Experience Seamless Full Ebook Downloads for Every Genre at textbookfull.

com

The Basics of Item Response Theory Using R 1st

Edition Frank B. Baker

https://textbookfull.com/product/the-basics-of-item-
response-theory-using-r-1st-edition-frank-b-baker/

OR CLICK BUTTON

DOWNLOAD NOW

Explore and download more ebook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Handbook of Item Response Theory, Three Volume Set - First

Edition Van Der Linden

https://textbookfull.com/product/handbook-of-item-response-theory-
three-volume-set-first-edition-van-der-linden/

textboxfull.com

Handbook of item response theory, volume three:

applications 1st Edition Wim J. Van Der Linden (Editor)

https://textbookfull.com/product/handbook-of-item-response-theory-
volume-three-applications-1st-edition-wim-j-van-der-linden-editor/

textboxfull.com

Land Subsidence Mitigation: Aquifer Recharge Using Treated

Wastewater Injection 1st Edition Frank R. Spellman

https://textbookfull.com/product/land-subsidence-mitigation-aquifer-
recharge-using-treated-wastewater-injection-1st-edition-frank-r-
spellman/
textboxfull.com

The Legal Environment of Business 10th Edition Frank B.

Cross

https://textbookfull.com/product/the-legal-environment-of-
business-10th-edition-frank-b-cross/

textboxfull.com
Afropessimism 1st Edition Frank B. Wilderson Iii

https://textbookfull.com/product/afropessimism-1st-edition-frank-b-
wilderson-iii/

textboxfull.com

The Handbook of Nature 2nd Edition Frank R. Spellman

https://textbookfull.com/product/the-handbook-of-nature-2nd-edition-
frank-r-spellman/

textboxfull.com

The Science of Water 4th Edition Frank R. Spellman

https://textbookfull.com/product/the-science-of-water-4th-edition-
frank-r-spellman/

textboxfull.com

The Science of Environmental Pollution, Third Edition

Frank R. Spellman

https://textbookfull.com/product/the-science-of-environmental-
pollution-third-edition-frank-r-spellman/

textboxfull.com

Using Corpora to Analyze Gender Paul Baker

https://textbookfull.com/product/using-corpora-to-analyze-gender-paul-
baker/

textboxfull.com
Statistics for Social and Behavioral Sciences

Frank B. Baker
Seock-Ho Kim

The Basics of
Item Response
Theory Using R
Statistics for Social and Behavioral Sciences

Series editor
Stephen E. Fienberg
Carnegie Mellon University
Pittsburgh
Pennsylvania
USA

Statistics for Social and Behavioral Sciences (SSBS) includes monographs and
advanced textbooks relating to education, psychology, sociology, political science,
public policy, and law.

More information about this series at http://www.springer.com/series/3463

Frank B. Baker • Seock-Ho Kim

The Basics of Item Response

Theory Using R

123
Frank B. Baker Seock-Ho Kim
Educational Psychology Educational Psychology
University of Wisconsin-Madison University of Georgia
Madison, WI, USA Athens, GA, USA

ISSN 2199-7357 ISSN 2199-7365 (electronic)

Statistics for Social and Behavioral Sciences
ISBN 978-3-319-54204-1 ISBN 978-3-319-54205-8 (eBook)
DOI 10.1007/978-3-319-54205-8

Library of Congress Control Number: 2017933685

© Springer International Publishing AG 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

This book is an update of the original book, The Basics of Item Response Theory, by
the first author. The original book by Frank B. Baker was based on the course given
during his tenure at the University of Wisconsin–Madison. It appeared in 1985.
The second edition of the original book by Frank B. Baker appeared in 2001 with
a publisher’s note by Lawrence A. Rudner. About 15 years have passed since the
last revision. So much has happened in the fields of educational measurement and
psychometrics as well as in the statistical computing technology. In the meantime,
we felt that the usefulness of the book would be increased by some further changes.
The main alterations are due to the use of the computing package R for the
illustration purpose and especially for the computer sessions. The treatment of the
original topics over eight chapters has not been changed.
The original object of the book was to make the book to be a tutorial for item
response theory suited to those who possess only a limited knowledge of educational
measurement and psychometrics. We have never lost sight of such an object. The
amendments in this book are not due to any alteration in the original object but
they are necessitated by the development of the statistical computing technology. In
particular, the book now aims at covering both the basics of item response theory
and the use of R for preparing graphical presentation in the item response theory
related writings.
We will be indebted to any reader who calls our attention to errors or obscurities.

Madison, WI, USA Frank B. Baker

Athens, GA, USA Seock-Ho Kim
January 2017

v
Acknowledgments of the Original Book (1985)

Over the past century, many people have contributed to the development of item
response theory. Three persons deserve special recognition. D. N. Lawley of the
University of Edinburgh published a paper in 1943 showing that many of the
constructs of classical test theory could be expressed in terms of parameters of the
item characteristic curve. This paper marks the beginning of item response theory
as a measurement theory. Dr. F. M. Lord of the Educational Testing Service has
been the driving force behind both the development of theory and its application
for the past 35 years. Over this period, he has systematically defined, expanded,
and explored the theory as well as developed the computer programs needed to put
the theory into practice. This effort culminated in his recent book on the practical
applications of item response theory. In the late 1960s, Dr. B. D. Wright of the
University of Chicago recognized the importance of the measurement work by the
Danish mathematician Georg Rasch. Since that time, he has played a key role in
bringing item response theory, the Rasch model in particular, to the attention of
practitioners. Without the work of these three individuals, the level of development
of item response theory would not be where it is today.
I am indebted to Mr. T. Seavey of Heinemann Educational Books for first
suggesting that I do a small book on item response theory. This suggestion allowed
me to fulfill a long-standing desire to develop an instructional software package
dealing with item response theory for a microcomputer. I must also acknowledge
the technical assistance of Mr. W. Vilberg in squeezing the maximum capability out
of the Apple II computer. Without his help, the computer software would be much
less sophisticated. Finally, the manuscript was prepared using the Screenwriter II
word processor program written by Mr. R. Kidwell of Sierra On-Line, Inc. Without
this marvelous package, the present book would never be written.

Madison, WI, USA Frank B. Baker

vii
Acknowledgements of the Second Edition of the
Original Book (2001)

Over the past century, many people have contributed to the development of item
response theory. Three persons deserve special recognition. D. N. Lawley of the
University of Edinburgh published a paper in 1943 showing that many of the
constructs of classical test theory could be expressed in terms of parameters of
the item characteristic curve. This paper marks the beginning of item response
theory as a measurement theory. The work of Dr. F. M. Lord of the Educational
Testing Service has been the driving force behind both the development of theory
and its application for the past 50 years. Dr. Lord systematically defined, expanded,
and explored the theory as well as developed the computer programs needed to
put the theory into practice. This effort culminated in his classic books (with
Dr. Melvin Novick, 1968; 1980) on the practical applications of item response
theory. In the late 1960s, Dr. B. D. Wright of the University of Chicago recognized
the importance of the measurement work by the Danish mathematician Georg
Rasch. Since that time, he has played a key role in bringing item response theory,
the Rasch model in particular, to the attention of practitioners. Without the work of
these three individuals, the level of development of item response theory would not
be where it is today.
I am indebted to Mr. T. Seavey of Heinemann Educational Books for first
suggesting that I do a small book on item response theory, which resulted in the
first edition of this book in 1985. This suggestion allowed me to fulfill a long-
standing desire to develop an instructional software package dealing with item
response theory for the then-state-of-the-art Apple II and IBM PC computers. An
upgraded version of this software has now been made available on the World Wide
Web (http://wricae.net/irt).

