Applying Genetic Algorithm to Generation of High-Dimensional Item Response Data

Byoungwook Kim, Ja Mee Kim, Won Gyu Lee

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

The item response data is the nm-dimensional data based on the responses made by m examinees to the questionnaire consisting of n items. It is used to estimate the ability of examinees and item parameters in educational evaluation. For estimates to be valid, the simulation input data must reflect reality. This paper presents the effective combination of the genetic algorithm (GA) and Monte Carlo methods for the generation of item response data as simulation input data similar to real data. To this end, we generated four types of item response data using Monte Carlo and the GA and evaluated how similarly the generated item response data represents the real item response data with the item parameters (item difficulty and discrimination). We adopt two types of measurement, which are root mean square error and Kullback-Leibler divergence, for comparison of item parameters between real data and four types of generated data. The results show that applying the GA to initial population generated by Monte Carlo is the most effective in generating item response data that is most similar to real item response data. This study is meaningful in that we found that the GA contributes to the generation of more realistic simulation input data.

Original languageEnglish
Article number589317
JournalMathematical Problems in Engineering
Volume2015
DOIs
Publication statusPublished - 2015

Fingerprint

High-dimensional
Genetic algorithms
Genetic Algorithm
Mean square error
Monte Carlo methods
Simulation
Kullback-Leibler Divergence
Questionnaire
Estimate
Monte Carlo method
Discrimination
Roots
Valid

ASJC Scopus subject areas

  • Mathematics(all)
  • Engineering(all)

Cite this

Applying Genetic Algorithm to Generation of High-Dimensional Item Response Data. / Kim, Byoungwook; Kim, Ja Mee; Lee, Won Gyu.

In: Mathematical Problems in Engineering, Vol. 2015, 589317, 2015.

Research output: Contribution to journalArticle

@article{ab5601208e6d46039bbbe863b1c086e1,
title = "Applying Genetic Algorithm to Generation of High-Dimensional Item Response Data",
abstract = "The item response data is the nm-dimensional data based on the responses made by m examinees to the questionnaire consisting of n items. It is used to estimate the ability of examinees and item parameters in educational evaluation. For estimates to be valid, the simulation input data must reflect reality. This paper presents the effective combination of the genetic algorithm (GA) and Monte Carlo methods for the generation of item response data as simulation input data similar to real data. To this end, we generated four types of item response data using Monte Carlo and the GA and evaluated how similarly the generated item response data represents the real item response data with the item parameters (item difficulty and discrimination). We adopt two types of measurement, which are root mean square error and Kullback-Leibler divergence, for comparison of item parameters between real data and four types of generated data. The results show that applying the GA to initial population generated by Monte Carlo is the most effective in generating item response data that is most similar to real item response data. This study is meaningful in that we found that the GA contributes to the generation of more realistic simulation input data.",
author = "Byoungwook Kim and Kim, {Ja Mee} and Lee, {Won Gyu}",
year = "2015",
doi = "10.1155/2015/589317",
language = "English",
volume = "2015",
journal = "Mathematical Problems in Engineering",
issn = "1024-123X",
publisher = "Hindawi Publishing Corporation",

}

TY - JOUR

T1 - Applying Genetic Algorithm to Generation of High-Dimensional Item Response Data

AU - Kim, Byoungwook

AU - Kim, Ja Mee

AU - Lee, Won Gyu

PY - 2015

Y1 - 2015

N2 - The item response data is the nm-dimensional data based on the responses made by m examinees to the questionnaire consisting of n items. It is used to estimate the ability of examinees and item parameters in educational evaluation. For estimates to be valid, the simulation input data must reflect reality. This paper presents the effective combination of the genetic algorithm (GA) and Monte Carlo methods for the generation of item response data as simulation input data similar to real data. To this end, we generated four types of item response data using Monte Carlo and the GA and evaluated how similarly the generated item response data represents the real item response data with the item parameters (item difficulty and discrimination). We adopt two types of measurement, which are root mean square error and Kullback-Leibler divergence, for comparison of item parameters between real data and four types of generated data. The results show that applying the GA to initial population generated by Monte Carlo is the most effective in generating item response data that is most similar to real item response data. This study is meaningful in that we found that the GA contributes to the generation of more realistic simulation input data.

AB - The item response data is the nm-dimensional data based on the responses made by m examinees to the questionnaire consisting of n items. It is used to estimate the ability of examinees and item parameters in educational evaluation. For estimates to be valid, the simulation input data must reflect reality. This paper presents the effective combination of the genetic algorithm (GA) and Monte Carlo methods for the generation of item response data as simulation input data similar to real data. To this end, we generated four types of item response data using Monte Carlo and the GA and evaluated how similarly the generated item response data represents the real item response data with the item parameters (item difficulty and discrimination). We adopt two types of measurement, which are root mean square error and Kullback-Leibler divergence, for comparison of item parameters between real data and four types of generated data. The results show that applying the GA to initial population generated by Monte Carlo is the most effective in generating item response data that is most similar to real item response data. This study is meaningful in that we found that the GA contributes to the generation of more realistic simulation input data.

UR - http://www.scopus.com/inward/record.url?scp=84946962669&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84946962669&partnerID=8YFLogxK

U2 - 10.1155/2015/589317

DO - 10.1155/2015/589317

M3 - Article

VL - 2015

JO - Mathematical Problems in Engineering

JF - Mathematical Problems in Engineering

SN - 1024-123X

M1 - 589317

ER -