Assessment And Intervention Advances In Learning And Behavioral Disabilities Thomas E Scruggs

Assessment And Intervention Advances In Learning
And Behavioral Disabilities Thomas E Scruggs
download
https://ebookbell.com/product/assessment-and-intervention-
advances-in-learning-and-behavioral-disabilities-thomas-e-
scruggs-2474472
Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Children And Youth With Autism Spectrum Disorder Asd Recent Advances
And Innovations In Assessment Education And Intervention James K
Luiselli Ed
https://ebookbell.com/product/children-and-youth-with-autism-spectrum-
disorder-asd-recent-advances-and-innovations-in-assessment-education-
and-intervention-james-k-luiselli-ed-4964210
Assessment And Intervention With Mothers And Partners Following Child
Sexual Abuse Empowering To Protect Jenny Still
https://ebookbell.com/product/assessment-and-intervention-with-
mothers-and-partners-following-child-sexual-abuse-empowering-to-
protect-jenny-still-47215828
Assessment And Intervention For English Language Learners Translating
Research Into Practice 1st Edition Susan Unruh
https://ebookbell.com/product/assessment-and-intervention-for-english-
language-learners-translating-research-into-practice-1st-edition-
susan-unruh-5882644
Assessment And Intervention For Executive Function Difficulties
Schoolbased Practice In Action Series Papcdr George Mccloskey
https://ebookbell.com/product/assessment-and-intervention-for-
executive-function-difficulties-schoolbased-practice-in-action-series-
papcdr-george-mccloskey-1858548

Mathematics Assessment And Intervention In A Plc At Work 2nd Edition
Sarah Schuhl
https://ebookbell.com/product/mathematics-assessment-and-intervention-
in-a-plc-at-work-2nd-edition-sarah-schuhl-53058044
Violence Assessment And Intervention The Practitioners Handbook 3rd
Edition James S Cawood
https://ebookbell.com/product/violence-assessment-and-intervention-
the-practitioners-handbook-3rd-edition-james-s-cawood-36430804
Violence Assessment And Intervention The Practitioners Handbook
Michael H Corcoran James S Cawood
https://ebookbell.com/product/violence-assessment-and-intervention-
the-practitioners-handbook-michael-h-corcoran-james-s-cawood-4109550
Clinical Assessment And Intervention For Autism Spectrum Disorders 1st
Edition Johnny L Matson
https://ebookbell.com/product/clinical-assessment-and-intervention-
for-autism-spectrum-disorders-1st-edition-johnny-l-matson-4646804
Literacy Assessment And Intervention For Classroom Teachers Paperback
Beverly A Devries
https://ebookbell.com/product/literacy-assessment-and-intervention-
for-classroom-teachers-paperback-beverly-a-devries-10130954

ASSESSMENT AND INTERVENTION

ADVANCES IN LEARNING AND
BEHAVIORAL DISABILITIES
Series Editors: Thomas E. Scruggs and
Margo A. Mastropieri
Recent Volumes:
Volume 12: Edited by Thomas E. Scruggs and Margo A. Mastropieri
Volume 13: Edited by Thomas E. Scruggs and Margo A. Mastropieri
Volume 14: Educational Interventions – Edited by
Thomas E. Scruggs and Margo A. Mastropieri
Volume 15: Technological Applications – Edited by
Thomas E. Scruggs and Margo A. Mastropieri
Volume 16: Identiﬁcation and Assessment – Edited by
Thomas E. Scruggs and Margo A. Mastropieri
Volume 17: Research in Secondary Schools – Edited by
Thomas E. Scruggs and Margo A. Mastropieri
Volume 18: Cognition and Learning in Diverse Settings – Edited
by Thomas E. Scruggs and Margo A. Mastropieri
Volume 19: Applications of Research Methodology – Edited
by Thomas E. Scruggs and Margo A. Mastropieri
Volume 20: International Perspectives – Edited by Thomas E. Scruggs
and Margo A. Mastropieri
Volume 21: Personnel Preparation – Edited by Thomas E. Scruggs
and Margo A. Mastropieri
Volume 22: Policy and Practice – Edited by Thomas E. Scruggs and
Margo A. Mastropieri
Volume 23: Literacy and Learning – Edited by Thomas E. Scruggs
and Margo A. Mastropieri

ADVANCES IN LEARNING AND BEHAVIORAL
DISABILITIES VOLUME 24
ASSESSMENT AND
INTERVENTION
EDITED BY
THOMAS E. SCRUGGS
George Mason University, Fairfax, VA, USA
MARGO A. MASTROPIERI
George Mason University, Fairfax, VA, USA
United Kingdom – North America – Japan
India – Malaysia – China

Emerald Group Publishing Limited
Howard House, Wagon Lane, Bingley BD16 1WA, UK
First edition 2011
Copyrightr2011 Emerald Group Publishing Limited
Reprints and permission service
Contact: [email protected]
No part of this book may be reproduced, stored in a retrieval system, transmitted in any
form or by any means electronic, mechanical, photocopying, recording or otherwise
without either the prior written permission of the publisher or a licence permitting
restricted copying issued in the UK by The Copyright Licensing Agency and in the USA
by The Copyright Clearance Center. No responsibility is accepted for the accuracy of
information contained in the text, illustrations or advertisements. The opinions
expressed in these chapters are not necessarily those of the Editor or the publisher.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
ISBN: 978-0-85724-829-9
ISSN: 0735-004X (Series)
Emerald Group Publishing
Limited, Howard House,
Environmental Management
System has been certified by
ISOQAR to ISO 14001:2004
standards
Awarded in recognition of Emerald’s production department’s adherence to quality systems and processes when preparing scholarly journals for print

CONTENTS
LIST OF CONTRIBUTORS vii
PREDICTIVE VALIDITY OF DYNAMIC TESTING
AND WORKING MEMORY AS IT RELATES TO
READING GROWTH IN CHILDREN WITH
READING DISABILITIES
H. Lee Swanson and Michael Orosco 1
APPLICATIONS OF CURRICULUM-BASED
MEASURES IN MAKING DECISIONS WITH
MULTIPLE REFERENCE POINTS
Gerald Tindal and Joseph F. T. Nese 31
EVOLVING STANDARDS OF DIAGNOSTIC
ACCURACY IN PREDICTING AND AVOIDING
ACADEMIC FAILURE
Amanda M. VanDerHeyden 59
HOW TO USE PSYCHOLOGICAL TESTS FOR
FUNCTIONAL DIAGNOSIS: THE CASE OF
ASSESSMENT OF LEARNING DISABILITIES
Giulia Balboni and Roberto Cubelli 79
A COMPARISON OF OBSERVATIONAL TECHNIQUES
FOR ASSESSING STUDENTS’ SOCIAL BEHAVIOR
Jugnu Agrawal, Dannette Allen-Bronaugh and Margo A.
Mastropieri
93
CURRICULUM-BASED MEASUREMENT FOR
BEGINNING WRITERS: RECENT DEVELOPMENTS
AND FUTURE DIRECTIONS
Kristen L. McMaster, Kristen D. Ritchey and Erica
Lembke
111
v

ISSUES IN ASSESSMENT FOR INTERVENTION IN
IMPLEMENTATION OF RESPONSIVENESS TO
INTERVENTION MODELS
Rollanda E. O’Connor and Victoria Sanchez 149
COMORBIDITY BETWEEN ATTENTION DEFICIT
HYPERACTIVITY DISORDER AND READING
DISABILITIES: IMPLICATIONS FOR ASSESSMENT
AND TREATMENT
Ana Miranda, Marı´a Jesu´s Presentacio´n, Rebeca
Siegenthaler, Carla Colomer and Vicente Pinto
171
ASSESSMENT AND INTERVENTION IN SELF-
DETERMINATION
Michael L. Wehmeyer 213
SUCCESS AND FAILURE WITH TIER-2 SRSD FOR
TIMED-WRITING TESTS AMONG SECOND-
THROUGH FIFTH-GRADE STUDENTS WITH
WRITING AND BEHAVIORAL DIFFICULTIES:
IMPLICATIONS FOR EVIDENCE-BASED PRACTICE
Karin Sandmel, Kristen D. Wilson, Karen R. Harris,
Kathleen Lynne Lane, Steve Graham, Wendy P. Oakes,
Sharlene A. Kiuhara and Trish D. Steinbrecher
251
DEVELOPING WRITING FLUENCY FOR
ADOLESCENTS WITH DISABILITIES
Linda H. Mason and Richard M. Kubina 295
ENGAGING YOUTH WITH DISABILITIES IN
SCHOOL: BUILDING AND SUSTAINING POSITIVE
RELATIONSHIPS
Elizabeth Talbott and Lisa S. Cushing 321
IMPROVING INTERVENTION EFFECTIVENESS
WITH UNIVERSITY–PUBLIC SCHOOL COHORT
PARTNERSHIPS
Margo A. Mastropieri, Thomas E. Scruggs, Nicole
Conners, Mary Kealy, Nancy Morrison, Tina Diamond
and Terry Werner
341
CONTENTSvi

LIST OF CONTRIBUTORS
Jugnu Agrawal College of Education and Human
Development, George Mason University,
Fairfax, VA, USA
Dannette
Allen-Bronaugh
College of Education and Human
Development, George Mason University,
Fairfax, VA, USA
Giulia Balboni Faculty of Psychology, University of Valle
d’Aosta, Aosta, Italy
Carla Colomer Department of Developmental and
Educational Psychology, University of
Valencia, Valencia, Spain
Nicole Conners Fairfax County Public Schools, Fairfax,
VA, USA
Roberto Cubelli Department of Cognitive Sciences and
Education, Center for Mind/Brain
Sciences, University of Trento, Trento,
Italy
Lisa S. Cushing Department of Special Education,
University of Illinois, Chicago, IL, USA
Tina Diamond College of Education and Human
Development, George Mason University,
Fairfax, VA, USA
Steve Graham Department of Special Education,
Vanderbilt University, Nashville, TN, USA
Karen R. Harris Department of Special Education,
Vanderbilt University, Nashville, TN, USA
Mary Kealy Loudoun County Public Schools,
Loudoun, VA, USA
vii

Sharlene A. KiuharaDepartment of Special Education,
Vanderbilt University, Nashville, TN, USA
Richard M. Kubina Department of Educational and School
Psychology and Special Education, the
Pennsylvania State University, University
Park, PA, USA
Kathleen Lynne LaneDepartment of Special Education,
Vanderbilt University, Nashville, TN, USA
Erica Lembke University of Missouri, Columbia, MO,
USA
Linda H. Mason Department of Educational and School
Psychology and Special Education, the
Pennsylvania State University,
University Park, PA, USA
Margo A. MastropieriCollege of Education and Human
Development, George Mason University,
Fairfax, VA, USA
Kristen L. McMasterUniversity of Minnesota, Minneapolis,
MN, USA
Ana Miranda Department of Developmental and
Educational Psychology, University
of Valencia, Spain
Nancy Morrison Fairfax County Public Schools, Fairfax,
VA, USA
Joseph F. T. Nese Department of Special Education,
University of Oregon, Eugene, OR, USA
Rollanda E. O’ConnorGraduate School of Education, University
of California at Riverside, CA, USA
Wendy P. Oakes Department of Special Education,
Vanderbilt University, Nashville, TN, USA
Michael Orosco Graduate School of Education, University
of California at Riverside, Riverside, CA,
USA
viii LIST OF CONTRIBUTORS

Vicente Pinto Department of Developmental,
Educational and Social Psychology, and
Methodology, Jaume I University,
Castello´n, Spain
Marı´a Jesu´s
Presentacio´n
Department of Developmental,
Educational and Social Psychology, and
Methodology, Jaume I University,
Castello´n, Spain
Kristen D. RitcheySchool of Education, University of
Delaware, Newark, DE, USA
Victoria Sanchez Graduate School of Education, University
of California at Riverside, Riverside, CA,
USA
Karin Sandmel School of Education, Johns Hopkins
University, Maryland, Baltimore, MD,
USA
Thomas E. Scruggs College of Education and Human
Development, George Mason University,
Virginia, Fairfax, VA, USA
Rebeca SiegenthalerDepartment of Developmental,
Educational and Social Psychology, and
Methodology, Jaume I University,
Castello´n, Spain
Trish D. SteinbrecherDepartment of Special Education,
Vanderbilt University, Nashville, TN, USA
H. Lee Swanson Graduate School of Education, University
of California, Riverside, CA, USA
Elizabeth Talbott Department of Special Education,
University of IL, Chicago, USA
Gerald Tindal Department of Special Education,
University of Oregon, Eugene, OR, USA
Amanda M.
VanDerHeyden
Education Research & Consulting,
Fairhope, AL, USA
List of Contributors ix