Madison, WI, USA Frank B. Baker

ix
Contents

1 The Item Characteristic Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2 Item Characteristic Curve Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3 Estimating Item Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4 The Test Characteristic Curve. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5 Estimating an Examinee’s Ability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6 The Information Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
7 Test Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
8 Specifying the Characteristics of a Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

A R Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

B Estimating Item Parameters Under the Two-Parameter

Model with Logistic Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

C Putting the Three Tests on a Common Ability Scale: Test Equating . . . 167

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

xi
Introduction

When the original book was first published in 1985, the fields of educational
measurement and psychometrics were in a transitional period. The majority of
practice was based upon the classical test theory developed during the 1920s.
However, a new test theory had been developing over the past 40 years that was
conceptually more powerful than classical test theory. Based upon items rather than
test scores, the new approach was known as item response theory. While the basic
concepts of item response theory are straightforward, the underlying mathematics
is somewhat advanced compared to that of classical test theory. As a result, it is
difficult to examine some of these concepts without performing a large number of
calculations to obtain usable information. The original book was designed to provide
the reader access to the basic concepts of item response theory freed of the tedious
underlying calculations through an Apple II computer program. The second edition
of the original published in 2001 used a version of computer program written in
Visual Basic 5.0 that could be obtained at http://ericae.net/irt. Readers accustomed
to sophisticated statistical and graphics packages might have found it utilitarian,
but nevertheless helpful in understanding various facets of the theory. This book
now uses R that is a freely available programming language for applied statistics
and data visualization. The file folder accompanying the book contains a set of R
functions that implement various facets of the theory. These R functions allow the
reader to explore the theory at the conceptual level.
The book is organized in a building block fashion. It proceeds from the simple to
the complex with each new topic building on the preceding topics. Within each
of the eight chapters, a basic concept is presented, the corresponding computer
session is explained, and a set of exploratory exercises are defined. Readers are then
strongly encouraged to use the computer session to explore the concept through a
series of exercises. A final section of each chapter, called “Things to Notice,” lists
some of the characteristics of the concept that you should have noticed and some
of the conclusions that you should have reached. If you do not understand the logic
underlying something in this section, you can return to the computer session and try
new variations and explorations until clarity is achieved.

xiii
xiv Introduction

When finished with the book and the computer sessions, the reader should have
a good working knowledge of the fundamentals of item response theory. This book
emphasizes the basics, minimizes the amount of mathematics, and does not pursue
technical details that are of interest only to the specialist. In some sense, you will
be shown only “what you need to know” rather than all the glorious details of the
theory. Upon completion of this book, the reader should be able to interpret test
results that have been analyzed under item response theory by means of programs
such as WINSTEPS (Linacre 2015), BILOG-MG (Zimowski et al. 2002), and
PCLOGIST (Wingersky et al. 1999). Note that WINSTEPS is a current descendant
of BICAL (Wright and Mead 1976), BILOG-MG is the extended version of BILOG
(Mislevy and Bock 1984), and PCLOGIST is the personal computer version of
LOGIST (Wingersky et al. 1982). In order to employ the theory in a practical setting,
the reader should study more advanced books on the applications of the theory such
as Baker and Kim (2004), de Ayala (2009), Embretson and Reise (2000), Nering and
Ostini (2010), Reckase (2009), Thissen and Wainer (2001), and van der Linden and
Glas (2000) as well as some earlier books including Hambleton and Swaminathan
(1984), Hambleton et al. (1991), Wright and Stone (1979), and Hulin et al. (1983).

Getting Started

R is a software package for data analysis and graphical representation. R provides

the language, functions, and the computing environment in one convenient package.
The main uniform resource locator (URL; i.e., webpage) of R is
http://www.r-project.org
You can download R by clicking one of the Comprehensive R Archive Network
(CRAN) mirror sites in
http://cran.r-project.org/mirrors.html
and following the instruction shown on your computer screen for your own operat-
ing system of Linux, Macintosh, or Windows. The base subdirectory contains the
binaries for R. Appendix A contains a brief introductory summary of R.
After installing R on your computer, you can perform all activities shown in
the respective chapters and the computer sessions by typing in the R command
lines exactly shown in the book. Alternatively, for larger R command lines that
may contain R functions in the book, you can obtain and use a zipped file folder
(BIRTRFunctions.zip) that contains R scripts from the publisher’s web site.
Chapter 1
The Item Characteristic Curve

1.1 Introduction

In many educational and psychological measurement situations there is an under-

lying variable of interest. This variable is often something that is intuitively
understood, such as “intelligence.” People can be described as being bright or
average and the listener has some idea as to what the speaker is conveying about
the object of the discussion. Similarly, one can talk about scholastic ability and
its attributes such as gets good grades, learns new material easily, relates various
sources of information, and uses study time effectively. In academic areas, one can
use descriptive terms such as reading ability and arithmetic ability. Each of these
is what psychometricians refer to as an unobservable or latent trait. While such
a variable is easily described and knowledgeable persons can list its attributes, it
cannot be measured directly as can height or weight, since the variable is a concept
rather than a physical dimension. A primary goal of educational and psychological
measurement is the determination of how much of such a latent trait a person
possesses. Since most of the research has dealt with variables such as scholastic,
reading, mathematical, and arithmetic abilities, the generic term “ability” is used
within item response theory to refer to such latent traits.
If one is going to measure how much of a latent trait a person has, it is necessary
to have a scale of measurement, that is, a ruler having a given metric. For a number
of technical reasons, defining the scale of measurement, the numbers on the scale,
and the amount of the trait that the numbers represent is a very difficult task. For the
purposes of the first six chapters, this problem shall be solved by simply defining
an arbitrary underlying ability scale. It will be assumed that, whatever the ability,
it can be measured on a scale having a midpoint of zero, a unit of measurement
of one, and a range from negative infinity to positive infinity. Since there is a unit
of measurement and an arbitrary zero point, such a scale is referred to as existing
at an interval level of measurement. The underlying idea here is that, if one could
physically ascertain the ability of a person, this ruler would be used to tell how much

© Springer International Publishing AG 2017 1

F.B. Baker, S.-H. Kim, The Basics of Item Response Theory Using R,
Statistics for Social and Behavioral Sciences, DOI 10.1007/978-3-319-54205-8_1
2 1 The Item Characteristic Curve

ability a given person has, and the ability of several persons could be compared.
While the theoretical range of ability is from negative infinity to positive infinity,
practical considerations usually limit the range of values from, say, 3 to C3.
Consequently, the discussions in the text and the computer sessions will only deal
with ability values within this range. However, you should be aware that values
beyond this range are possible.
The usual approach taken to measure an ability is to develop a test consisting
of a number of items (i.e., questions). Each of these items measures some facet of
the particular ability of interest. From a purely technical point of view such items
should be free response items where the examinee can write any response that seems
appropriate. The person scoring the test then must decide whether the response
is correct or not. When the item response is determined to be correct, the examinee
receives a score of one, an incorrect answer receives a score of zero, that is, the item
is dichotomously scored. Under classical test theory, the examinee’s raw test score
would be the sum of the scores received on the items in the test. Under item response
theory, the primary interest is in whether an examinee got each individual item
correct or not rather than in the raw test score. This is because the basic concepts of
item response theory rest upon the individual items of a test rather than upon some
aggregate of the item responses such as a test score.
From a practical point of view, free response items are difficult to use in a test.
In particular, they are difficult to score in a reliable manner. As a result, most tests
used under item response theory consist of multiple-choice items. These are scored
dichotomously with the correct answer receiving a score of one and each of the
distractors yielding a score of zero. Items scored dichotomously are often referred
to as binary items.

