Computerizing reading training

Evaluation of a latent semantic analysis space for science text

Christopher A. Kurby, Katja Wiembr-Hastings, Nagasai Ganduri, Joseph P. Magliano, Keith K. Millis, Danielle McNamara

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

The effectiveness of a domain-specific latent semantic analysis (LSA) in assessing reading strategies was examined. Students were given self-explanation reading training (SERT) and asked to think aloud after each sentence in a science text. Novice and expert human raters and two LSA spaces (general reading, science) rated the similarity of each think-aloud protocol to benchmarks representing three different reading strategies (minimal, local, and global). The science LSA space correlated highly with human judgments, and more highly than did the general reading space. Also, cosines from the science LSA spaces can distinguish between different levels of semantic similarity, but may have trouble in distinguishing local processing protocols. Thus, a domain-specific LSA space is advantageous regardless of the size of the space. The results are discussed in the context of applying the science LSA to a computer-based version of SERT that gives online feedback based on LSA cosines.

Original languageEnglish (US)
Pages (from-to)244-250
Number of pages7
JournalBehavior Research Methods, Instruments, and Computers
Volume35
Issue number2
StatePublished - May 2003
Externally publishedYes

Fingerprint

Semantics
Reading
Benchmarking
Evaluation
Latent Semantic Analysis
Students

ASJC Scopus subject areas

  • Psychology(all)
  • Experimental and Cognitive Psychology
  • Psychology (miscellaneous)

Cite this

Computerizing reading training : Evaluation of a latent semantic analysis space for science text. / Kurby, Christopher A.; Wiembr-Hastings, Katja; Ganduri, Nagasai; Magliano, Joseph P.; Millis, Keith K.; McNamara, Danielle.

In: Behavior Research Methods, Instruments, and Computers, Vol. 35, No. 2, 05.2003, p. 244-250.

Research output: Contribution to journalArticle

Kurby, CA, Wiembr-Hastings, K, Ganduri, N, Magliano, JP, Millis, KK & McNamara, D 2003, 'Computerizing reading training: Evaluation of a latent semantic analysis space for science text', Behavior Research Methods, Instruments, and Computers, vol. 35, no. 2, pp. 244-250.
Kurby, Christopher A. ; Wiembr-Hastings, Katja ; Ganduri, Nagasai ; Magliano, Joseph P. ; Millis, Keith K. ; McNamara, Danielle. / Computerizing reading training : Evaluation of a latent semantic analysis space for science text. In: Behavior Research Methods, Instruments, and Computers. 2003 ; Vol. 35, No. 2. pp. 244-250.
@article{cc9ee34d3d984701800fba5b431d2ac9,
title = "Computerizing reading training: Evaluation of a latent semantic analysis space for science text",
abstract = "The effectiveness of a domain-specific latent semantic analysis (LSA) in assessing reading strategies was examined. Students were given self-explanation reading training (SERT) and asked to think aloud after each sentence in a science text. Novice and expert human raters and two LSA spaces (general reading, science) rated the similarity of each think-aloud protocol to benchmarks representing three different reading strategies (minimal, local, and global). The science LSA space correlated highly with human judgments, and more highly than did the general reading space. Also, cosines from the science LSA spaces can distinguish between different levels of semantic similarity, but may have trouble in distinguishing local processing protocols. Thus, a domain-specific LSA space is advantageous regardless of the size of the space. The results are discussed in the context of applying the science LSA to a computer-based version of SERT that gives online feedback based on LSA cosines.",
author = "Kurby, {Christopher A.} and Katja Wiembr-Hastings and Nagasai Ganduri and Magliano, {Joseph P.} and Millis, {Keith K.} and Danielle McNamara",
year = "2003",
month = "5",
language = "English (US)",
volume = "35",
pages = "244--250",
journal = "Behavior Research Methods, Instruments, and Computers",
issn = "1554-351X",
publisher = "Springer New York",
number = "2",

}

TY - JOUR

T1 - Computerizing reading training

T2 - Evaluation of a latent semantic analysis space for science text

AU - Kurby, Christopher A.

AU - Wiembr-Hastings, Katja

AU - Ganduri, Nagasai

AU - Magliano, Joseph P.

AU - Millis, Keith K.

AU - McNamara, Danielle

PY - 2003/5

Y1 - 2003/5

N2 - The effectiveness of a domain-specific latent semantic analysis (LSA) in assessing reading strategies was examined. Students were given self-explanation reading training (SERT) and asked to think aloud after each sentence in a science text. Novice and expert human raters and two LSA spaces (general reading, science) rated the similarity of each think-aloud protocol to benchmarks representing three different reading strategies (minimal, local, and global). The science LSA space correlated highly with human judgments, and more highly than did the general reading space. Also, cosines from the science LSA spaces can distinguish between different levels of semantic similarity, but may have trouble in distinguishing local processing protocols. Thus, a domain-specific LSA space is advantageous regardless of the size of the space. The results are discussed in the context of applying the science LSA to a computer-based version of SERT that gives online feedback based on LSA cosines.

AB - The effectiveness of a domain-specific latent semantic analysis (LSA) in assessing reading strategies was examined. Students were given self-explanation reading training (SERT) and asked to think aloud after each sentence in a science text. Novice and expert human raters and two LSA spaces (general reading, science) rated the similarity of each think-aloud protocol to benchmarks representing three different reading strategies (minimal, local, and global). The science LSA space correlated highly with human judgments, and more highly than did the general reading space. Also, cosines from the science LSA spaces can distinguish between different levels of semantic similarity, but may have trouble in distinguishing local processing protocols. Thus, a domain-specific LSA space is advantageous regardless of the size of the space. The results are discussed in the context of applying the science LSA to a computer-based version of SERT that gives online feedback based on LSA cosines.

UR - http://www.scopus.com/inward/record.url?scp=0041475865&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0041475865&partnerID=8YFLogxK

M3 - Article

VL - 35

SP - 244

EP - 250

JO - Behavior Research Methods, Instruments, and Computers

JF - Behavior Research Methods, Instruments, and Computers

SN - 1554-351X

IS - 2

ER -