Michael L. WehmeyerDepartment of Special Education,
University of Kansas, Lawrence, KS, USA
Terry Werner College of Education and Human
Development, George Mason University,
Fairfax, VA, USA
Kristen D. Wilson Department of Special Education,
Vanderbilt University, Nashville, TN, USA
x LIST OF CONTRIBUTORS

PREDICTIVE VALIDITY OF
DYNAMIC TESTING AND
WORKING MEMORY AS IT
RELATES TO READING GROWTH
IN CHILDREN WITH READING
DISABILITIES
H. Lee Swanson and Michael Orosco
ABSTRACT
The purpose of this chapter is to review our ﬁndings related to the
question ‘‘Do outcomes related to dynamic assessment on a cognitive
measure predict reading growth?’’ Our discussion related to the predictive
validity of such procedures focused on outcomes related to a battery of
memory and reading measures administered over a three-year period to 78
children (11.6 years) with and without reading disabilities (RD).
Working memory (WM) tasks were presented under initial, gain, and
maintenance testing conditions. The preliminary results suggested that
maintenance testing conditions were signiﬁcant moderators of compre-
hension and vocabulary growth, whereas probe scores and gain testing
Assessment and Intervention
Advances in Learning and Behavioral Disabilities, Volume 24, 1–30
Copyrightr2011 by Emerald Group Publishing Limited
All rights of reproduction in any form reserved
ISSN: 0735-004X/doi:10.1108/S0735-004X(2011)0000024003
1

conditions were signiﬁcant moderators of nonword ﬂuency growth.
Overall, the results suggested that the dynamic assessment of WM added
signiﬁcant variance in predicting later reading performance.
Children with normal intelligence but who suffer from reading disabilities
(RD) experience considerable difﬁculty on working memory (WM) tasks
(e.g., see De Beni, Palladino, Pazzaglia,&Cornoldi,1998;deJong,1998;
Gathercole, Alloway, Willis, & Adams, 2006; Siegel & Ryan, 1989;
Swanson & Siegel, 2001; Willcutt, Pennington, Olson, Chhabildas, &
Hulslander, 2005). Although WM is integrally related to a number of
academic behaviors (e.g., Gathercole & Pickering, 2000a, 2000b; Pickering,
2006;Swanson,2008;Swanson,Jerman, & Zheng, 2008), relatively few
studies have been undertaken to systematically explore whether WM can be
improved upon and whether such improvements predict later academic
performance. The purpose of this chapter is to review our ﬁndings on
whether procedures that attempt to improve WM performance in children
with RD predict reading growth.
The procedures we use to assess improvements in WM performance come
under the rubric of dynamic assessment (DA). Dynamic assessment is one of
clearest links between assessment and instruction (e.g., Caffrey, Fuchs, &
Fuchs, 2008; Sternberg & Grigorenko, 2002; Swanson & Lussier, 2001)
because it incorporates a child’s response to instructional feedback. Although
DA is a term used to characterize a number of approaches, its distinctive
feature is to determine the learners’ potential for change when given assistance
(e.g., Embretson, 1992; see Grigorenko & Sternberg, 1998 for a review), and
to provide a prospective measure of performance change independent of
assistance (e.g., Brown & Ferrara, 1999). Unlike traditional testing
procedures, score changes due to DA methods are not viewed as threatening
task validity. In fact, some authors argue that construct validity will increase
(e.g., Carlson & Wiedl, 1979; however see Caffrey et al., 2008; Swanson &
Lussier, 2001, for a critical review). For example, DA procedures have been
found to contribute unique variance to concurrent criterion measures of
achievement even when the inﬂuence of traditional assessment procedures was
partialed out in the analysis (e.g., Swanson, 1999).
Although the literature is clear that impairments in WM in children with
RD are related to reading performance, whether DA procedures add
additional information to the prediction of growth in reading in children
with RD beyond traditional testing procedures has not been tested. Perhaps
H. LEE SWANSON AND MICHAEL OROSCO2

the closest study on this issue is a cross-sectional study by Swanson (2003)
which determined whether improvements in WM differentiated children
with and without RD. The study compared RD and skilled readers
(N¼226) across four (7, 10, 13, 20) age groups on WM tasks under various
WM conditions: initial (no probes or cues), gain (cues that bring
performance to an asymptotic level), and maintenance conditions (asymp-
totic conditions without cues). The general ﬁndings were that skilled readers
performed better than children with RD in all processing conditions and
that concurrent reading comprehension performance was best predicted by
the maintenance testing than the other testing conditions. Further, the
magnitude of the difference (effect size) between RD and skilled readers
increased on gain and maintenance testing conditions when compared with
the initial conditions, suggesting that performance differences between
ability groups were enhanced by using DA procedures. However, there was
no follow-up testing of these children, raising the questions as to whether the
measures were predictive of later reading performance.
The work we describe in this chapter outlines our attempts to improve
WM performance, via adapting procedures used in DA, and then predict
later reading performance. The question of interest was whether DA
procedures provide additional information (contributed unique variance) in
predicting children’s reading performance relative to traditional testing
procedures. To answer this question we assessed WM performance under
three conditions across three years. These conditions included: (1) presenta-
tion of WM tasks without cues to assess initial performance (initial
condition), (2) presentation of graduated cues to help participants access
forgotten information from the initial condition and to continue the use of
cues until span scores could no longer be improved upon (referred to as the
gain or asymptotic testing condition), and (3) presentation of the highest
span level achieved for the gain condition after a brief interlude, but without
the support of cues (referred to as the maintenance condition). The
administration of WM tasks under initial testing conditions were assumed
to tap individual differences in idiosyncratic processing, as well as individual
differences in accessing newly presented items in storage. To assess individual
differences in item accessibility, probes (probes and cues are used
interchangeably) were presented to help children reinstate the memory trace
and/or retrieve forgotten items. This condition, referred to as the gain testing
condition, allowed participants to use as many cues as necessary to access
previously forgotten information. The probes were ordered in terms of
explicitness, with the general prompts given ﬁrst and more explicit prompts
later. To assess constraints in activating previously retrieved information, the
Predictive Validity of Dynamic Testing 3

testing procedures reinstated the highest level achieved successfully under
gain conditions, but without cues. This condition, referred to as the
maintenance condition, presented the same items retrieved in the gain
conditions. The presentation of items in the maintenance testing condition
was calibrated to the individual’s highest span level and allowed for the
assessment of individual processing differences beyond the simple learning of
items.
Prior to reviewing the outcomes, we brieﬂy describe our sample,
procedures, and analysis. In this study, subgroups of children with RD
were tested across three testing waves approximately one year apart. We
predicted that DA procedures would differentiate WM performance among
subgroups of children with RD. For example, a previous study by Swanson
and Howard (2005) found that children with RD-only and children with
general achievement problems could be separated in terms of performance
using DA procedures. They showed that skilled readers and children
considered poor readers (children with low IQ and reading scores) were
more likely to change and maintain their WM score from initial testing
conditions than children with RD-only or children with a combination of
math/reading disabilities. Overall, the results suggested that WM deﬁcits
were less changeable for children with RD when compared to low IQ
readers and skilled readers. To this end, our most recent study included 78
children selected from a school in Southern California. The mean age of the
sample in the ﬁrst testing wave was 11.62 years (SD¼2.28) and included 27
girls and 51 boys. Sixty-six children were Anglo-American, eight African-
American, and four were of Hispanic heritage. All children were from
middle-upper to upper class homes. The criterion for the average reading
group was a score above the 45th percentile in reading word recognition, as
well as an IQ score above 90.
Classiﬁcation of subgroups of children with RD was based upon standard
scores from the word-reading and arithmetic subtests of the Wide Range
Achievement Test 3 (WRAT3, Wilkinson, 1993). Operational criteria for
children deﬁned as the RD-only group were based upon: (a) IQ scores (in
this case scores on the Raven Colored Progressive Matrices Test, Raven,
1976) above 90, (b) reading recognition scores at or below the 25th
percentile, and (c) math calculation scores above the 25th percentile. These
cut-off score criteria for children with RD-only matched the operational
deﬁnition of RD outlined by Siegel and Ryan (1989). Children with average
ﬂuid intelligence (scores on the Raven Progressive Colored Matrices Test)
and verbal IQs but whose performance on the WRAT3 in both word
H. LEE SWANSON AND MICHAEL OROSCO4

recognition and calculation were below the 25th percentile were classiﬁed as
RD children with comorbid deﬁcits (RDþMD). Children whose ﬂuid
intelligence (Raven Progressive Colored Matrices Test) were in the average
range but tested low (o25th percentile) in both arithmetic and word
recognition and obtained low verbal IQs (o85, subtests on the WISCIII that
included information and vocabulary), were deﬁned as low verbal IQ
readers. All children were monolingual in English.
For comparison purposes, participants were divided into four ability
groups: children with RD-only (low word recognition/average arithmetic,
n¼13), children with both reading and calculation deﬁcits (comorbid
group,n¼15), low verbal IQ readers (low verbal IQ, low word recognition,
and low arithmetic,n¼27), and skilled readers (high word recognition, high
arithmetic,n¼23). No signiﬁcant differences emerged among subgroups
related to gender, ethnicity, or chronological age.
Because ability groups were categorized on measures of word recognition
and math skills, it was important to follow growth on measures related to
other forms of reading achievement independent of the classiﬁcation
measures. Three criterion measures were selected: reading comprehension,
nonword reading ﬂuency, and receptive vocabulary. Because real sight
words were used in the classiﬁcation of reading groups, we selected pseudo-
word reading speed because of its strong association with phonological
skills. Further, because expressive vocabulary was used to establish verbal
IQ, we selected the receptive language test (PPVT-III) as a criterion
measure. Reading comprehension was assessed by the Passage Comprehen-
sion subtest from theWoodcock Reading Mastery Test Revised(WRMT-R;
Woodcock, 1998). Nonword reading efﬁciency was assessed by the
Phonemic Decoding Efﬁciency subtest from theTest of Word Reading
Efﬁciency(TOWRE; Wagner & Torgesen, 1999). Receptive vocabulary was
assessed byThe Peabody Picture Vocabulary Test-Third Edition(PPVT-III;
Dunn, Dunn, Dunn, & Dunn 1997).
WORKING MEMORY MEASURES AND TESTING
CONDITIONS
Four WM subtests from a battery of 11 of the S-Cognitive Processing Test
were selected because they represent verbal and visual-spatial processing
(Swanson, 1995a, 1995b).
Predictive Validity of Dynamic Testing 5