1.2 The Item Characteristic Curve

A reasonable assumption is that each examinee responding to a test item possesses

some amount of the underlying ability. Thus, one can consider each examinee to
have a numerical value, a score, that places the examinee somewhere on the ability
scale. This ability score will be denoted by the Greek letter theta, . At each ability
level there will be a certain probability that an examinee with that ability will give
a correct answer to the item. This probability will be denoted by P. /. In the case
of a typical test item, this probability will be small for examinees of low ability and
large for examinees of high ability. If one plotted P. / as a function of ability, the
result would be a smooth S-shaped curve such as shown in Fig. 1.1. The probability
of correct response is near zero at the lowest levels of ability and increases until
at the highest levels of ability the probability of correct response approaches unity.
This S-shaped curve describes the relationship between the probability of correct
response to an item and the ability scale. In item response theory, it is known as the
item characteristic curve. Each item in a test will have its own item characteristic
curve.
1.3 Item Difficulty and Item Discrimination 3

1.0
Probability of Correct Response, P(θ)

0.5
0.0

−3 −2 −1 0 1 2 3

Ability, θ

Fig. 1.1 A typical item characteristic curve

The item characteristic curve is the basic building block of item response theory
and all the other constructs of the theory depend upon this curve. Because of this,
considerable attention will be devoted to this curve and its role within the theory.
There are two technical properties of an item characteristic curve that are used to
describe it. The first is the difficulty of the item. Under item response theory, the
difficulty of an item describes where the item functions along the ability scale. For
example, an easy item functions among the low-ability examinees while a hard
item would function among the high-ability examinees; thus, item difficulty is a
location index. The second technical property is the discrimination of an item, which
describes how well an item can differentiate between examinees having abilities
below the item location and those having abilities above the item location. This
property essentially reflects the steepness of the item characteristic curve in its
middle section. The steeper the curve the better the item can discriminate. The flatter
the curve the less the item is able to discriminate since the probability of correct
response at low ability levels is nearly the same as it is at high ability levels. Using
these two descriptors, one can describe the general form of the item characteristic
curve. These descriptors are also used to discuss the technical properties of an item.
It should be noted that these two properties say nothing about whether the item really
measures some facet of the underlying ability or not; that is a question of validity.
These two properties simply describe the form of the item characteristic curve.

1.3 Item Difficulty and Item Discrimination

The idea of item difficulty as a location index will be examined first. In Fig. 1.2,
three item characteristic curves are presented on the same graph. All have the same
level of item discrimination but differ with respect to item difficulty. The left-hand
4 1 The Item Characteristic Curve

1.0
Probability of Correct Response

0.5
0.0

−3 −2 −1 0 1 2 3

Ability

Fig. 1.2 Three item characteristic curves with the same item discrimination but different levels of
item difficulty

curve represents an easy item because the probability of correct response is high
for low-ability examinees and approaches 1 for high-ability examinees. The center
curve represents an item of medium item difficulty because the probability of correct
response is low at the lower ability levels, around 0:5 in the middle of the ability
scale, and near 1 at the highest ability level. The right-hand curve represents a hard
item. The probability of correct response is low for most of the ability scale and
increases only when the higher ability levels are reached. Even at the highest ability
level shown (i.e.,  D 3) the probability of correct response is only 0:8 for the most
difficult item.
The concept of item discrimination is illustrated in Fig. 1.3. This figure contains
three item characteristic curves having the same item difficulty but differing with
respect to item discrimination. The upper curve on the positive ability range has a
high level of item discrimination since the curve is quite steep in the middle where
the probability of correct response changes very rapidly as ability increases. Just
a short distance to the left of the middle of the curve, the probability of correct
response is much less than 0:5; and a short distance to the right, the probability is
much greater than 0:5. The middle curve represents an item with a moderate level
of item discrimination. The slope of this curve is much less than the previous curve
and the probability of correct response changes less dramatically than the previous
curve as the ability level increases. However, the probability of correct response
is near zero for the lowest-ability examinees and near unity for the highest-ability
examinees. The third curve represents an item with low item discrimination. The
curve has a very small slope and the probability of correct response changes slowly
over the full range of abilities shown. Even at low ability levels, the probability of
correct response is reasonably large and it increases only slightly when high ability
levels are reached. The reader should be warned that although the figures only show
a range of ability from 3 to C3, the theoretical range of ability is from negative
1.3 Item Difficulty and Item Discrimination 5

1.0
Probability of Correct Response

0.5
0.0

−3 −2 −1 0 1 2 3

Ability

Fig. 1.3 Three item characteristic curves with the same item difficulty but different levels of item
discrimination
1.0
Probability of Correct Response

0.5
0.0

−3 −2 −1 0 1 2 3

Ability

Fig. 1.4 An item that discriminates perfectly at  D 1:5

infinity to positive infinity. Thus, all item characteristic curves of the type used here
actually become asymptotic to a probability of zero at one tail and to unity at the
other tail. The restricted range employed in the figures is necessary to fit the curves
on the computer screen reasonably and to provide a uniform frame of reference.
One special case is of interest; namely, that of an item with perfect discrimina-
tion. The item characteristic curve of such an item is a vertical line at some point
along the ability scale. Figure 1.4 shows such an item. To the left of the vertical line
at  D 1:5, the probability of correct response is zero and to the right of the line
the probability of correct response is unity. Thus, the item discriminates perfectly
6 1 The Item Characteristic Curve

between examinees whose abilities are above and below an ability score of 1:5. Such
items would be ideal for distinguishing between examinees with abilities just above
and below 1:5. However, such an item makes no distinction among those examinees
with abilities above 1:5 nor among those examinees with abilities below 1:5.

1.4 Verbal Terms of Item Difficulty and Item Discrimination

At the present point in the presentation of item response theory, the goal is to allow
you to develop an intuitive understanding of the item characteristic curve and its
properties. In keeping with this goal, item difficulty and item discrimination will be
defined in verbal terms. Item difficulty will have the following levels:
very easy
easy
medium
hard
very hard
Item discrimination will have the following levels:
none
low
moderate
high
perfect
These terms will be used in the computer session to specify item characteristic
curves.

