國立虎尾科技大學 |

Laboratory experiments in information retrieval = sample sizes, effect sizes, and statistical power /

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Laboratory experiments in information retrieval/ by Tetsuya Sakai.
Reminder of title:	sample sizes, effect sizes, and statistical power /
Author:	Sakai, Tetsuya.
Published:	Singapore :Springer Singapore : : 2018.,
Description:	ix, 150 p. :ill., digital ; : 24 cm.;
Contained By:	Springer eBooks
Subject:	Information retrieval - Statistical methods. -
Online resource:	https://doi.org/10.1007/978-981-13-1199-4
ISBN:	9789811311994

Laboratory experiments in information retrieval = sample sizes, effect sizes, and statistical power /
Sakai, Tetsuya.

Laboratory experiments in information retrievalsample sizes, effect sizes, and statistical power /[electronic resource] :by Tetsuya Sakai. - Singapore :Springer Singapore :2018. - ix, 150 p. :ill., digital ;24 cm. - The information retrieval series,v.401387-5264 ;. - Information retrieval series ;v.37..

1 Preliminaries -- 2 t-tests -- 3 Analysis of Variance -- 4 Multiple Comparison Procedures -- 5 The Correct Ways to Use Significance Tests -- 6 Topic Set Size Design Using Excel -- 7 Power Analysis Using R -- 8 Conclusions.

Covering aspects from principles and limitations of statistical significance tests to topic set size design and power analysis, this book guides readers to statistically well-designed experiments. Although classical statistical significance tests are to some extent useful in information retrieval (IR) evaluation, they can harm research unless they are used appropriately with the right sample sizes and statistical power and unless the test results are reported properly. The first half of the book is mainly targeted at undergraduate students, and the second half is suitable for graduate students and researchers who regularly conduct laboratory experiments in IR, natural language processing, recommendations, and related fields. Chapters 1-5 review parametric significance tests for comparing system means, namely, t-tests and ANOVAs, and show how easily they can be conducted using Microsoft Excel or R. These chapters also discuss a few multiple comparison procedures for researchers who are interested in comparing every system pair, including a randomised version of Tukey's Honestly Significant Difference test. The chapters then deal with known limitations of classical significance testing and provide practical guidelines for reporting research results regarding comparison of means. Chapters 6 and 7 discuss statistical power. Chapter 6 introduces topic set size design to enable test collection builders to determine an appropriate number of topics to create. Readers can easily use the author's Excel tools for topic set size design based on the paired and two-sample t-tests, one-way ANOVA, and confidence intervals. Chapter 7 describes power-analysis-based methods for determining an appropriate sample size for a new experiment based on a similar experiment done in the past, detailing how to utilize the author's R tools for power analysis and how to interpret the results. Case studies from IR for both Excel-based topic set size design and R-based power analysis are also provided.

ISBN: 9789811311994

Standard No.: 10.1007/978-981-13-1199-4doiSubjects--Topical Terms:

1209182
Information retrieval
--Statistical methods.

LC Class. No.: ZA3075

Dewey Class. No.: 025.04

Laboratory experiments in information retrieval = sample sizes, effect sizes, and statistical power /
LDR:03320nam a2200349 a 4500 001 928931
003 DE-He213
005 20180923021022.0
006 m d
007 cr nn 008maaau
008 190626s2018 si s 0 eng d
020 $a 9789811311994 $q (electronic bk.)
020 $a 9789811311987 $q (paper)
024 7 $a 10.1007/978-981-13-1199-4 $2 doi
035 $a 978-981-13-1199-4
040 $a GP $c GP
041 0 $a eng
050 4 $a ZA3075
072 7 $a UNH $2 bicssc
072 7 $a COM030000 $2 bisacsh
072 7 $a UNH $2 thema
072 7 $a UND $2 thema
082 0 4 $a 025.04 $2 23
090 $a ZA3075 $b .S158 2018
100 1 $a Sakai, Tetsuya. $3 1113521
245 1 0 $a Laboratory experiments in information retrieval $h [electronic resource] : $b sample sizes, effect sizes, and statistical power / $c by Tetsuya Sakai.
260 $a Singapore : $c 2018. $b Springer Singapore : $b Imprint: Springer,
300 $a ix, 150 p. : $b ill., digital ; $c 24 cm.
490 1 $a The information retrieval series, $x 1387-5264 ; $v v.40
505 0 $a 1 Preliminaries -- 2 t-tests -- 3 Analysis of Variance -- 4 Multiple Comparison Procedures -- 5 The Correct Ways to Use Significance Tests -- 6 Topic Set Size Design Using Excel -- 7 Power Analysis Using R -- 8 Conclusions.
520 $a Covering aspects from principles and limitations of statistical significance tests to topic set size design and power analysis, this book guides readers to statistically well-designed experiments. Although classical statistical significance tests are to some extent useful in information retrieval (IR) evaluation, they can harm research unless they are used appropriately with the right sample sizes and statistical power and unless the test results are reported properly. The first half of the book is mainly targeted at undergraduate students, and the second half is suitable for graduate students and researchers who regularly conduct laboratory experiments in IR, natural language processing, recommendations, and related fields. Chapters 1-5 review parametric significance tests for comparing system means, namely, t-tests and ANOVAs, and show how easily they can be conducted using Microsoft Excel or R. These chapters also discuss a few multiple comparison procedures for researchers who are interested in comparing every system pair, including a randomised version of Tukey's Honestly Significant Difference test. The chapters then deal with known limitations of classical significance testing and provide practical guidelines for reporting research results regarding comparison of means. Chapters 6 and 7 discuss statistical power. Chapter 6 introduces topic set size design to enable test collection builders to determine an appropriate number of topics to create. Readers can easily use the author's Excel tools for topic set size design based on the paired and two-sample t-tests, one-way ANOVA, and confidence intervals. Chapter 7 describes power-analysis-based methods for determining an appropriate sample size for a new experiment based on a similar experiment done in the past, detailing how to utilize the author's R tools for power analysis and how to interpret the results. Case studies from IR for both Excel-based topic set size design and R-based power analysis are also provided.
650 0 $a Information retrieval $x Statistical methods. $3 1209182
650 0 $a Information storage and retrieval systems $x Statistical methods. $3 1209183
650 1 4 $a Information Storage and Retrieval. $3 593926
650 2 4 $a Statistics for Engineering, Physics, Computer Science, Chemistry and Earth Sciences. $3 782247
710 2 $a SpringerLink (Online service) $3 593884
773 0 $t Springer eBooks
830 0 $a Information retrieval series ; $v v.37. $3 1139662
856 4 0 $u https://doi.org/10.1007/978-981-13-1199-4
950 $a Computer Science (Springer-11645)