Verbal Working Memory
Rhyming
The purpose of this task was to assess the child’s recall of acoustically
similar words. The child listened to sets of words that rhyme. Each
successive word in the set was presented every 2 seconds. There were 9 word
sets that ranged from 2 to 14 monosyllabic words. The dependent measure
was the number of sets recalled. Before recalling the words, the child was
asked whether a particular word was included in the set. For example, the
child was presented the words ‘‘lip-slip-clip’’ and then asked if ‘‘ship or lip’’
was presented in the word set. They were then asked to recall the previously
presented words (lip-slip-clip) in order. The dependent measure was the
number of sets recalled correctly (range of 0–9).
If the child omitted, inserted, or incorrectly ordered the words, a series of
probe responses were presented. Probe responses continued until the child
could no longer provide the correct response. For example, consider the
sample item ‘‘car-star-bar-far’’ (Item #3) and the process question ‘‘Which
word did I say – jar or star?’’ Consider the Probe Sequence:
1. The last word in the sequence was ‘‘far,’’ now can you tell me all the
words in order?
2. The ﬁrst word in the sequence was ‘‘car,’’ now can you tell me all the
words in order?
3. The middle words in the sequence are ‘‘star’’ and ‘‘bar,’’ now can you tell
me all the words in order?
4. All the words in order are ‘‘car-star-bar-far,’’ now can you tell me all the
words?
For each set of items not recalled in the correct order or for items left out
or substituted, the experimenter provided a series of hints based on the error
that was closest to Probe 1. That is, probes went from the least obvious hint
(Probe 1) to the next explicit hint that facilitated recall of the answer. Once
the appropriate hint had been identiﬁed, based on the location of the error,
probes were presented in order until the correct sequence was given. For
example, suppose the child for Item #3 responded car-bar-far. The child
obviously left out a word in the middle so the experimenter would provide a
hint related to the middle words (Probe 3 in this case). If Probe 3 did not
provide the correct response, the experimenter would then move to Probe 4.
In contrast, if a child responded initially by saying only car, the sequence
began with Probe 1 and proceeded through all probes until the correct
response was given. If a correct response did not occur after probing, the
H. LEE SWANSON AND MICHAEL OROSCO6

task was discontinued and the next task was administered. If a correct
response did occur, the next set of items of increased difﬁculty was
presented. The format of the probe procedures above was used for the other
subtests.
Digit/Sentence Task
The purpose of this task was to assess the participant’s ability to remember
numerical information embedded in a short sentence. The administration of
items and probes followed the same format as the rhyming task. Prior to
stimulus presentation, the participant was shown a ﬁgure (see Swanson,
1993, Fig. 1) depicting four strategies for recalling numerical information.
These strategies were pictorial representations of rehearsal, chunking,
associating, and elaborating of information. The general instructions for
introducing the strategies were as follows:
I’m going to read you some sentences that have information I want you to remember. All
the sentences have to do with remembering an address, but I would like you to pay
attention to all the information in the sentence because I will ask you a question about
the sentence. After I present this information, and before you recall it, I will ask you to
choose a strategy (for children under ten – the phrase, ‘‘A way of remembering the
information’’ was used) that you think will best help you remember.
The experimenter then showed four pictures each depicting a person
thinking about using one of the four strategies (see Swanson, 1993). As the
experimenter explained each strategy, they pointed to the picture that
matched the description. The experimenter stated that
Some of the ways that may help you remember are: (1) saying the numbers over
to yourself. For example, if I say ‘‘2-4-6-3 Bader Street,’’ you would say to yourself ‘‘2-
4-6-3’’ over and over again, or (2) you might say some numbers together in pairs. For
example, if I say the numbers ‘‘2-4-6-3 Bader Street,’’ you would say ‘‘24 and 63,’’ or
(3) you may just want to remember that the numbers go with a particular street and
location. For example, if I say ‘‘2-4-6-3 Bader Street’’ you would remember that 2-4-6-3
and Bader Street goes together, or (4) you might think of other things that go with
the numbers. For example, if I say ‘‘2-4-6-3’’ you might think 2-4-6-3 I have to go climb
a tree.
These four pictorial representations of strategies generally reﬂect
rehearsal, chunking, associating, and elaborating of information, respec-
tively. After all strategies had been explained, participants were then
presented item sets that included numbers in a sentence context. They were
then told that they must recall the numbers in the sentence in order shortly
after they select from (point to) a pictorial array representing the strategy that
best approximates how he or she will attempt to remember the information.
Predictive Validity of Dynamic Testing 7

No further information about the strategies shown in the picture was provided
to the participant. Participants were allowed 10 seconds in which to make a
decision. The range of recall difﬁculty was 3 digits to 14 digits, and the
dependent measure was the highest number of sets correctly recalled (range of
difﬁculty 0–9).
Thus, the sequence of the steps for administration after the introduction
was as follows: (1) The participant was orally read a sentence (the numbers
in the sentence were presented at the rate of approximately 1 every 2
seconds), (2) the participant was asked a process question which required
them to give the name of the street referred to in the target sentence, (3) the
participant was asked to select one of the four strategies that were
represented pictorially that were most like the one they would use to
remember the order of the street numbers, (4) the participant was asked to
recall the numbers of the address in the order in which they were originally
presented, and (5) if an error in recall occurred, the probe questions were
implemented.
Probing procedures followed the same format as the rhyming task: hints
were provided sequentially based on the type of error, and ranged from least
obvious hint (Probe 1) to the next explicit hint that facilitated recall of the
answer. If probing did not elicit a correct response, the task was discontinued
and the next task was administered. If a correct response did occur, the next
set of items of increased difﬁculty was presented.
Visual-Spatial Working Memory
Visual Matrix
The purpose of this task was to assess the participant’s ability to remember
visual sequences within a matrix. The participant was presented a series of
dots in a matrix and was allowed 5 seconds to study the matrix. The matrix
was removed and the participant was asked a discrimination question, i.e.,
‘‘Are there any dots in the ﬁrst column?’’ To ensure the understanding of
columns, the experimenter pointed to the ﬁrst column on a blank matrix (a
grid with no dots). After answering the discrimination question, the
participant was asked to draw the dots in the correct boxes on the blank
matrix. The task difﬁculty ranged from a matrix of 4 squares and 2 dots to
a matrix of 45 squares and 12 dots. The dependent measure was the number
of matrices recalled correctly (range of 0–11). If an error occurred, probe
questions were started.
H. LEE SWANSON AND MICHAEL OROSCO8

Probe procedures followed the same format as the verbal subtests, except
that the matrices were represented as columns that reﬂect recency, primacy,
and middle positions. For example, when probing for errors that occurred in
the recency position, the experimenter drew dots on a blank matrix in the
appropriate last columns and said, ‘‘Now can you show me where the rest of
the dots go?’’ During probing, the matrices response form in which initial
scores were established was turned over. For each hint provided to the
participant, anewmatrix was presented, the experimenter provided the hints
(demonstrates by ﬁlling in the appropriate dots) on this matrix and then the
participant was asked to ﬁll in the correct remaining dots. If some hints do
not have dots in which to demonstrate, the participant was told so. The
instructions related to probing were as follows:
‘‘You missed placing some dots in the right boxes. I think you can do it correctly if I
provide you with some hints.’’
Probe 1. On the blank matrix, the experimenter correctly draws in the last column(s) of
dots. Then the experimenter says, ‘‘Now can you draw where the rest of the dots go?’’
Probe 2. On a blank matrix, the experimenter draws in dots for the ﬁrst column. Then
the experimenter says, ‘‘Now can you draw where the rest of the dots go?’’
Probe 3. On a blank matrix, the experimenter draws in the dots for the middle (between
the ﬁrst and last column) and says, ‘‘Now can you draw where the rest of the dots go?’’
Probe 4. The stimulus card is shown for 2 seconds. The model matrix is removed and the
participant is asked to ﬁll in all the dots on a blank matrix.
Mapping and Directions
The purpose of this task was to determine whether the participants could
remember a sequence of directions on a map that is void of labels (see
Swanson, 1995a for detailed instructions). The experimenter presented the
participant with a street map with lines connected to a number of dots
that illustrated the direction a car would go to get out of the city (see
Swanson, 1993). The dots represented stoplights, and the lines represented
the direction the car should go. The participant was given 5 seconds to study
the map. After the map was removed, the participant was then asked a
process question and to point to the strategy (picture) they thought they
would use to remember the street directions. Finally, they were asked to
draw on a blank map the street directions (lines) and stop lights (dots). The
process question was ‘‘Were there any dots in the ﬁrst street (column)?’’
Using the same pictorial format as the Digit/Sentence Task, strategies were
pictorial representations of elemental, global, sectional, or backward
processing of patterns. The range of difﬁculty included dots that ranged
Predictive Validity of Dynamic Testing 9

in number from 4 to 19. The dependent measure was the number of maps
drawn correctly (range of 0–9).
If an error was made, a series of probes was presented following the same
format as the visual-matrix task procedures. Probing continued until the
participant could no longer provide the correct response. The task was
subsequently discontinued and the next task was administered.
Validity and Reliability of WM Measures
The psychometric qualities of the measures are reviewed in Swanson (1995a,
1995b, 1996). Cronbach alpha measures on the subtest vary from .78 to .96.
The reliabilities for the WM measures have been sustained when translated
into Spanish (e.g., Swanson, Saez, Gerber, & Leafstedt, 2004) and Japanese
(e.g., Hwang et al., 2006). Correlations with each WM subtest vary on
intelligence, achievement, and Listening Span (Daneman & Carpenter,
1980) measures from .40 to .80 (e.g., see Swanson, 1992, Tables 4 and 7,
Swanson, 1996, see Table 4).
Procedure
Individual testing of children was performed by ﬁve doctoral students over
the course of three years. All items for the WM measures for the initial
testing condition were administered until (a) a process question was missed
or (b) an error in retrieval occurred. If an error in retrieval occurred
(a participant omitted, inserted, or incorrectly ordered the numbers,
dots, words, related to the appropriate task), probes were administered.
The only stipulation for instituting the probing condition was that the
process question be answered correctly. Probes were administered
based on the type of error made (i.e., whether the error was related to
recency, primacy, or middle items), and probing procedures continued
until all targeted items could not be recalled correctly. After all four
subtests from the S-CPT were administered under initial and gain testing
conditions, participants were re-administered the items for the highest
successful set (highest set of items established under gain conditions) for
each task.
In order to compare ability groups across verbal and visual-spatial WM
measures, mean compositez-scores for verbal (rhymingþauditory sequence)
H. LEE SWANSON AND MICHAEL OROSCO10

and visual-spatial (matrixþmapping/direction) WM performance from
the initial testing conditions were computed.Z-scores for the gain and
maintenance testing conditions were based on the means and SDs from the
initial condition. The compositez-scores for wave (year) 1, 2, and 3 as a
function of ability group and testing condition are shown in the appendix.
Thez-scores for wave 2 and 3 were based on the mean and standard
deviations of span scores in the initial testing condition at wave 1.
As a preliminary analysis of the ﬁndings we calculated effect sizes for the
gain [(mean gainmean initial)/SD
gain] and maintenance testing conditions
[(mean gainmean initial)/SD
gain] relative to the initial testing conditions.
The standard deviations from the DA conditions (in this case gain or
maintenance scores) were selected because pretest scores (in this case initial
testing scores) are generally inﬂated (see Swanson & Lussier, 2001, for
discussion).
1
Further adjustments were made in the calculations because of
the high correlations among tasks with repeated measures (see Rosenthal,
1994, p. 241, for formula).
Table 1 shows the magnitude of the effect sizes (ES) as a function of gain
or maintenance testing conditions. Also shown are the ES corrected for the
correlation between the initial and DA conditions. Also shown is the ratio of
inﬂation in standard deviations by dividing the standard deviation for the
DA measures (gain or maintenance) by the standard deviation for initial
testing measure. A previous synthesis of DA measures (Swanson & Lussier,
2001) suggested that ratios greater than 1.25 reﬂect practice effects rather
than responsiveness to DA techniques. That is, initial testing conditions
rather than DA served as the learning event.
Using Cohen’s (1988) criterion for large (.80) and moderate (.50) effect
sizes, Table 1 shows that all ability groups yielded high effect sizes for the
gain testing condition and moderate effect sizes for the maintenance testing
conditions. Using 1.25 ratio as a critical value, practice effects from the
initial test condition were most notable for the RDþMD group (4 of 6
values were greater than 1.25). Another important ﬁnding was related to the
reduction in performance from the gain to maintenance testing condition.
The largest average drop in performance from the gain to the maintenance
testing condition (ES for gain condition minus ES for maintenance
condition) occurred for the RD-only group (.86 ES difference), whereas
the remaining groups yielded a reduction in ES that were in the same range
[RDþMD (.54 ES difference), low verbal IQ reader (.56 difference), and
skilled readers (.48 difference). Thus, the RD-only group showed the largest
drop in performance (3/4 of SD) from the gain to the maintenance testing
condition when compared to the other groups.
Predictive Validity of Dynamic Testing 11

Summary
A descriptive analysis of ES between gain and maintenance conditions
suggested that the RD-only group was less likely to maintain performance
when compared to the other groups.
Table 1.Effect Sizes (ES), Retesting Sensitivity, and Corrected Effect
Sizes by Reading Group.
Gain-
ES
Maintenance-
ES
Gain
Ratio
a
Maintenance
Ratio
a
Corrected
Gain ES
Corrected
Maintenance
ES
RD-only
Wave
1
1.64 1.01 1.02 0.90 1.03 0.63
Wave
2
1.88 0.48 0.86 1.35 1.19 0.3
Wave
3
2.69 0.64 0.92 1.43 1.7 0.4
RDþMD
Wave
1
1.2 0.45 1.67 1.46 0.76 0.28
Wave
2
1.41 0.55 1.98 2.10 0.89 0.35
Wave
3
2.10 1.18 0.92 0.94 1.33 0.74
Low VIQ
Wave
1
1.21 0.51 1.32 1.09 0.77 0.32
Wave
2
1.76 0.73 1.33 1.23 1.12 0.46
Wave
3
1.69 0.72 1.12 1.24 1.07 0.45
Skilled
Wave
1
1.71 0.69 0.89 0.96 1.08 0.43
Wave
2
1.27 0.7 1.74 1.08 0.81 0.44
Wave
3
1.67 1.02 0.99 1.11 1.06 0.64
a
Ratio of inﬂation in standard deviations by dividing the standard deviation for the DA
measures (gain or maintenance) by the standard deviation for initial testing measure.
H. LEE SWANSON AND MICHAEL OROSCO12