1.5 Computer Session

The purpose of this session is to enable you to develop a sense of how the shape of
the item characteristic curve is related to item difficulty and item discrimination.
To accomplish this, you will be able to select verbal terms describing the item
difficulty and item discrimination of an item. The computer program R will then
calculate and display the corresponding item characteristic curve on the screen. You
should do examples in this section and exercises in the next section, then try various
combinations of levels of item difficulty and item discrimination and relate these to
the resulting curves. After a bit of such exploratory practice, you should be able to
predict what the item characteristic curve will look like for a given combination of
item difficulty and item discrimination.
1.5 Computer Session 7

1.5.1 Procedures for an Example Case

When R is ready for input through the R console window, it prints out its prompt
character with an invisible, horizontal space after it:
>
A command line in R will be executed by pressing the enter key:
Enter
-

Such a special character that indicates the end of a command line, usually entered
by pressing the enter or return key will be treated as an invisible character here.
When an incomplete command line (e.g., the end of the expression cannot have
occurred yet) gets the enter key, R prints out the continuation prompt character with
an invisible space after it:
+
To improve readability and because a rather long command line can be typed in
without pressing the enter key in the middle, a long command line will not be
separated by the continuation prompt character but will be continued to the next
line with indentation.
The followings are the simplest command lines, each with the R prompt in front,
to display an item characteristic curve for an item with medium item difficulty and
moderate item discrimination:
> theta <- seq(-3, 3, .1)
> bmedium <- 0
> amoderate <- 1
> P <- 1 / (1 + exp(-amoderate * (theta - bmedium)))
> plot(theta, P, type="l")
By pressing the enter key in the end of each line, five times as a total, the computer
will display an item characteristic curve shown in Fig. 1.5 in the graphics window.
By pressing the enter key in the end of the first line, a sequence of numbers (i.e., a
vector) will be created with 3 as a starting number and 3 as an ending number with
an increment of 0:1. The length or the total number of elements of the sequence is
61. The name of such a sequence is given as theta, and it is to be done by using the
assign function <- for which the less than character < and the hyphen character -
cannot be separated with a space. You can read the first line as “theta gets a sequence
of numbers . . .” or “a sequence of numbers . . . is saved under the name theta.” The
first line is equivalent to:
> assign("theta", seq(-3,3,.1))
or
> seq(-3, 3, .1) -> theta
8 1 The Item Characteristic Curve

0.8
0.6
P

0.4
0.2

−3 −2 −1 0 1 2 3

theta

Fig. 1.5 An example item characteristic curve

Use of spaces is optional in the R command line as shown in seq(-3,3,.1).

Especially, spaces before and after a comma, a left or right parenthesis, an
elementary arithmetic operator (e.g., +, -, *, /), an exponentiation operator (i.e., ^,
**), a common mathematical function (e.g., log, exp, sin, cos, tan, sqrt), a
logical operator (e.g., <, <=, >, >=, ==, !=, &, |), and an assign function (e.g., ->,
<-) are all optional. To improve readability of the code, at least one space can be
put before and after some operators or symbols.
Multiple command lines can be typed in a single line by separating them with a
semicolon. Real numbers in the integer form can be entered with a decimal place
(i.e., a period) in the end of the number, and the number 0 can be added after the
decimal place. The number 0 can be added before the decimal place if the number
is less than unity that is to be expressed as a decimal. The second and third lines of
the above command lines can be combined as:
> bmedium <- 0; amoderate <- 1
Equivelently, we may use the numbers with a decimal place as:
> bmedium <- 0.0; amoderate <- 1.0
The item characteristic curve is a function of ability and item characteristics, that
is, item difficulty and item discrimination. Because ability is a latent variable, it was
denoted by  and the variable name theta was used in the command line. The
item difficulty and item discrimination are denoted as numerical values of b and a,
respectively, and the verbal terms contain the letters b and a as mnemonic clues.
The names of objects in the R command can be constructed with the upper- and
lower-case roman letters as an initial character and the digits and the period as any
non-initial characters. Names of built-in, intrinsic functions and system variables
(e.g., seq, exp, plot, type in the above lines, and c, q, s, t, C, D, F, I, T,
1.5 Computer Session 9

diff, length, mean, sd, pi, range, rank, time, tree, var, etc.) should
not be used as the names of variables or functions of your own. Note that R is case
sensitive, so the variable P will be different from the variable p unless both are
defined to be equivalent.
The command line
P <- 1 / (1 + exp(-amoderate * (theta - bmedium))))
will create a vector of 61 values of the probabilities of correct response based on the
respective 61 ability points, item difficulty, and item discrimination. We will explore
the exact meaning of this function in the subsequent chapter.
The plot based on the 61 sets of points from the ability variable as the horizontal
axis, abscissa, and the probability of correct response as a vertical axis, ordinate,
can be constructed via:
> plot(theta, P)
With the enter key pressed, the above line can open an R graphics window that
contains the plot of the two variables. The default setting of the function plot will
create a plot with a symbol ı as each point. Because we want to have a plot with
connected lines that ultimately yield a curve, the optional argument type="l"
was added in the earlier command line. Whenever R opens up its graphics window,
it treats the graphics window as a current window. If you want to continue to use
command lines, you may click the R console window (especially the caption on the
top of the R console window or any inside portion of the R console window) to
make it current before you type in a new command line.
You may notice that the number of ticks used in the default setting of the function
plot may not be an appropriate one you want to use in your own figure. The
numbers of ticks in the horizontal and vertical axes can be modified with the use
of the graphical parameters function par and its labels argument lab, for example:
> par(lab=c(7,3,3))
The labels argument was defined by the three parameters in the above line. The set
of the three parameters (i.e., the three numbers separated with two commas in the
combine function c) specifies the number of ticks on the horizontal axes to be 7,
that on the vertical axes to be 3, and the length of characters in the labels to be 3
(but the character length will be most likely ignored in R). If you want to explore
or read the full description of the R function, for example par, you can obtain it by
typing:
> ?par
or
> help(par)
Assuming that your computer is connected to Internet, such a command line will
open up a file in html, the HyperText Markup Language, that explains the function
in a default browser you are using.
10 1 The Item Characteristic Curve

It can be noticed in Fig. 1.5 that the variable names are appeared as the respective
labeling texts along the horizontal and vertical axes. You can change them with the
arguments of, for example, xlab="Ability" and ylab="Probability of
Correct Response" in the function plot. The ranges of the horizontal and
vertical variables can be precisely controlled with the use of the arguments and
parameters of xlim=c(-3,3) and ylim=c(0,1), by specifying the lower limit
number and the upper limit number in the combine or concatenate function c.
The following command lines can be used to obtain Fig. 1.1:
> par(lab=c(7,3,3))
> theta <- seq(-3, 3, .1)
> b <- 0
> a <- 1
> P <- 1 / (1 + exp(-a * (theta - b)))
> plot(theta, P, type="l", xlim=c(-3,3), ylim=c(0,1),
xlab=expression(paste("Ability, ",theta)),
ylab=expression(paste(
"Probability of Correct Response, P(",theta,")")))
As shown in Fig. 1.1, you may use the function expression to add characters in
Greek. We may simply use xlab="Ability" and ylab="Probability of
Correct Response". You can add the main heading on the top of the plot and
the subheading to the bottom of the plot with the arguments in the function plot.
These will be helpful when you are making figures for a presentation purpose. For
example, you may construct a figure (n.b., a figure is not displayed here) that is more
elaborate but nevertheless similar to Fig. 1.5 using the following command lines:
> par(lab=c(7,3,3))
> theta <- seq(-3, 3, .1)
> bmedium <- 0
> amoderate <- 1
> P <- 1 / (1 + exp(-amoderate * (theta - bmedium)))
> plot(theta, P, type="l", xlim=c(-3,3), ylim=c(0,1),
xlab="Ability", ylab="Probability of Correct Response",
main="Figure 1. An Item Characteristic Curve with
Medium Item Difficulty and Moderate Item Discrimination",
sub="See Baker and Kim (2016).")

