Recent Changes - Search:

ECD

ACED

Bayesian Networks in Educational Assessment

Cognition and Assessment

PmWiki

pmwiki.org

edit SideBar

BN /

MiniACED

This is a stripped down version of the full ACED example. Go to to the ACED pages for a full description. In particular, this is a simplified version with fewer nodes so that it runs in the student version of Netica?.

All of the nets can be downloaded from http://pluto.coe.fsu.edu/BNinEA/Examples/miniACED (tarball, zip)

Proficiency Model

The proficiency model for ACED only contains the nodes for the geometric sequences part of the full proficiency model. Note that in the motifs below, certain of the proficiency nodes are removed to make the net small enough for the student version.

ACED geometric sequence only proficiency model

Netica network

Evidence models

All of the ACED tasks were dichotomously scored, but the task authors classified them as easy, medium and difficult. The three evidence model fragments below correspond to those three difficulty levels. Note that each of these must be docked? with the proficiency model by attaching the edge to the appropriate node in the proficiency model. After evidence is entered into the observable variable, then it can be absorbed? into the proficiency model.

Easy evidence model Medium evidence model Har evidence model

In this simplified version, only evidence models which tap a single proficiency variable are provided.

Motifs

Two custom motifs are provided for experimenting with expected weight of evidence? and mutual information?.

Motif 1

This motif contains easy, medium and hard tasks for only two proficiency variables. It is good for exploring the effect of task difficulty on information metrics.

Netica network

Motif 2

This motif contains only medium difficulty tasks attached to a number of proficiency variables. It is good for exploring the effect of network topology on information metrics.

Netica network

Edit - History - Print - Recent Changes - Search
Page last modified on November 28, 2014, at 07:34 PM