Repeated Measures
Prior to determining whether the various testing conditions of WM
contributed to predictions of growth on measures of reading comprehension,
nonword ﬂuency, and receptive language, we used mixed modeling with
repeated measures across testing waves to determine if signiﬁcant differences
in growth emerged between ability groups as a function of testing conditions.
The meanz-scores for each reading group as a function of domain (verbal vs.
visual) across the testing waves are shown in the appendix. Because span scores
were greater in the verbal than visual-spatial domain under DA conditions at
wave 1, only verbal WM span scores were considered in the subsequent
analysis. In addition, because initial testing may have served as a learning
event, initial testing scores were partialed in the analysis. Further, because
chronological age at initial testing in wave 1 may interact with the amount of
change across testing waves, chronological age at wave 1 was a covariate in the
analysis. The results indicated that skilled readers outperformed the RD
subgroups across all testing waves and DA conditions. The DA conditions did
not differentiate between subgroups of less skilled readers. Because the RD
subgroups could not be differentiated across the testing waves, we combined
the performance RD subgroups in the subsequent analyses.
Correlations
An important question to address was whether DA procedures in wave 1
were related to achievement measures in wave 3. As shown in Table 2, the
majority of verbal WM measures in wave 1 were signiﬁcantly related to the
criterion measures in wave 3. However, it is unclear from the correlation
analysis which of the testing conditions in wave 1 uniquely moderated
growth on the criterion measures. To address this issue, latent growth
modeling was conducted to determine which of the testing conditions in
wave 1 moderated growth on the criterion measures. Verbal WM measures
were used since the majority of signiﬁcant correlations shown in Table 2
were related to the verbal WM measures.
Growth Modeling
To examine growth in reading, we used a multilevel framework (Singer,
2002; Singer & Willet, 2003). We applied individual growth modeling to the
Predictive Validity of Dynamic Testing 13

study of intra-individual change in reading as function of WM testing
conditions over the three-year period via the PROC MIXED program (SAS,
2003). This hierarchical linear modeling (HLM, Bryk & Raudenbush, 2002)
allowed us to determine both the average rate of change and individual
variability in change over time. Age was the variable that represented the
passage of time in our growth model. To interpret the results, we centered
age at 12.48 (sample mean chronological age at wave 2), so that intercepts
reﬂected the expected performance at that age. Our growth model yielded
parameter estimates that deﬁned both the overall trajectory of the sample
(ﬁxed effects) and deviations in the overall trajectories (random effects).
This unconditional linear random regression model was expressed as:
y
ij
¼b
0þb
1ðage
ijÞþU 0jþU1jðage
ijÞþR ij
wherey ijis the dependent variable (e.g., passage comprehension) measured
at timeiin childj, age
ijis childj’s age at timei,b
0is the average intercept at
12.48.b
1is the linear slope.U
0jis the random intercept for childjin the
sample as a whole.U
1jis the random age slope for childj. R
ijis the residual
for childjat timei. The between-child variance components,t
2
0
¼
VarðU
0jÞ;t
2
01
¼VarðU 01jÞandt
2
1
¼VarðU 1jÞreﬂected individual differences
Table 2.Correlations with Working Memory Measures as a Function
of Testing Conditions (Initial, Gain, Maintenance) and Criterion
Measures for the Total Sample.
Wave 3 Measures
Passage comprehension Nonword ﬂuency Receptive language
Composite wave 1
Verbal WM measures
Initial 0.51

0.27 0.56

Gain 0.42

0.28 0.51

Main 0.50

0.34

0.37

Probes 0.13 0.02 0.20
Visual-spatial WM measures
Initial 0.29

0.11 0.31

Gain 0.23 0.06 0.29

Maintenance 0.27

0.05 0.38

Probes 0.06 0.06 0.17

po.05,

po.01,

po.001.
H. LEE SWANSON AND MICHAEL OROSCO14

in level of performance for the sample as a whole, and the rate of change
between children, respectively.
We estimated the association between the outcome (e.g., reading
comprehension) and repeated measures of age across the three-year time
periods. Both unconditional and conditional models are reported. For the
unconditional model, reported are the ﬁxed effects for the intercept value for
wave 2 at age 12.48 and the average rate of change across individuals. For
the random effects, the estimates are the variance around the sample
intercept and the slope related to change overtime. Signiﬁcant random
effects indicated that children differed in intercepts and/or rate of change
(slopes).
For the conditional model, we tested whether entering WM span scores
from the DA conditions into the model explained any statistically signiﬁcant
associations related to ﬁxed and random effects. When one or more
predictors are introduced in the conditional model, there are reductions in
the magnitude of the various components. These reductions are analogous
to effect sizes (Snijders & Bosker, 1999).
To evaluate the compatibility of the data with our conditional model, we
tested the signiﬁcance of the model change. This was done by comparing the
differences between the deviance values (i.e., the likelihood value for the
correspondence between model and data) from the unconditional and
conditional growth models. These are chi-square values, and the number of
parameters added for the conditional model serves as degrees of freedom. In
general, models with lower deviance ﬁt better than models with higher
deviance values. For the present study, a signiﬁcantly lower deviance score
for the conditional model indicated that the conditional model showed a
better ﬁt to the data than the unconditional model.

Prior to the analysis, we converted raw scores for the criterion measures
(passage comprehension, nonword ﬂuency, receptive language) toz-scores.
Wave 2 and 3 measures werez-scored based upon the means and standard
deviations of wave 1. The results were centered on wave 2. The values were
centered at wave 2 because group differences were ﬁrmly established on a
number of measures at wave 1. Age in months at each testing wave was used
to measure change.
Tables 3–5 show the estimates related to growth and the moderating
effects of DA conditions for reading comprehension, ﬂuency, and
vocabulary, respectively. For the conditional models, Model 2 included the
categorical variable for skilled versus children with RD (combined
performance of RD-only, RDþMD, low verbal IQ poor readers). Model
3 added the testing conditions for the verbal WM measures. Also of interest
Predictive Validity of Dynamic Testing 15

was whether DA testing conditions interacted with ability group
performance.
Passage Comprehension (WRMT)
Table 3 shows the outcomes for one unconditional and two conditional
models. To interpret the table, ﬁrst consider the unconditional model for the
ﬁxed effects. The ﬁxed effects model provided an estimate of the intercept
(averagez-score score at wave 2) for the total sample. As shown, the average
Table 3.Hierarchical Linear Modeling for Growth in Passage
Comprehension.
Model Passage Comprehension
12 3
Random effects
Variance SE Variance SE Variance SE
Intercepts .28

.16 .34

.17 .26

.04
Slope .10

.02 .06

.02 .04

.01
Residual .17

.02 .16

.02 .14

.02
Fit-index Fit-index Fit-index
Deviance 338.0 296.1 255.9
Fixed effects
Estimate SE Estimate SE Estimate SE
Intercept .15

.05 .69

.09 .30

.14
RD .75

.11 .23 .14
Slope 0.01

0.001 .014

.004 .01

.004
RD .0002 .0001 .001 .001
Moderators
Initial .29

.11
Gain .11 .16
Main. .31

.14
Probe .17 .12
Interaction
Gain .32

.14
Main. .13 .15

po.05,

po.01.
Note:Interactions, ability groupdynamic testing condition; main., maintenance testing
condition; gain, gain testing condition; initial, initial testing condition; probe, number of
probes used to establish gain condition; RD, less skilled readers (combined subgroups of
childrenwithRD).
H. LEE SWANSON AND MICHAEL OROSCO16

passage comprehensionz-score (meanz-score of test administered at time
2) for all children was .15. The slope (.01) indicated that the average child in
the total sample improved .01 units for each testing wave. Also presented in
Table 3 for the unconditional model were the random effects portions of the
model for the total sample. The random effects included the variance of
conditions related to the intercept (.28) and slope (.10). The residual error
was .17. The random effects reveal that 70% [.28þ.10)/(.28þ.10þ.17)] of
the between-subject variance in passage comprehension was related to
random effects of child differences at the intercept and slope.
Table 4.Hierarchical Linear Modeling for Growth in Nonword
Fluency.
Model TOWRE: Nonword Fluency
12 3
Random effects
Variance SE Variance SE Variance SE
Intercept 1.07

0.29 0.88

0.22 .73

.32
Slope 0.27

0.11 0.22

0.08 .07

.02
Residual 0.32

0.04 0.24

0.03 .20

.03
Fit-index Fit-index Fit-index
Deviance 447.9 383.5 365.1
Fixed effects
Estimate SE Estimate SE Estimate SE
Intercept 0.20

0.07 1.06

0.12 .57

.20
RD 1.20

0.15 .83

.19
Slope .011 .002 0.01

0.004 .007 .005
RD 0.002 0.005 .0002 .002
Moderators
Initial .29 .16
Gain .57

.22
Main. .14 .19
Probe .35

.18
Interaction
Gain .34 .20
Main. .18 .21
po.05,

po.01.
Note:Interactions, ability groupdynamic testing condition; main., maintenance testing
condition; gain, gain testing condition; initial, initial testing condition; probe, number of
probes used to establish gain condition; RD, less skilled readers (combined subgroups of
childrenwithRD).
Predictive Validity of Dynamic Testing 17

The next column (Model 2) in Table 3 shows the conditional model that
included the entry of the classiﬁcation variable (skilled vs. children with
RD). The model indicated that the average passage comprehensionz-score
for the children in wave 2withoutreading problems was .69. Thus, the
skilled reader at wave 2 had gained approximately 3/4 of a standard
deviation in passage comprehension performance. Model 2 also showed that
without entering moderator variables related to testing conditions, passage
Table 5.Hierarchical Linear Modeling for Growth in Receptive
Language.
Model Receptive Language
123
Random effects
Variance SE Variance SE Variance SE
Intercept 0.44

0.15 0.34

0.12 .33

.11
Slope 0.08

0.02 0.06

0.01 .04

.01
Residual 0.13

0.02 0.13

0.02 .11

.01
Fit-index Fit-index Fit-index
Deviance 320.7 301.5 265.1
Fixed effects
Estimate SE Estimate SE Estimate SE
Intercept 0.17

0.05 0.55

0.10 .30

.16
RD 0.53

0.11 .22 .15
Slope 0.028

0.002 0.02

0.003 .007 .004
RD 0.001 0.004 .003

.001
Moderators
Initial .29

.12
Gain .07 .18
Main. .37

.15
Probe .08 .14
Interactions
Gain .10 .15
Main. .48

.16

po.05,

po.01.
Note:Interactions, ability groupdynamic testing condition; main., maintenance testing
condition; gain, gain testing condition; initial, initial testing condition; probe, number of
probes used to establish gain condition; RD, less skilled readers (combined subgroups of
childrenwithRD).
H. LEE SWANSON AND MICHAEL OROSCO18

reading comprehension performance was .75z-score pointslower(estimate
was.75) for children with RD when compared to skilled readers. The
growth rate for skilled readers was approximately .02 (.014) units of
improvement per testing session. Although the estimate for children with
RD showed a lower growth rate (.0002), the estimate was not signiﬁcantly
different from skilled readers.
A comparison was made between the unconditional and conditional
models on the random intercept (.28 vs. .34) and slope (.10 vs. .06). The
addition of the classiﬁcation measure accounted for approximately 40% of
the explainable variance in growth [(.10–.06)/.10)]. However, the variance
increased between children on the intercept values, suggesting the model was
not a good ﬁt to the data.
Model 3 added probe scores and span scores from the WM testing
conditions. Also included were potential interactions between reading
ability group and the DA condition. The results indicated that WM
performance related to initial and maintenance testing for the verbal tasks
and the ability groupgain testing condition moderated passage compre-
hension. To interpret these ﬁndings it is important to note that predictor
variables were centered. That is, the estimate for passage comprehension
was .30 when predictors were set to zero. The estimate for the maintenance
testing condition was .31. Taken together, the results indicated that children
who differ by 1 point on WM for the maintenance testing condition differ
by .31 points on passage comprehension. The negative estimates for the
interaction indicated that skilled readers were more likely to beneﬁt from the
gain conditions than children with RD. A deviance test was computed to
determine the ﬁt of Model 3. A deviance test indicated that the Model 3
provided a better ﬁt to the data than Model 2 for passage comprehension,w
2
(df¼6)¼40.9 (296.1–255.9),po.001. When compared to Model 2, the
measurement of WM across various testing conditions in Model 3 accounted
for 24% of the explainable variance in intercepts [(.34–.26)/.34] and 33% of
the explainable variance between children in growth [(.06–.04)/.06].
Nonword Fluency (TOWRE)
Table 4 shows the results for predicting reading ﬂuency for nonwords. The
total sample averagez-score (meanz-score of test administered at time 2) for
all children was .20. The slope (.01) indicated that children improved .01
units for each testing wave. The random effects for the unconditional model
were signiﬁcant for the intercept (1.07), slope (.27), and the residual error
(.32.) The random effects reveal that 81% [1.07þ.27)/(1.07þ.27þ.32)] of
Predictive Validity of Dynamic Testing 19