1.5.2 An R Function for Item Characteristic Curves

It is possible to create your own R functions. Each time the plot of the item
characteristic curve is created, you may notice that a nearly identical set of R
command lines are executed. To avoid the repetition of typing in the same command
1.5 Computer Session 11

lines, an R function for plotting an item characteristic curve can be constructed.

Consider the following function named iccplot:
> iccplot <- function(b, a) {
par(lab=c(7,3,3))
theta <- seq(-3, 3, .1)
P <- 1 / (1 + exp(-a * (theta - b)))
plot(theta, P, type="l", xlim=c(-3,3), ylim=c(0,1),
xlab="Ability", ylab="Probability of Correct Response")
}

After defining the function, by specifying the numerical values of item difficulty
and item discrimination, a plot of the item characteristic curve can be constructed
in the R graphics window. For example, you may use the following line to obtain
a plot of item characteristic curve with medium item difficulty and moderate item
discrimination:
> iccplot(0, 1)
The two arguments, b and a, are the named actual arguments in R. When the
numerical values are specified without the arguments, R recognizes the first number
to be the value of item difficulty and the second number to be the value of item
discrimination. The arguments in the function can be specified in arbitrary order by
exactly defining them with names. You can obtain the same plot by typing:
> iccplot(a=1, b=0)
Instead of using the numerical values to define item difficulty and item discrim-
ination, the verbal terms described earlier can be used to plot item characteristic
curves. The numerical definitions of the verbal terms for item difficulty are as
follows:
> bveryeasy <- -2.625
> beasy <- -1.5
> bmedium <- 0
> bhard <- 1.5
> bveryhard <- 2.625
The numerical definitions of the verbal terms for item discrimination are as follows:
> anone <- 0
> alow <- 0.4
> amoderate <- 1
> ahigh <- 2.1
> aperfect <- 999
The following command lines now can display two item characteristic curves
in the R graphics window (see Fig. 1.6). Note that you should click the R console
window after pressing the enter key in the end of the first line, that is, after creating
the R graphics window.
12 1 The Item Characteristic Curve

1.0
Probability of Correct Response

0.5
0.0

−3 −2 −1 0 1 2 3

Ability

Fig. 1.6 Two item characteristic curves

> iccplot(bmedium, amoderate)

> par(new=T)
> iccplot(beasy, alow)
The first item characteristic curve is for an item with medium item difficulty and
moderate item discrimination. The line
> par(new=T)
is an overlying figures parameter and equivalent to:
> par(new=TRUE)
The line resets a graphics parameter new to be TRUE, meaning that the R graphics
window is treated now as a new graphics device, so it is assumed that there are
currently no plots on it. In such a case, a call to a high-level plotting function
could not erase the canvas of the graphics window before putting up a plot. After
entering the third line, you may notice that the second item characteristic curve
was overlaid on the same graph as the previous curve and you can compare the
two. The new curve is rather flat and has higher probabilities of correct response in
the lower range of abilities than did the previous item characteristic curve. This is
because it was an easier item and low-ability examinees should do well on it. The
low item discrimination shows up in the curve having only a slight bend over the
range of ability scores employed. At the high ability levels the probability of correct
response was somewhat lower than that of the previous item. This is reflection of
the lower discrimination on the new item. When the function is specified without
par(new=T), a new plot is created in the R graphics window.
Note that you can obtain a figure that contains several item characteristic curves.
For example, to obtain Fig. 1.2 you can type the following command lines with
1.6 Exercises 13

clicking the console window after the first line and the third line, respectively:
> iccplot(beasy, amoderate)
> par(new=T)
> iccplot(bmedium, amoderate)
> par(new=T)
> iccplot(bhard, amoderate)

1.6 Exercises

For the exercises, it is assumed that you have defined the function iccplot and
the numerical values of the verbal terms of item difficulty and item discrimination
by typing in the respective command lines.
1. An item with easy item difficulty and high item discrimination is to be plotted.
(a) Use the function iccplot to plot an item characteristic curve of an item
with easy item difficulty and high item discrimination.
(b) From the graph it can be seen that the probability of correct response will be
rather high over most of the ability scale. The item characteristic curve will
be steep in the lower part of the ability scale.
(c) After you have studied the curve, make sure to click the R console window
to make it current and to type in a new command line.
(d) The next graph will be plotted in a new graphics window.
2. An item with hard item difficulty and low item discrimination is to be plotted.
(a) Use the function iccplot to plot an item characteristic curve of an item
with hard item difficulty and low item discrimination.
(b) From the graph it can be seen that the probability of correct response will
have a low general level over most of the ability scale. The item characteristic
curve will not be very steep.
(c) After you have studied the curve, make sure to click the R console window
to make it current and to type in a new command line.
(d) The next graph will be plotted in a new graphics window.
3. An item with medium item difficulty and low item discrimination is to be
plotted.
(a) Use the function iccplot to plot an item characteristic curve of an item
with medium item difficulty and low item discrimination.
(b) From the graph it can be seen that the probability of correct response will be
between 0.2 and 0.8 over the range of ability shown. The item characteristic
curve will be nearly linear over the range of ability employed.
(c) After you have studied the curve, make sure to click the R console window
to make it current and to type in a new command line.
(d) The next graph will be plotted in a new graphics window.
14 1 The Item Characteristic Curve

4. In this exercise, all the items will have the same level of item difficulty but
different levels of item discrimination. The intent is to relate the steepness of
the curve to the level of item discrimination.
(a) Use the function iccplot to plot an item characteristic curve of an item
with medium item difficulty and moderate item discrimination.
(b) From the graph it can be seen that the probability of correct response will
be small at low ability levels and large at high ability levels. The item
characteristic curve will be moderately steep in the middle part of the ability
scale.
(c) After you have studied the curve, make sure to click the R console window
to make it current and to type in a new command line.
(d) The next graph will be plotted on the same graph in the graphics window.
Type in:
> par(new=T)
(e) Now repeat steps a through d several times using medium item difficulty
for each item and item discrimination levels of your choice (e.g., none, low,
high, perfect).
(f) The next graph will be plotted in a new graphics window.
5. In this exercise, all the items will have the same level of item discrimination but
different levels of item difficulty. The intent is to relate the location of the item
on the ability scale to its level of item difficulty.
(a) Use the function iccplot to plot an item characteristic curve of an item
with very easy item difficulty and moderate item discrimination.
(b) From the graph it can be seen that the probability of correct response will be
reasonably large over most of the ability scale. The item characteristic curve
will be moderately steep in the lower part of the ability scale.
(c) After you have studied the curve, make sure to click the R console window
to make it current and to type in a new command line.
(d) The next graph will be plotted on the same graph in the graphics window.
Type in:
> par(new=T)
(e) Now repeat steps a through d several times using items with moderate item
discrimination and item difficulty levels of your choice (e.g., easy, medium,
hard, very hard).
(f) The next graph will be plotted in a new graphics window.
6. Experiment with various combinations of item difficulty and item discrimination
of your own choice until you are confident that you can predict the shape of
the item characteristic curve corresponding to the levels chosen. You may find it
useful to make a rough sketch of what you think the curve will look like before
you have the computer display it on the screen.
1.7 Things to Notice 15

