that would lead to a general form “error reduction” process', in which measured incorrect knowledge is reduced when α is positive. Any non-constant terms in α produce a positive gain–pretest correlation and a curved-down pattern at low pretest scores.

Design Implications¶

Context¶

Requirements¶

• Score-based pre-post measurement in a fixed measurement window

Constraints¶

• The article reports details of this generalization and other extensions (including connections to the Rescorla-Wagner model) will be reported in future papers

Target Learners¶

• physics students in pre-post assessed courses

Target Learning Objectives¶

• reduction of measured incorrect knowledge through instruction

Claims¶

• Gamma Process Positive Gain Pretest Correlation [+M]

Related Theories¶

• Two-level measurement-based dynamic model of learning with α, β, and γ transition processes

Examples¶

-

Key Sources¶

• Lei Bao. (2007). Dynamic Models of Learning and Education Measurement. https://arxiv.org/abs/0710.1375