the between-subject variance in nonword ﬂuency was related to child
differences at the intercept and slope.
Model 2 indicated that the nonword ﬂuency z-scores for skilled
reading children at wave 2 was 1.06. Children with RD were1.20z-score
points lower than skilled readers. However, no differences were found
between the two groups in growth. The addition of the classiﬁcation
measure in Model 2 explained about 19% of the variance in growth
[(.27–.22)/.27)].
Model 3 added probe scores and span scores for the verbal WM testing
conditions. As shown for the random effect, the intercepts and slopes
decreased when compared to Model 2. The results indicated that span scores
from the gain condition and probe scores signiﬁcantly moderated nonword
ﬂuency. The results indicated children who differed by 1.0 with respect to
the scores on the gain testing condition for verbal WM tasks differed in
intercept values by .57 in the nonword ﬂuency. The deviance test indicated
that Model 3 showed a good ﬁt to the data when compared to Model 2,w
2
(df¼6)¼18.41 (383.50–365.1),po.001. When compared to Model 2, the
measurement of WM across various testing conditions in Model 3 accounted
for 17% of the explainable variance in intercepts [(.88–.73)/.88] and 68% of
the explainable variance between children in growth [(.22–.07)/.22].
Vocabulary
As shown in Table 5, the ﬁxed effects in Model 3 indicated that the slopes
were higher for children with RD than skilled readers. The important
ﬁndings were that the initial and maintenance conditions moderated
receptive vocabulary. As shown, the ﬁxed effects for Model 3 eliminated
group differences at wave 2 for the intercept estimates. Performance on the
maintenance testing condition was a signiﬁcant moderator of receptive
vocabulary as was the ability groupmaintenance testing condition
interaction. The results suggested that children skilled in reading were
better able to sustain WM performance at their highest level (based on gain
testing conditions) without cues than children with RD. The deviance test
indicated that the Model 3 showed a good ﬁt to the data when compared to
Model 2,w
2
(df¼6)¼36.40 (301.50–265.1),po.001. When compared
to Model 2, the measurement of WM across various testing conditions in
Model 3 accounted for only 3% of the explainable variance in intercepts
[(.34–.33)/.24], but 33% of the explainable variance between children in
growth [(.06–.04)/.06].
H. LEE SWANSON AND MICHAEL OROSCO20

Summary
The growth modeling showed that various DA conditions moderated
outcomes on the criterion measures. Maintenance testing conditions were
important moderators in the growth model of passage comprehension and
receptive vocabulary, whereas gain testing conditions and probing moderated
nonword ﬂuency. The conditional model that included measures of WM
accounted for approximately 33% of the explainable growth variance in
passage comprehension and receptive vocabulary and 68% of growth in
nonword ﬂuency. Ability group interactions emerged for gain testing
conditions for passage comprehension, and maintenance testing conditions
for vocabulary (receptive language). These interactions indicated that skilled
readers were more likely to beneﬁt from those DA conditions than children
with RD.
SUMMARY OF FINDINGS AND IMPLICATIONS
We will now consider the two questions that motivated this preliminary work.
1. Does dynamic assessment contribute to later predictions of achievement
in children with RD?
The results of the HLM analysis support the hypothesis that DA adds
important variance in predicting later reading achievement. These results
showed that various DA conditions (e.g., conditions related to gain,
probing, and/or maintenance) signiﬁcantly moderated passage comprehen-
sion, nonword reading ﬂuency, and receptive language. Interestingly,
although intercept values favored skilled readers on all criterion measures,
growth on the reading (passage comprehension, nonword ﬂuency) and
receptive language measures was statistically comparable between skilled
and children with RD.
These ﬁndings complement some of our earlier work. Our earliest study
(Swanson, 1992, Experiment 2) assessed whether dynamic testing of WM
predicted concurrent reading recognition performance when initial testing
conditions were partialed out in the analysis. In this study, children (mean
age 10 years) were administered WM tasks under initial, gain, and
maintenance conditions. Effect sizes (ES) for gain and maintenance
performance relative to initial conditions (ES .90 for gain scores and .47
for maintenance scores) were in the same range as the present study. The
Predictive Validity of Dynamic Testing 21

regression analysis found that DA conditions contributed a signiﬁcant
variance (approximately 5%) to concurrent performance on measures of
word recognition (see Sternberg & Grigorenko, 2002, for a critical analysis
of this ﬁnding). This earlier study was broaden to a larger sample ranging in
ages 5–85 (Swanson, 1999) and that again found that DA procedures
contributed approximately 5% of the variance reading and math
performance.
So what does this small amount of variance buy us? To answer that
question we need to consider the logic of the testing conditions (also see
Swanson, 2003, for a more detailed discussion). The initial testing condition
reﬂects the baseline for each participant’s self-initiated processes to access
information. The gain testing condition enhances the access to stored items
by tailoring cues to help participants reinstate memory traces or to retrieve
forgotten items from the initial (or baseline) conditions. Our previous studies
(Swanson, 1992, 1999; Swanson, Ashbaker, & Lee, 1996) and the current
study have shown that the gain conditions improve performance by as much
as 1 standard deviation. We assume this occurs because the systematic cuing
procedures emphasize sequential processing strategies and thereby reduces
the number of competing strategies employed. We also assume that if the
locus of WM problems in children with RD is in the retrieval phase, one
would expect a reduction in residual difference between children with and
without RD when compared to the initial testing condition. As shown in this
study and our earlier studies we have been unable to reduce the residual
differences between reading groups in the gain condition. Thus, one could
argue that children with RD suffer WM deﬁcits related to retrieval efﬁciency.
One reservation in arguing that WM problems in children with RD are
related to problems in retrieval efﬁciency (i.e., improved access to items
previously forgotten), however, is that the manipulations between the initial
(noncued) and gain (cued) condition are limited to individual differences as
to what was actually attended to and stored by the child. Thus, the
maintenance condition allowed for us to examine whether WM deﬁcits reﬂect
capacity constraints in accessing what has been previously stored. For this
condition, thesameWM tasks that matched each participant’s highest WM
span level (gain score) are again administered, but without cues. Thus, each
participant is presented items calibrated to their asymptotic level of WM
performance. Calibrating this condition allows us to capture processing
differences between groups beyond the simple learning of items. The inability
to access items that have been previously learned (stored and retrieved) and
have been calibrated to the individual span level we assume reﬂects
fundamental constraints in WM capacity.
H. LEE SWANSON AND MICHAEL OROSCO22

Based on the above logic, we consider the results that are related to the
growth curve analysis. The conditional models showed that maintenance
testing conditions moderated passage comprehension and receptive lan-
guage, whereas gain testing conditions moderated nonword ﬂuency.
Consistent with the above logic, we assume that individual differences as
a function of maintenance testing conditions reﬂects constraints in WM
capacity. These constraints were not directly tied to simple retrieval
strategies (e.g., feedback) or previous learning (e.g., item familiarity). In
contrast, individual differences in gain conditions reﬂect limitations in the
activation of new information. Thus, DA informs us to some degree that
constraints in accessing previously remember information may play a role in
some areas of reading (comprehension) and the efﬁciency in accessing new
information in other areas of reading (nonword ﬂuency). However, these
results are qualiﬁed by the ability groupDA interactions. The interaction
with reading groupgain testing condition when predicting passage
comprehension suggests that a failure to activate new information in
storage is an important determinant of reading skill. Further it appears from
the reading groupmaintenance interaction that when predicting receptive
vocabulary the ability to sustain performance after environmental (probes)
support is removed maybe constrained in children with RD. Clearly, these
conclusions are speculative on our part but the results do suggest that
different processes may be activated as a function of DA when predicting
performance on the criterion measures.
2. Do children with RD differ from children with more generalized
achievement deﬁcits as a function of dynamic assessment conditions?
As expected, skilled readers outperformed children with RD on the WM
measures. This occurred even though children were statistically matched on
measures of ﬂuid intelligence. Interestingly, no signiﬁcant differences were
found on measures of WM between children with RD who exhibited low
verbal IQ from children with RD-only and RDþMD. This ﬁnding is in
contrast to other studies (e.g., van der Sluis, van der Leij, & de Jong, 2005)
which have suggested that RD-only children exhibit no problems across
domains that draw upon executive processing (such as updating as required
by WM tasks, see Miyake, Friedman, Emerson, Witzki, & Howerter, 2000),
whereas children who show reading plus arithmetic difﬁculties have
problems in executive processing (domain general processing). We did not
ﬁnd support for the notion that more generalized WM deﬁcits occur only in
the comorbid group when compared to children with RD only. Further,
because more generalized deﬁcits in achievement (problems in both reading
Predictive Validity of Dynamic Testing 23

and math) may be associated with verbal IQ, we also explored whether
group differences in WM measures were merely an artifact of verbal ability.
No signiﬁcant differences were found between poor readers with low verbal
IQ and children with RD-only on measures of WM. Thus, we did not ﬁnd
support for the notion that WM deﬁcits in children with RD are merely an
artifact of verbal IQ or comorbidity. Although the results indicated that
ability group differences were larger on verbal WM tasks than visual-spatial
WM tasks, performance among subgroups of children with RD was in the
same low range. We now consider in more detail the ﬁndings related to the
weak differentiation of subgroups as a function of DA and the domain
general ﬁndings related to WM tasks.
Subgroups and Dynamic Assessment
The present ﬁndings of this study partially support the previous work of
Swanson and Howard (2005), which showed that that WM performance
and DA procedures enhanced predictions of reading. The results of
Swanson and Howard showed that poor readers and children with RD-
only or MDþRD generally perform in the same low range on verbal WM
tasks. These results alsocoincide with others’ showing that poor readers
and children with RD are difﬁcult to separate on cognitive measures (e.g.,
Siegel, 1992; see Hoskyn & Swanson, 2000, for a review). The Swanson
and Howard study showed children with RD and/or RDþMD were more
likely than the other subgroups to return to their initial score performance
after presentation of probes (feedback) had been stopped. In contrast, the
present study found that effect size scores related to gain and maintenance
testing for all ability groups were in the same range. However, the present
results did suggest that a drop in performance from the gain to
maintenance conditions was more likely to occur for the RD-only group
and the RDþMD group was more likely inﬂuenced by initial testing
conditions than the actual probing. No doubt, the differences between the
previous study and the present study may be due to sampling. The earlier
study focused on younger children of middle to low SES background. In
contrast, the present sample was older, came from higher SES homes (in
terms of parent education), more schooling, and therefore both cognitive
and reading performance may have been more stable across the testing
sessions.
It is important to note that our inability to separate children with RD-
only from other children with reading problems is not unique to this study.
H. LEE SWANSON AND MICHAEL OROSCO24