1.7 Things to Notice

1. When item discrimination is less than moderate, the item characteristic curve is
nearly linear and appears rather flat.
2. When item discrimination is greater than moderate, the item characteristic curve
is S-shaped and rather steep in its middle section.
3. When item difficulty is less than medium, most of the item characteristic curve
has a probability of correct response that is greater than 0:5.
4. When item difficulty is greater than medium, most of the item characteristic
curve has a probability of correct response less than 0:5.
5. Regardless of the level of item discrimination, item difficulty locates the item
along the ability scale. Therefore item difficulty and item discrimination are
independent of each other.
6. When an item has no item discrimination, all choices of item difficulty yield the
same horizontal line at a value of P. / D 0:5. This is because the value of item
difficulty for an item with no item discrimination is undefined.
7. If you have been very observant, you may have noticed the point at which P. / D
0:5 corresponds to item difficulty. When an item is easy, this value occurs at a
low ability level. When an item is hard, this value corresponds to a high ability
level.
Chapter 2
Item Characteristic Curve Models

2.1 Introduction

In the first chapter the properties of the item characteristic curve were defined in
terms of verbal descriptors. While this is useful to obtain an intuitive understanding
of item characteristic curves, it lacks the precision and rigor needed by a theory.
Consequently, in this chapter the reader will be introduced to three mathematical
models for the item characteristic curve. These models provide mathematical
equations for the relation of the probability of correct response to ability. Each
model employs one or more item parameters whose numerical values define a
particular item characteristic curve. Such mathematical models are needed if one
is to develop a measurement theory that can be rigorously defined and is amenable
to further growth. In addition, these models and their parameters provide a vehicle
for communicating information about an item’s technical properties. For each of the
three models, the mathematical equation will be used to compute the probability of
correct response at several ability levels. Then the graph of the corresponding item
characteristic curve will be shown. The goal of the chapter is to have you develop a
sense of how the numerical values of the item parameters for a given model relate
to the shape of the item characteristic curve.

2.2 The Two-Parameter Model

Under item response theory the standard mathematical model for the item char-
acteristic curve is the cumulative form of the logistic function. It defines a family
of curves having the general shape of the item characteristic curves shown in the
first chapter. The logistic function was first derived in 1844 by Pierre François
Verhulst and has been widely used in the biological sciences to model the growth
of plants and animals from birth to maturity. It was first used as a model for the

© Springer International Publishing AG 2017 17

F.B. Baker, S.-H. Kim, The Basics of Item Response Theory Using R,
Statistics for Social and Behavioral Sciences, DOI 10.1007/978-3-319-54205-8_2
18 2 Item Characteristic Curve Models

item characteristic curve in the late 1950s and, because of its simplicity, has become
the preferred model. The equation for the two-parameter logistic model is given in
Eq. (2.1) below:

1 1
P. / D L
D a.b/
; (2.1)
1Ce 1Ce

where
e is the base of the natural logarithm that is a constant 2.718,
b is the item difficulty parameter,
a is the item discrimination parameter,1
L D a.  b/ is the logistic deviate (logit), and
 is an ability level.
This equation represents a family of curves whose individual members are defined
by specific numerical values of the item parameters b and a; hence, it is called the
two-parameter model. It is the model that was used in Chap. 1.
The item difficulty parameter, denoted by b, is defined as the point on the ability
scale at which the probability of correct response to the item is 0.5. The theoretical
range of the values of this parameter is 1  b  C1 (i.e., a set of extended real
numbers). However, typical values have the range of 3  b  C3.
Due to the shape of the item characteristic curve, the slope of the curve changes
as a function of the ability level and reaches a maximum value when the ability
level equals the item difficulty parameter. Because of this, the item discrimination
parameter does not represent the general slope of the item characteristic curve as was
indicated in Chap. 1. The technical definition of the item discrimination parameter is
beyond the level of this book. However, a usable definition is that this parameter is
proportional to the slope of the item characteristic curve at  D b. The actual slope
at  D b is a=4 under the two-parameter model, but considering a to be the slope at b
is an acceptable approximation that makes interpretation of the item discrimination
parameter easier in practice. The theoretical range of the values of this parameter is
1  a  C1, but the usual range seen in practice is 2:80 to C2:80.

1
In much of the item response theory literature, the logistic value of the item discrimination
parameter a is divided by 1:702 or 1:7 to obtain the corresponding normal ogive model value.
This is done to make the two-parameter logistic ogive similar to the normal ogive. However, this
was not done in this book as it introduces two frames of reference for interpreting the numerical
values of the item discrimination parameter. All item parameters in this book and the associated
computer programs are interpreted in terms of the logistic function.
2.2 The Two-Parameter Model 19

2.2.1 Computational Example

To illustrate how the two-parameter model is used to compute the points on an item
characteristic curve, consider the following example problem. The values of the item
parameters are:
b D 1:0 is the item difficulty parameter
a D 0:5 is the item discrimination parameter
The illustrative computation is performed at the ability level:
 D 3:0
The first term to be computed is the logistic deviate (logit), L, where:
L D a.  b/
Substituting the appropriate values yields:
L D 0:5.3:0  1:0/ D 2:0
The next term computed is e raised to the power L. If you have a pocket calculator
that can compute ex or exp.x/ you can verify this calculation. Substituting yields:
eL  exp.L/ D exp.2:0/ D 7:389
Now the denominator of Eq. (2.1) can be computed as:
1 C exp.L/ D 1 C 7:389 D 8:389
Finally, the value of P. / is:
P. / D 1=Œ1 C exp.L/ D 1=8:389 D 0:119
Thus, at an ability level of  D 3:0, the probability of responding correctly to this
item is 0:119.
From the above, it can be seen that computing the probability of correct response
at a given ability level is very easy using the logistic model. Table 2.1 shows the
calculations for this item at seven ability levels evenly spaced over the range of
ability levels from 3 to C3. You should perform the computations at several of
these ability levels to become familiar with the procedure.

Table 2.1 Item characteristic Ability,  Logit, L exp.L/ 1 C exp.L/ P. /

curve calculations under the
two-parameter model, 3.0 2.0 7.389 8.389 0.119
b D 1:0 and a D 0:5 2.0 1.5 4.482 5.482 0.182
1.0 1.0 2.718 3.718 0.269
0.0 0.5 1.649 2.649 0.378
1.0 0.0 1.000 2.000 0.500
2.0 0.5 0.607 1.607 0.622
3.0 1.0 0.368 1.368 0.731
20 2 Item Characteristic Curve Models

1.0
Probability of Correct Response

0.5
0.0

−3 −2 −1 0 1 2 3

Ability

Fig. 2.1 The item characteristic curve for the two-parameter model with b D 1:0 and a D 0:5

The item characteristic curve for the item of Table 2.1 is shown in Fig. 2.1. The
vertical dotted line corresponds to the value of the item difficulty parameter.

2.3 The Rasch Model

The next model of interest was first published by the Danish mathematician Georg
Rasch in the 1960s. Rasch approached the analysis of test data from a probability
theory point of view. Although he started from a very different frame of reference,
the resultant item characteristic curve model was a logistic model. In Chap. 8,
Rasch’s approach will be explored in greater detail; our present interest is only
in his item characteristic curve model. Under this model, the item discrimination
parameter of the two-parameter model is fixed at a value of a D 1 for all items and
only the item difficulty parameter can take on different values. Because of this, the
Rasch model is often referred to as the one-parameter logistic model.
The equation for the Rasch model is given by the following:

1
P. / D ; (2.2)
1C e1.b/

where
b is the item difficulty parameter and
 is the ability level.
It should be noted that the item discrimination parameter was used in Eq. (2.2). But
because it always has a value of 1 it usually is not shown in the formula.
Exploring the Variety of Random
Documents with Different Content
KASTEEN HELMEITÄ

Aamusella astuelin tuolla kukkanurmella, uinaileva luonto

siinä henkäili niin hiljaisna.

Lintuin äänet yksin raikui

elon tointa ilmaisten,
ne kun lauloi laulujansa
Herraa hyvää kiitellen.

Aamun kirkas valo voitti,

öisen sumut karkoitti,
aurinkoisen säde koitti,
kaiki eloon heräsi.

Kukka kaunis päätään nosti,

kyynel sillä silmästään,
kirkkahana helmiellen
alas maahan vierähtää:

Miksi itket, kaunis kukka?

Suutelonhan hellän soi
sulle armas aurinkoinen
ja sun umpus auki loi.
 Sulo-tuoksullasi sitten
hurmaat meidät ihmiset.
Etkö tuota nähdessäsi
iloitse? Oi, etpä, et. —

Ah, jo tiedän, miksi itket:

onhan luonto turmeltu;
viatonta siin' ei löydy,
hyv' on pahaks muutettu.

Eikä ihme, että sulla kyyneleet käy vuotamaan; — paremp'

ois sun, ettet enää silmääs auki loisikaan!
KUKAN KUISKE

Kukka ihanainen kasvoi metsässä, siinä valittaen yksin

jääneensä: "Sisko mulla tuossa oli vieressäin, nyt on hänkin
poissa, ainut ystäväin".

Neito hiljallensa astui metsässä, kuuli kukan kuiskeen, otti

käteensä, armahalleen antoi kotiin tullessaan: nuorukainen
painoi vasten poveaan.
KUKAN HOITO

Kukkaa kaunokaista hoida hellästi; varo, ettei helle häntä

kuivaisi, vesain juuret hennot kätke multahan, mutta latvat
nosta ylös ilmahan.

Ja kun maahan juuret syvään tunkeutuu,

näet, miten niistä nousee kukkapuu.
Toivos täyttyneeksi vasta silloin näät,
edessäs kun näet kukat täyteläät.

Niinkuin kukka auringosta kasvun saa, ihminenkin valon

lahjaa kaipajaa. Valon lahjan Luoja hälle parhaan suo, kun
hän maansa eteen työnsä uhriks tuo.
SYDÄMENI LAATU

Ah, jo tunnen sydämeni laadun: niinkuin virtaa tuuli velloaa,

sydämeni samoin kuohahtaapi, sen kun tunteen tuuli liikuttaa.

Tai kuin aurinkoisen valo taittaa

sätehiään veden kalvossa,
sydän onnen-kuviaan myös luopi,
jotka särkyvät taas samassa.

Ja kun myrskyn sauhu pauhinalla vettä kohden rantaa

kuohuttaa, niin myös rinnastani myrskyn aallot huokauksiin
ilmi puhkeaa.
KUKKANURMI

Kukat nurmella loistavat hohdossaan, kesäpäivyt kun lämpöä

suo; vaan kumma, kun katselen tarkemmin, miks sulkemat
nuppunsa nuo?

Yön usma ja viima niin kylmästi

kävi hienoihin helpeisiin,
siks painuvi runko ja peittyvi pää
suruharsosen suojuksiin.

Vaan toivoni onpi, nietos ei voi sen juuria jäädyttää: Maa,

josta se elonsa nesteen joi, sen uudesti synnyttää.
IHANTEET

Kuu heitti valoansa ja tähdet kirkkahat ne veteen kuvastuivat,

"kuin siin' ois taivahat". — Niin mietti pieni tyttö, mi lähteen
reunalla siin' ihmetellen seisoi, ihaillen taivasta.

"Oi onko taivas tuossa, tuon veden pohjassa, ja onko siellä

maata ja puita, kukkia? Oi, kuinka haluaisin ma nähdä
näitäkin! Nyt äidille mä kerron: on taivas toinenkin."

Vaan vieno tuulen henki veen pintaan värähtää, ja kuva-

taivahasta et mitään silmään jää. Ei ollutkaan se totta: tuo
lähde vietteli, ja tosi-taivas yksin tuoll' ylhääll' yleni.

Monesti ihminenkin luo luulo-onnelan, luo siihen

ihanteensa, tuon sarjan kirjavan, vaan onnettaren oikku ne
pian runtelee, ja kauniit unelmamme niin tyhjään raukenee.
KOSKEN KUOHU

Oi, kosken kuohu vaahteva, sa miten liet niin raivoisa, mi yhä

riehut, rauhaton, kun luonto muuten tyyni on?

Ah, jopa tiedänkin mä sen:

sa olet kuva sydämen,
kenenkä sydän rauhaton,
kuin kuuma koski on.

Oi, sydän, miksi saanutkaan

oot lailla kosken pauhaamaan,
mi aaltos aina kuohuun luot
ja tuntees suotta ilmi tuot?

Vaan viimein Luojan käskien tuo koski talttuu kuohuinen,

kun taisteluista vapaana on sydän löynnyt lepoa.
SALOLAMPI

Vuoren jylhäkön kupehella salolampi niin tyynenä on, sen

kalvoa tuulosen henki vain tuutinut on lepohon.

Ja taivahan sinervä laki

sen pintahan loistoa luo;
sitä tuuhea lehvikkö suojaa,
kukat rannalla tuoksuaan tuo.

Ma mietin: ah lammen jos lailla muin sydän myös rauhaisa

ois, johon vihurin viima ei koskis, ja taivaskin loistoa lois!
ILTALAULU

Hymyellen lempeänä aurinkoinen laskeuu lännen rantaan,

sille siinä kehto vieno valmistuu.

Illan hohde kukkulamme

pukee punapurppuraan,
vieno tuuli tuutielee
kaikki kasvit nukkumaan.

Linnun laulu sointuvainen

lehdossa jo lakkasi,
öisen henki helmahansa
luonnon kaiken anasti,

kunnes taasen valon voima pimeyden poistaapi, kuolleet

eloon herättääpi, — aamu nuori loistaapi.
ILTATÄHTI

Kun iltasella istun ja tähtiä katselen,

niin johtuu mieleheni tuo aika entinen:
kuin moni onnen tähti sen taivaan valaisi —
vaan virvatulta ollen, pian tyhjään hälveni.