For example, a meta-analysis comparing children with MD-only, RD-only,
and RDþMD on measures independent of the classiﬁcation measures
placed their performance on cognitive measures in the same range
(Swanson & Jerman, 2006). For example, Swanson and Jerman found
that the magnitude of ES across cognitive measures were miniscule
between children with RD-only and MD-only (.10) and between MD-
only and RDMD (M ES¼.26). More important, the means ES between
RD-only and MD-only groups were .06 for STM of words, .03 for STM
digits,.07 verbal WM, and.30 for visual-spatial WM. In general, these
results suggested that MD and RD children share common memory
deﬁcits. Unraveling how the common memory deﬁcits manifest themselves
in different forms on achievement measures, of course, needs to be
explored.
Limitations
There are, of course, several limitations to our preliminary work. Four are
apparent. First, we focused on children who came from average to high SES
homes with highly educated parents. We assumed that previous studies that
have tested children with RD had failed to control for SES (or focused on
low SES participants) and therefore differences in reading in previous
studies could be attributed to low SES and/or instructional factors.
Regardless, our ﬁndings can only be generalized to a select group of
children. Second, children with undiagnosed reading problems may yield
different results. The children with RD in this study had the diagnosis of
learning disabilities and the information from their cumulative ﬁle indicated
a persistent reading problem in their elementary school years. This study
sample was also unique in that all children had ﬂuid intelligence scores
(Raven test performance) in the normal range. If ﬂuid intelligence scores
were in the same low range as their reading scores, the results may have been
different. Third, attrition in the sample occurred at wave 3. Although we did
not ﬁnd signiﬁcant differences between the retained and nonretained
samples across the various measures, the study was limited in terms of
sample size when correlating wave 1 and wave 3 testing. Finally, these
results need to be established by researchers independent of our research
team. Although our DA measures are normed referenced (Swanson, 1995a),
the majority of studies independent of our research team using the measures
have been isolated to dissertations.
Predictive Validity of Dynamic Testing 25

Summary
In summary, DA procedures moderated later performance on measures of
passage comprehension, nonword ﬂuency, and receptive language. The
results indicated that children with RD and skilled readers both beneﬁted
from DA procedures, but DA procedures did not statistically differentiate
between the subgroups of children with RD. A descriptive analysis suggested
that greater reductions occur between gain and maintenance testing conditions
for the RD only group and the effect sizes related to DA conditions RDþMD
group are more likely related initial testing sensitivity than to procedures that
test the limits. Clearly, additional research on the properties of DA and
cognitive measures is needed, especially studies linking the malleability of WM
to long-term instructional gains.
NOTE
1. Under most training circumstances with no ceiling or ﬂoor effects, pretest and
posttest variable standard deviations are expected to be similar (see Carlson &
Schmidt, 1999, p. 853; for a review). Carlson and Schmidt (1999) have argued that if
exposure to the pretest acted as a learning event, then this would have caused
posttest training scores to be higher, but responses to be more variable. Posttest
dependent variable standard deviations would be systematically larger than pretest
standard deviations, and as a result posttest standard deviations would result in
smaller effect sizes than those calculated using pretest standard deviations. On the
other hand, if changes in performance are a function of treatment (in this case
feedback to the examinee) rather than merely retesting, then posttest means would be
higher than pretest means, but posttest standard deviations would be comparable to
pretest standard deviations (Carlson & Schmidt, 1999). In our case, we assumed that
initial scores served as a pretest measure and gain and maintenance reﬂected
posttesting.
ACKNOWLEDGMENTS
The authors thank Marshal Raskin and Eleanor Higgins for their
administrative assistance in testing children over the three- to ﬁve-year
period. The authors also thank the various testers (Crystal Howard, Leilani
Saez, Olga Jerman, Diane Luxenberg, and Diane Dowds) who as doctoral
students played an important role in monitoring children’s progress. The
technical aspects of this preliminary work are detailed for reading
comprehension (Swanson, in press) and ﬂuency (Swanson, 2010) elsewhere.
H. LEE SWANSON AND MICHAEL OROSCO26

REFERENCES
Brown, A. L., & Ferrara, R. A. (1999). Diagnosing zones of proximal development. In: P. Lloyd
(Ed.),L.Vygotsky: Critical assessments: The zones of proximal development(Vol. III,
pp. 225–256). New York: Routledge.
Bryk, A., & Raudenbush, S. W. (2002).Hierarchical linear models(ed.). Newbury Park, CA: Sage.
Caffrey, E., Fuchs, D., & Fuchs, L. S. (2008). The predictive validity of dynamic assessment:
A review.The Journal of Special Education,41, 254–270.
Carlson, J. S., & Wiedl, K. H. (1979). Toward a differential testing approach: Testing-the-limits
employing the Raven matrices.Intelligence,3, 323–344.
Carlson, K. D., & Schmidt, F. L. (1999). Impact of experimental design on effect size: Findings
from the research literature on training.Journal of Applied Psychology,84, 851–862.
Cohen, J. (1988).Statistical power analysis for the behavioral sciences(2nd ed.). Hillsdale, NJ:
Erlbaum.
Daneman, M., & Carpenter, P. A. (1980). Individual differences in working memory and
reading.Journal of Verbal Learning and Verbal Behavior,19, 450–466.
De Beni, R., Palladino, P., Pazzaglia, F., & Cornoldi, C. (1998). Increases in intrusion errors
and working memory deﬁcit of poor comprehenders.Quarterly Journal of Experimental
Psychology: Human Experimental Psychology,51, 305–320.
De Jong, P. (1998). Working memory deﬁcits of reading disabled children.Journal of
Experimental Child Psychology,70, 75–95.
Dunn, L., Dunn, L. M., Dunn, L. M., & Dunn, D. M. (1997).The peabody picture vocabulary
test(3rd ed.). Circle Pines, MN: American Guidance Services, Inc.
Embretson, S. E. (1992). Measuring and validating cognitive modiﬁability as an ability: A study
in the spatial domain.Journal of Educational Measurement,29, 25–50.
Gathercole, S. E., Alloway, T. P., Willis, C., & Adams, A. (2006). Working memory in children
with reading disabilities.Journal of Experimental Child Psychology,93, 265–281.
Gathercole, S. E., & Pickering, S. J. (2000a). Assessment of working memory in six- and seven-
year-old children.Journal of Educational Psychology,92, 377–390.
Gathercole, S. E., & Pickering, S. J. (2000b). Working memory deﬁcits in children with low
achievements in the national curriculum at 7 years of age.British Journal of Education
Psychology,70, 177–194.
Grigorenko, E. L., & Sternberg, R. J. (1998). Dynamic testing.Psychological Bulletin,124,
75–111.
Hoskyn, M., & Swanson, H. L. (2000). Cognitive processing of low achievers and children with
reading disabilities: A selective meta-analytic review of the published literature.School
Psychology Review,29, 102–119.
Hwang, Y., Hosokawa, T., Swanson, H. L., Ishizaka, I., Kifune, N., Ohira, D., et al. (2006). A
Japanese short form of the Swanson-Cognitive Processing Test to measure working
memory: Reliability, validity, and differences in scores between primary school children
of the United States and Japan.Psychological Reports,99, 27–38.
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., & Howerter, A. (2000). The unity
and diversity of executive functions and their contributions to complex frontal lobe
tasks: A latent variable analysis.CognitivePsychology,41, 49–100.
Pickering,S.
J. (2006). Working memory in dyslexia. In: T. P. Alloway & S. E. Gathercole
(Eds),Working memory and neurodevelopmental disorders(pp. 7–40). New York:
Psychology Press.
Predictive Validity of Dynamic Testing 27

Raven, J. C. (1976).Colored progressive matrices. London, England: H. K. Lewis & Co. Ltd.
Rosenthal, R. (1994). Parametric measures of effect size. In: H. Cooper & L. V. Hedges (Eds),
The handbook of research synthesis(pp. 232–260). New York: Russell Sage Foundation.
SAS Institute Inc. (2003).SAS/STAT user’s guide. Cary, NC: SAS Institute.
Siegel, L. S. (1992). An evaluation of the discrepancy deﬁnition if dyslexia.Journal of Learning
Disabilities,25, 618–629.
Siegel, L. S., & Ryan, E. B. (1989). The development of working memory in normally achieving
and subtypes of learning disabled.Child Development,60, 973–980.
Singer, J. D. (2002). Fitting individual growth models using SAS PROC MIXED. In:
Moskowitz & S. Hersberger (Eds),Modeling intraindividual variability with repeated
measures data: Methods and applications(pp. 135–170). Mahwah, NJ: Erlbaum.
Singer, J. D., & Willett, J. B. (2003).Applied longitudinal data analysis. New York: Oxford
University Press.
Snijders, T. A., & Bosker, R. J. (1999).Multilevel analysis. Thousand Oaks, CA: Sage.
Sternberg, R. J., & Grigorenko, E. L. (2002).Dynamic testing: The nature and measurement of
learning potential. New York: Cambridge University Press.
Swanson, H. L. (1992). Generality and modiﬁability of working memory among skilled and less
skilled readers.Journal of Educational Psychology,84, 473–488.
Swanson, H. L. (1993). Working memory in learning disability subgroups.Journal of
Experimental Child Psychology,56, 87–114.
Swanson, H. L. (1995a).S-cognitive processing test. Austin, TX: PRO-ED.
Swanson, H. L. (1995b). Using the cognitive processing test to assess ability: Development of a
dynamic assessment measure.School Psychology Review,24, 672–693.
Swanson, H. L. (1996). Individual and age-related differences in children’s working memory.
Memory & Cognition,24, 70–82.
Swanson, H. L. (1999). What develops in working memory? A life span perspective.
Developmental Psychology,35, 986–1000.
Swanson, H. L. (2003). Age-related differences in learning disabled and skilled readers’ working
memory.Journal of Experimental Child Psychology,85, 1–31.
Swanson, H. L. (2008). Working memory and intelligence in children: What develops?Journal
of Educational Psychology,100, 581–602.
Swanson, H. L. (2010). Does the dynamic testing of working memory predict growth in
nonword ﬂuency and vocabulary in children with reading disabilities.Journal of
Cognitive Education and Psychology,9, 51–77.
Swanson, H. L. (in press). Dynamic testing, working memory and reading comprehension
growth in children.Journal of Learning Disabilities.
Swanson, H. L., Ashbaker, M., & Lee, C. (1996). Working-memory in learning disabled readers
as a function of processing demands.Journal of Child Experimental Psychology,61,
242–275.
Swanson, H. L., & Howard, C. B. (2005). Children with reading disabilities: Does dynamic
assessment help in the classiﬁcation?Learning Disability Quarterly,28,17–34.
Swanson, H. L.,
& Jerman, O. (2006). Math disabilities: A selective meta-analysis of the
literature.Review of Educational Research,76, 249–274.
Swanson, H. L., Jerman, O., & Zheng, X. (2008). Growth in working memory and
mathematical problem solving in children at risk and not at risk for serious math
difﬁculties.Journal of Educational Psychology,100, 343–379.
H. LEE SWANSON AND MICHAEL OROSCO28

Swanson, H. L., & Lussier, C. M. (2001). A selective synthesis of the experimental literature on
dynamic assessment.Review of Educational Research,71, 321–363.
Swanson, H. L., Saez, L., Gerber, M., & Leafstedt, J. (2004). Literacy and cognitive functioning
in bilingual and nonbilingual children at or not at risk for reading disabilities.Journal of
Educational Psychology,96, 3–18.
Swanson, H. L., & Siegel, L. (2001). Learning disabilities as a working memory deﬁcit.Issues in
Education: Contributions from Educational Psychology,7, 1–48.
van der Sluis, S., van der Leij, A., & de Jong, P. F. (2005). Working memory in Dutch children
with reading- and arithmetic-related LD.Journal of Learning Disabilities,38(3), 207–221.
Wagner, R., & Torgesen, J. (1999).Test of word reading efﬁciency. Austin, TX: Pro-Ed.
Wilkinson, G. S. (1993).The Wide Range Achievement Test-3. Wilmington, DE: Wide Range, Inc.
Willcutt, E. G., Pennington, B. F., Olson, R. K., Chhabildas, N., & Hulslander, J. (2005).
Neuropsychological analyses of comorbidity between reading disability and attention
deﬁcit hyperactivity disorder: In search of the common deﬁcit.Developmental
Neuropsychology,27, 35–78.
Woodcock, R. W. (1998).Woodcock reading mastery test – Revised. Minneapolis, MN:
American Guidance.
APPENDIX. COMPOSITE SCORES ( Z-SCORES) FOR
WORKING MEMORY TASKS AS A FUNCTION OF
TESTING CONDITIONS AND TESTING WAVES
Variable RD-Only RD þMD Low VIQ Readers Skilled Readers
Mean SD Mean SD Mean SD Mean SD
Wave 1
V
initial0.39 0.830.25 0.720.13 0.75 0.55 0.73
V
gain 1.0 0.85 1.19 1.2 1.07 0.99 1.66 0.65
V
main 0.36 0.75 0.22 1.05 0.29 0.82 1.03 0.7
V
probe0.01 0.76 0.2 1.16 0.01 0.76 0.11 0.67
V
isinitial0.21 0.650.14 0.860.1 0.75 0.39 0.64
V
isgain0.69 0.83 0.82 1.28 0.64 0.87 1.0 0.88
V
ismain0.01 0.69 0.26 1.01 0.28 0.74 0.75 0.81
V
isprobe0.14 0.74 0.25 1.030.14 0.55 0.02 0.9
Wave 2
V
initial0.11 0.840.24 0.61 0.14 0.64 0.59 0.65
V
gain 1.46 0.72 1.47 1.21 1.64 0.85 2.02 1.13
Predictive Validity of Dynamic Testing 29