Sa iltatähti kaunis ja kaikist' ihanin,

tien osote oot varmaan, mi viittaat kotihin.
Siis sammukoot muut tähdet, nuo harhavalot, vaan!
Sun valos armahaisin ei sammu konsanaan!
AURINGON LASKU

Aurinkoinen armahainen, länteen olet lähtevä, jäähyväiset

olet meille jälkehesi jättävä. Matka sull' on edessäsi, vaikka
toivot lepoa, lännen piirto purppurainen peittää sinut
vaippaansa. Kultakehdoss' et sä viihdy, unen luotas karkoitat,
aamuruskon kultasiivin ylöspäin taas kohoat.

III. LEMMEN MUISTOJA JA

TARINOITA
VALITUS

Sammu, valo aurinkoisen, piile pilven taakse vaan! Öisen

usma, levittele varjosi nyt maailmaan!

Siinä sydämeni tunteet

ehkä uneen nukkuvi,
vaan tuo kaunis kesäilta
säilyy muistiss' iäti.

Ilta kaunis nyt kuin silloin,

koska kultain erosi:
aurinko se lähti mailleen,
vuorten kupeet kultaili.

Nyt on kaikki autiota, poissa sulo entinen; kussa ennen

onnen löysin, vaivun valtaan kaihojen!
POIS, POIS!

Pois, pois! Näin rinnassani soi.

Mi kumma tunne minut valtioi!
Pois, pois! Ma täältä toivoisin
yön ikikeston tummaisiin tupihin.

Pois, pois! — Ei ollut ennen niin,

sun vaivuin minä onnen unelmiin.
Pois, pois! — Voi, ettei vienyt ois,
mun armast' ystäväistäni Tuoni, pois!

Pois, pois! — On lempi viehkeä,

mut ystävä nyt lepää kylmänä,
Pois, pois! Sen silmä suloinen
nyt vaipunut on iäks unehen.

Pois, pois! Mun salli kanssasi

nyt jättää elon levoton maininki!
Pois, pois! Jo kutsuu kuolo tuo,
se levätä mun luonas jälleen suo.
HILJAA

Hiljaa, hiljaa soitto kuuluu tuolta laakson helmasta. —

Lemmen jumalatar siellä soittaneeko kannelta?

Hiljaa, hiljaa soitto kuuluu,

kuuluu yhä hiljempään:
impi siinä istui nuori, —
mitä mietti mielessään?

Hiljaa, Hiljaa soitto kuuluu;

immen poski kalpenee,
ja hän nurmikolle vaipuu
viereen vienon kanteleen.

Hiljaa, hiljaa soitto kuuluu, kuuluu vain kuin valitus. —

Nousee mulla'i rinnastani, hiljaa tumma huokaus.
KEINUSSA

Ma keinussa yksinäin istun näin illalla kuutamoyön. — Oi,

annanko keinua hiljaa ja mieleni mietteihin lyön? Ei, vaan
korkeelle nouskohon keinu, kuin ennenkin muinoisin vaan,
kun impeni kanssa ma kiidin, ja unholaan jätimme maan!

Vaan hervonnut käteni vaipuu, pois, poissa mun impeni on;

unelmoissa nyt viihtyä tahdon: seis, keinuni malttamaton! viel'
impeni valkean liinan toki kerran heiluvan näin, mut lyhyt oli
lempeni muisto, — ja yksinpä taasen ma jäin.
KULTAANSA IKÄVÖIVÄ RANNALLA

Mä rannalle käyskelen yksinäin vaan, kun ilta jo varjonsa

luopi. Ah, tokkopa sieltä mä lohtua saan, vai tuskanko mulle
se tuopi? Oi, aalto, sa sääliös sydäntäni! — Vaan huoaten
virkat sa viestejäsi.

Sä sanoman saatat tuon murheellisen: täss' armaani hauta

nyt luodaan. Ma rannalla istun ja itkien ma katson kuin luitasi
tuodaan, ne rannalle lainehet lakaisee ja muistoksi minulle
heittelee.

Ja meri se synkän katseensa luo ja surusta ain' yhä

pauhaa, ja hoivaksi mulle taivaalta suo kuu himmeä hymyään
lauhaa, ja tähdet mua vienosti katselevat, vaan kaikk' ovat
lohdut tenhottomat.

Ma nimesi rakkahan kirjoittaa nyt santahan tahdon, oi kulta,

ja kukan myös tahdon ma istuttaa: se lemmen muisto on
multa. Taas rannalle sitten mä käyskellä saan, mun kukkaani
itkulla kostuttamaan.
PETETTY POIKA

Ulapalla aallokossa vene nopsaan liitääpi; venehessä istuu

poika, kullan luo mi kiitääpi. Lemmestä kun sydän sykkii, rinta
riemuin aaltoaa, odotellen onnen rantaa silmä ahnas
katsastaa.

Määräpäähän päästyänsä. toivoi saavans' sulkea

rakkahansa rinnoillehen, iki-siteen solmita; vaan — oi, horna!
— petettynä poika maahan vaipuupi; tyhjä virvatuli oli
onnensa, min kuvaili.

"Laine, mulle lempes anna, viihdytä mun tunteeni! Raivoava

olet sinä, niinkuin oma rintani. Lemmestä se ennen sykki,
uskollisna alati, vaan mun kultain minut petti: turha siis on
eloni."
LEMPIVÄISET

Jo öisen huntu nyt hulmahtaapi ja lepoon luonto jo uinahtaa,

vaan aatoksissansa impi yksin tuoll' asteleepi päin rantamaa.
Mi hänet valtaa nyt surun kauhu? — Ah, suoko lepoa aallon
pauhu?

Ja Aallotarten luo tultuansa

Hän niiden kuuntelee laulantaa
ja näkee Vellamon tuutivana:
se häntä luokseen nyt viittoaa, —
"Jos syliin Vellamon tuonne vaipuis,
niin silloin huoletkin kaikki haipuis!”

"Jää hyvästi, sinä isän koti!

Kenties en näe sua konsanaan.
Ma sulo-muistona rintaan suljen
tuon lapsuus-onnelan ainiaan.
Oi, lempi, onni, ne elon luovat;
vaan miksi mulle ne tuskan tuovat?"

Näin impi istuvi hetken vielä ja aallon loisketta kuuntelee ja

odotellen hän katsoo, katsoo: jo vene verkkahan lähenee. Se
sulhons' onpi, mi siellä rientää ja hänet noutaa ja tuskat
lientää.

"Oi armahin, sinut kerran omaan!"

Näin nuorukainen nyt huudahtaa
ja povellensa hän immen painaa:
"Ei meitä kukaan voi erottaa!
Sun lempes kautta voin kaikki voittaa,
viel' onnen tähtikin meille koittaa!"

"Oi sinne rietkäämme, missä myrskyt ei lemmen kukkia

riistää voi, miss' armahainen on iki-kesä ja linnut kauniisti
lauleloi! Ja Onnetar se nyt meitä johtaa, siks kunne ruusujen
nurmi hohtaa."
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

textbookfull.com