Variable RD-Only RD þMD Low VIQ Readers Skilled Readers
Mean SD Mean SD Mean SD Mean SD
V
main 0.65 1.13 0.46 1.28 0.72 0.79 1.08 0.7
V
probe0.11 0.610.02 0.840.22 0.75 0.28 0.8
V
isinitial0.21 0.69 0.07 1.14 0.39 0.83 0.68 0.76
V
isgain0.95 0.9 0.75 1.47 1.08 1.09 1.55 0.9
V
ismain0.76 0.93 0.65 1.34 0.73 1.11 1.23 0.89
V
isprobe0.05 0.650.18 0.720.25 0.67 0.12 0.67
Wave 3
V
initial0.26 0.6 0.34 0.66 0.13 0.74 0.78 0.96
V
gain 1.74 0.55 1.62 0.61 1.53 0.83 2.37 0.95
V
main 0.81 0.86 1.07 0.62 0.79 0.92 1.87 1.07
V
probe0.06 0.690.39 0.70.45 0.73 0.37 0.79
V
isinitial0.57 0.51 0.83 0.77 0.53 0.88 0.77 0.53
V
isgain1.39 0.52 1.92 0.63 1.73 1.08 1.7 0.86
V
ismain1.15 0.61 1.62 0.7 1.33 1.15 1.44 0.7
V
isprobe0.04 0.850.23 0.580.08 0.92 0.35 0.68
Note: Vinitial, initial testing on verbal measures;V isinitial, initial testing on visual-spatial
measures;V
gain, gain testing on verbal measures;V
isgain, gain testing on visual-spatial measures;
V
main, maintenance testing on verbal measures;V ismain, maintenance testing on visual-spatial
measures.
H. LEE SWANSON AND MICHAEL OROSCO30

APPLICATIONS OF CURRICULUM-
BASED MEASURES IN MAKING
DECISIONS WITH MULTIPLE
REFERENCE POINTS
Gerald Tindal and Joseph F. T. Nese
ABSTRACT
We write this chapter using a historical discourse, both in the chronology
of research and in the development that has occurred over the years with
curriculum-based measurement (CBM). More practically, however, we
depict the chronology in terms of the sequence of decisions that educators
make as they provide special services to students with disabilities. In the
ﬁrst part of the chapter, we begin with a pair of seminal documents that
were written in the late 1970s to begin the story of CBM. In the second
part of the chapter, we begin with the ﬁrst decision an educator needs to
make in providing special services and then we continue through the
chronology of decisions to affect change in learning for individual
students. In the end, we conclude with the need to integrate these decisions
with multiple references for interpreting data: normative to allocate
resources, criterion to diagnose skill deﬁcits, and individual to evaluate
instruction.
Assessment and Intervention
Advances in Learning and Behavioral Disabilities, Volume 24, 31–58
Copyrightr2011 by Emerald Group Publishing Limited
All rights of reproduction in any form reserved
ISSN: 0735-004X/doi:10.1108/S0735-004X(2011)0000024004
31

EARLY VIEWS ON INSTRUCTIONAL EVALUATION
In the late 1970s, Deno and Mirkin developed a concept they entitled data-
based program modiﬁcation, espousing the need for measures that teachers
could use to evaluate instruction over time (Deno & Mirkin, 1977). They
emphasized speciﬁc features of these measures to allow generating alternate
forms so that a time series data display could be used to determine if
programs were working or needed to be adjusted. For example, the
measures had to be easy to create, quick to administer, usable by all, from
the curriculum, and with sufﬁcient technical adequacy. In the end, they laid
the groundwork for an experimental view of teaching in which any
intervention was considered a hypothesis that needed vindication.
At nearly the same time, Engelmann, Granzin, and Severson (1979) wrote
about the need for diagnosing instruction. Their thesis was that problems in
learning were seldom the fault of students; rather they often resulted from
ineffective programs even though we spend an inordinate amount of time
directly trying to diagnose the problem as the student’s. A central warrant of
their thesis was that, while it is impossible to ascertain the certainty of
problems ‘‘inside’’ the student, it is very much possible to determine
programs are not working, change them, and only upon continued lack of
success, make inferences about the problem being with the student.
Both approaches relied upon the scientiﬁc method in which a null
hypothesis (a notion that is not really believed with much certainty) is
posited and countered by an alternative hypothesis. Rather than prove
anything in a deﬁnitive manner, teachers needed to disprove that their
actions are ineffective (posit a null hypothesis and then provide data to
counter it). For Deno and Mirkin (1977), experimental teaching provided a
time series design in which each student’s performance is compared to
previous performance (in level, slope, and variation), with the impetus for
instructional change residing in the lack of student change. For Engelmann
et al. (1979), the null hypothesis is similarly situated with repeated failure to
affect change as the only convincing evidence that supports the problem as
being with the student (and not an instructional one).
Ironically, they approached the problem from opposite ends of the
continuum. For Deno and Mirkin (1977), instruction was tentative and
ascertained only through measurement. For Engelmann et al. (1979),
diagnosis was entirely instructional with measurement only implied. Yet, in
an interesting conﬂuence of history, both have become signatures of
effective teaching that has current play in any ‘‘response to intervention’’
(RTI) model. In this chapter, we provide a brief history of this perspective
GERALD TINDAL AND JOSEPH F. T. NESE32

and then discuss the next generation of curriculum-based measurement
(CBM) that features alignment to content standards and rigorous statistical
modeling analytics. We review the empirical literature that has resulted from
more widespread adoption of CBM, in both the kinds of problems being
investigated as well as the variety of applications appearing in educational
systems. We conclude with classroom and student case studies that echo
these developments.
CBM research initially focused on measurement validation, different
sampling plans, and measurement designs. A number of studies also focused
on highlighting the central feature of CBM as reﬂecting a ‘‘general
outcome’’ measure rather than a curriculum-based one. Although efforts
to change the language failed and CBM became the parlance in use, more
recent efforts with the National Center on Response to Intervention
(NCRTI) (www.rti4success.org) reﬂect a return to more precise language
that references progress monitoring as oriented to either mastery monitoring
or general outcomes.
Mastery monitoring was initially deﬁned by Deno and Mirkin (1977) in
terms of the graphic display: ‘‘A progress (or mastery) graph is constructed
to display the time it is taking a student to master a set (usually ordered over
time in terms of sequence and/or complexity) of instructional objectives’’
(p. 33) with they-axis reﬂecting units traversed and thex-axis reﬂecting time
taken (usually days). The most signiﬁcant feature is that the domain for
measurement comes from a constricted sample of items (problems) being
taught during the evaluation period. The classic example of mastery
monitoring is the weekly spelling list in which 10–15 words are taught during
the week, and then students receive a test on this list of words at the end of
the week. Another feature of mastery monitoring is the graphic display,
reﬂecting a series of step increases plotted against some deﬁnition of
mastery; with mastery, performance moves up one unit for the next unit of
time, and without mastery, performance is recorded at the same level for the
next unit of time. Without incremental movement up in mastery, the graphic
display of performance would simply hover at a steady level over time.
General outcome measures were likewise described by Deno and Mirkin
(1977) in terms of performance graphs: ‘‘designed to display how a student’s
behavior changes on a single taskyover time’’ (p. 34) with they-axis
reﬂecting how well the student performed on the task and thex-axis
reﬂecting time (in days or weeks). In this system, items (problems) are
sampled across an alternate unit of instructional time, from the early
through late skills that are being introduced over a year (or possibly over
years). The sampling plan effectively includes items that preview what is to
Applications of CBMs in Making Decisions 33

be taught as well as reviews what has been taught. The graphic display
allows raw scores to change without converting them to a common scale.
The graphic display includes three metrics to evaluate programs: (a) slope of
improvement, (b) variation among successive data values, and (c) immediate
changes after a program has been introduced. By looking for increases in
slope, decreases in variation, and large changes in level (from pre- to
immediately postintervention), teachers can determine whether or not their
programs are effective.
THREE DECADES OF RESEARCH ON CBM FROM
THE 1980s THROUGH 2010
In one of the early deﬁnitions of CBM, Deno (1987) stated:
the term curriculum-based assessment, generally refers to any approach that uses direct
observation and recording of a student’s performance in the local school curriculum as a
basis for gathering information to make instructional decisionsyAs noted earlier, the
term curriculum-based measurement refers to a speciﬁc set of procedures created
through a research and development programyand grew out of theData-Based
Program Modiﬁcationsystem developed by Deno and Mirkin (1977) (p. 41).
CBMs, which sample skills related to the curriculum material covered in a
given year of instruction, provide teachers with a snapshot of their students’
current level of performance as well as a mechanism for tracking the
progress students make in gaining desired academic skills. Historically,
CBMs typically are very brief individually administered measures (Deno,
2003).
CBM is distinct in two important respects: (a) the procedures reﬂect
technically adequate measures with reliability and validity present to a
degree that equals or exceeds that of most achievement tests and
(b) ‘‘growth is described by an increasing score on a standard, or constant,
task. The most common application of CBM requires that a student’s
performance in each curriculum area be measured on a single global task
repeatedly across time’’ (Deno, 1987, p. 41). After the initial research
conducted at the Institute for Research on Learning Disabilities (IRLD) at
the University of Minnesota, scores of the technical reports became
publications in the refereed literature; eventually, as the practice of CBM
became more widespread, scores of chapters were written with two
important books published with Shinn as the editor and most of the
original researchers contributing chapters (1989, 1998). Much of this
GERALD TINDAL AND JOSEPH F. T. NESE34

research has been with and on the use of performance measures or general
outcome measures and not the mastery measures.
Reading Research
The history of research on CBM is rich and rather extensive, from the initial
work in the late 1970s and early 1980s on through more recent research in
the 2000s. Wallace, Espin, McMaster, and Deno (2007) summarize this
research in a recent special series on curriculum-based measurement. Over
the initial 20 years of research on reading CBM, a number of technical
characteristics were investigated, including:
1. Content-related evidence addressing domain size (Fuchs, Tindal, &
Deno, 1984) and difﬁculty of material (Fuchs & Deno, 1992; Shinn,
Gleason, & Tindal, 1989).
2. Criterion-related evidence in the relation of oral reading ﬂuency to
reading comprehension (Shinn, Good, Knutson, Tilly, & Colllins, 1992)
and predictions of performance (Fewster & Macmillan, 2002).
3. Construct validation with respect to rival hypotheses related to
psychological variables (Kranzler, Brownell, & Miller, 1998), as well as
other psychometric concerns (Fuchs, Fuchs, & Maxwell, 1988; Mehrens &
Clarizio, 1993).
4. The effects of curriculum-based planning on grouping students (Wesson,
Vierthaler, & Haubrich, 1989) and making instructional decisions
(Allinder, 1996; Fuchs, Fuchs, & Stecker, 1989; Wesson, et al., 1988).
As a useful formative assessment method, CBM has been extensively
researched, a summary of which is beyond the scope of this chapter; these
topics are simply illustrative of the kind of research that has been done over
the past 30 years. Much of this research, however, has been conducted with
elementary students and has focused on assessment of reading. With the
establishment of CBM as a useful means of assessment for decision-making,
we can apply its utility to mathematics, as it has been applied to reading.
Mathematics Research
The research on mathematics did not begin as early as that done in reading
and writing. Rather, work began after the initial research done with the
IRLD with the best summary provided by Foegen, Jiban, and Deno (2007).
Applications of CBMs in Making Decisions 35

In this publication, the research reported from 32 studies is summarized in
terms of three stages (Fuchs, 2004): (a) stage 1, with an emphasis on initial
technical adequacy, particularly reliability and criterion validity evidence;
(b) stage 2, with the focus on the technical characteristics of slope (change
over time); and (c) stage 3, in which instructional utility of the measures is
addressed.
1. The dominance of research is on elementary grade levels in which 17 studies
provided evidence on the technical characteristics of the measures (stage 1)
with monitoring basic skills progress measures (Fuchs, Hamlett, & Fuchs,
1998). The studieswere split between progress measures(reﬂecting curriculum
sampling within grades) and performance measures (reﬂecting robust
measures across grades). The results conﬁrmed many of the measures being
developed as both reliable and with adequate relations to other measures of
achievement.
2. In the early mathematics measures for students in prekindergarten to ﬁrst
grade, all of the studies were published since 2000 and most of them have
relied upon performance measures. The focus has been on number
identiﬁcation, quantity discrimination, and missing numbers.
3. The least well-developed research in mathematics has been with middle-
school grades. Calhoon (2008) has most aptly phrased the problem at the
middle- and high-school level: ‘‘The needs for CBM assessments at the
secondary level are limitless’’ (p. 238). Out of 578 articles, dissertations,
and reports related to CBM, there have been only four published studies
on math progress monitoring for middle school students (Foegen et al.,
2007). Of these four, two investigated measures involving facts and
estimation (Foegen, 2000; Foegen, et al., 2007), and two focused on
measures that reﬂected a conceptual understanding of math (Helwig,
Anderson, & Tindal, 2002; Helwig & Tindal, 2002). All four studies
showed promising validity and/or reliability.
The essential features of progress monitoring assessments have changed
very little since the 1970s. Such measures are still expected to sample from
the year’s worth of curriculum. They are meant to provide teachers with
meaningful information about the progress students are making in
mastering that material. In addition, to enhance their utility, progress
monitoring measures are intended to be easy to administer, score, and
interpret. However, whereas four decades ago the researchers deemed CBMs
as not requiring any particular expertise to develop, the increasing stakes
associated with assessment results as well as advances in psychometrics
have signiﬁcantly altered one perspective: We now recognize that the
GERALD TINDAL AND JOSEPH F. T. NESE36

creation of reliable and valid progress monitoring measures requires specialized
knowledge beyond what most public school teachers possess. This realization
spurred our creation of ‘‘next generation’’ elementary school reading CBMs,
measures created using rigorous statistical modeling analytics previously the
domain of large-scale assessments.
NEXT GENERATION CBM: ALIGNMENT TO
STANDARDS, ITEM SCALING, AND TECHNICAL
ADEQUACY
Next-generation CBM needs to reﬂect the changing educational landscape
with its emphasis on accountability. We developed easyCBM to reﬂect this
high-stakes environment; in the process, we focused on the need to integrate
the decisions teachers would make so that the same data system could be
used across multiple decisions and at the same time be predictive of
performance on a large-scale summative test. Rather than make a referral
decision based on a single measure used only once for one decision (to locate
students at risk), we developed screening measures that also would be
related to formative progress measures. Not only would the measures be
appropriate for all students so teachers could understand their differences,
but they also would be useful for an individual student so teachers could
understand what instructional programs were effective and make an
individual difference (Deno, 1990).
To accomplish this outcome and relate decisions using common measures,
however, next-generation CBMs were needed that avoided three signiﬁcant
problems represented in the research just cited. The ﬁrst problem was
alignment to standards, the second was proper scaling, and the third and
last was the documentation of technical adequacy with the most recent
efforts being promulgated by the NCRTI.
Alignment to Standards
In the most recent reauthorization of the Elementary and Secondary
Education Act (ESEA), the federal government announced that a key
priority in education is for all states to develop and adopt standards in
reading and mathematics that build through high school (U.S. Department
of Education Ofﬁce of Planning Evaluation and Policy Development, 2010).
Applications of CBMs in Making Decisions 37

Academic standards stipulate and deﬁne what students must know and be
able to do at each grade level. The U.S. Department of Education also
pledged support for the development and use of assessmentsalignedto
‘‘standards in English language arts and mathematics that accurately
measure student academic achievement and growth, provide feedback to
support and improve teaching, and measure school success and progress’’
(U.S. Department of Education, 2010, p. 11).
Thus, the call has been made in education for curriculum standards, and
formative assessments aligned to those standards. Progress is meaningful
when it is directed, and CBMs that are aligned with content standards may
provide more useful information for teachers about student performance
and progress. CBM aligned to standards can put the direction and rate of
progress into context to best address individual student learning and make
informed instructional decisions.
Many states and several groups have made concerted efforts to establish
curriculum standards, notably the Common Core State Standards Initiative,
a state-led effort coordinated by the National Governors Association Center
for Best Practices (NGA Center) and the Council of Chief State School
Ofﬁcers (CCSSO). Yet to date, there are no published research studies on
the alignment of CBMs to content standards, and easyCBM is the only
CBM system that has developed assessments to be aligned to national or
state standards (Alonzo, Tindal, Ulmer, & Glasgow, 2006). The easyCBM
mathematics system was developed based on the National Council of
Teachers of Mathematics (NCTM) Curriculum Focal Points for Kinder-
garten through Grade 8 Mathematics (National Council of Teachers of
Mathematics, 2006). The NCTM Focal Points outline the areas of
mathematics instruction focus and objective for each grade level that have
been adopted by many states as the basis of their state content standards in
mathematics (e.g., Oregon). The easyCBM math items were written directly
to the NCTM focal point standards; more speciﬁcally, an item was written
to a particular objective within a single focal point standard. The results of
an alignment study indicated that the easyCBM math items consistently
and strongly aligned with the NCTM Focal Points standards (Nese et al.,
2010). These results provide content validation of the easyCBM
math assessment system, and also lend support to the feasibility and
beneﬁts of integrating standards into the process of CBM development. But
as content standards become more prevalent and the technical properties of
CBMs become more robust, the ﬁeld needs to better study and understand
the importance of alignment of CBM items and systems to content
standards.
GERALD TINDAL AND JOSEPH F. T. NESE38

Alignment with content standards also allows CBMs and high-stakes
state tests to be complementary. If CBMs and the state achievement test are
both developed to be aligned to the same content standards, the predictive
and criterion validity of the CBMs will be enhanced, effectively increasing
teachers’ instructional decisions. But the purpose of the two tests are
discrete, as high-stakes state tests aim to assess year-end content mastery,
and CBMs aim to assess academic status and growth across the year.
Because CBMs need to be sensitive to academic growth, their ceilings should
not be parallel to the knowledge proﬁciency cut-point associated with state
tests. CBMs need to offer a large sample of items with a range of difﬁculties
in order to accurately measure student growth. Thus, an effective CBM
system must develop items that are representative of the classroom or state
curriculum standards, and also items that represent a wide range of
difﬁculties.
Use of Item Response Theory for Scaling CBM
In easyCBM, the potential exists to assist teachers in moving from
individual differences to making an individual difference through analysis
of item-level data in making diagnostic decisions about what to teach and
when. As we noted in the opening review of Deno and Mirkin (1977) and
Engelmann et al. (1979), instruction is the focus of any progress monitoring
system. However, rather than simply wait for data to be compiled over time
as instructional programs are delivered, we have designed a data system that
allows teachers to target speciﬁc skills based on the data, using item
response theory (IRT).
IRT has become the theoretical foundation for most current test
development given its theoretical measurement principles and its potential
to address measurement difﬁculties (Embretson & Reise, 2000). Embretson
and Reise (2000) outlined several advantageous psychometric assessment
properties that IRT offers, including shorter, more reliable tests; equating
different test forms; providing an unbiased estimates of model parameters
that can be obtained from unrepresentative samples; obtaining scores to
meaningfully compare students even when initial status scores differ; and
providing a standard error of measurement (SEM) that can differ across
scores but generalize across populations. This last property is particularly
important because it concerns the quality of the test as well as the
interpretation of individual student scores. As a consequence, it allows the
SEM to vary across students such that the measurement error is smaller (and
Applications of CBMs in Making Decisions 39

Other documents randomly have
different content

FRANCIS JAMES CHILD
Francis James Child was born in Boston on the first day of February,
1825. He was the third in a family of eight children. His father was a
sailmaker, "one of that class of intelligent and independent
mechanics," writes Professor Norton, "which has had a large share in
determining the character of our democratic community, as of old
the same class had in Athens and in Florence." The boy attended the
public schools, as a matter of course; and, his parents having no
thought of sending him to college, he went, in due time, not to the
Latin School, but to the English High School of his native town. At
that time the head master of the Boston Latin School was Mr Epes
Sargent Dixwell, who is still living, at a ripe old age, one of the most
respected citizens of Cambridge. Mr Dixwell had a keen eye for
scholarly possibilities in boys, and, falling in with young Francis
Child, was immediately struck with his extraordinary mental ability.
At his suggestion, the boy was transferred to the Latin School, where
he entered upon the regular preparation for admission to Harvard
College. His delight in his new studies was unbounded, and the
freshness of it never faded from his memory. "He speedily caught up
with the boys who had already made considerable progress in Greek
and Latin, and soon took the first place here, as he had done in the
schools which he had previously attended." Mr Dixwell strongly
advised his father to permit him to continue his studies, and made
arrangements by which his college expenses should be provided for.
The money Professor Child repaid, with interest, as soon as his
means allowed. His gratitude to Mr. Dixwell and the friendship
between them lasted through his life.
In 1842 Mr Child entered Harvard College. The intellectual condition
of the college at that time and the undergraduate career of Mr Child
have been admirably described by his classmate and lifelong friend,
Professor Norton, in a passage which must be quoted in full
[1]
:—

"Harvard was then still a comparatively small institution, with no
claims to the title of University; but she had her traditions of good
learning as an inspiration for the studious youth, and still better she
had teachers who were examples of devotion to intellectual pursuits,
and who cared for those ends the attainment of which makes life
worth living. Josiah Quincy was approaching the close of his term of
service as President of the College, and stood before the eyes of the
students as the type of a great public servant, embodying the spirit
of patriotism, of integrity, and of fidelity in the discharge of whatever
duty he might be called to perform. Among the Professors were
Walker, Felton, Peirce, Channing, Beck, and Longfellow, men of
utmost variety of temperament, but each an instructor who secured
the respect no less than the gratitude of his pupils.
"The class to which Child belonged numbered hardly over sixty. The
prescribed course of study which was then the rule brought all the
members of the class together in recitations and lectures, and every
man soon knew the relative standing of each of his fellows. Child at
once took the lead and kept it. His excellence was not confined to
any one special branch of study; he was equally superior in all. He
was the best in the classics, he was Peirce's favorite in mathematics,
he wrote better English than any of his classmates. His intellectual
interests were wider than theirs, he was a great reader, and his
tastes in reading were mature. He read for amusement as well as for
learning, but he did not waste his time or dissipate his mental
energies over worthless or pernicious books. He made good use of
the social no less than of the intellectual opportunities which college
life affords, and became as great a favorite with his classmates as he
had been with his schoolfellows.
"The close of his college course was marked by the exceptional
distinction of his being chosen by his classmates as their Orator, and
by his having the first part at Commencement as the highest scholar
in the class. His class oration was remarkable for its maturity of
thought and of style. Its manliness of spirit, its simple directness of
presentation of the true objects of life, and of the motives by which
the educated man, whatever might be his chosen career, should be

inspired, together with the serious and eloquent earnestness with
which it was delivered, gave to his discourse peculiar impressiveness
and effect."
Graduating with the degree of Bachelor of Arts in 1846, Mr Child
immediately entered the service of the college, in which he
continued till the day of his death. From 1846 to 1848 he was tutor
in mathematics. In 1848 he was transferred, at his own request, to a
tutorship in history and political economy, to which were annexed
certain duties of instruction in English. In 1849 he obtained leave of
absence for travel and study in Europe. He remained in Europe for
about two years, returning, late in 1851, to receive an appointment
to the Boylston Professorship of Rhetoric and Oratory, then falling
vacant by the resignation of Professor Edward T. Channing.
The tutorships which Mr Child had held were not entirely in
accordance with his tastes, which had always led him in the direction
of literary and linguistic study. The faculty of the college was small,
however, and it was not always possible to assign an instructor to
the department that would have been most to his mind. But the
governors of the institution were glad to secure the services of so
promising a scholar; and Mr Child, whose preference for an
academic career was decided, had felt that it was wise to accept
such positions as the college could offer, leaving exacter adjustments
to time and circumstances. Meantime he had devoted his whole
leisure to the pursuit of his favorite studies. His first fruits were a
volume entitled Four Old Plays
[2]
published in 1848, when he was
but twenty-three years old. This was a remarkably competent
performance. The texts are edited with judgment and accuracy; the
introduction shows literary discrimination as well as sound
scholarship, and the glossary and brief notes are thoroughly good.
There are no signs of immaturity in the book, and it is still valued by
students of our early drama.
The leave of absence granted to Mr Child in 1849 came at a most
favorable moment. His health had suffered from close application to
work, and a change of climate had been advised by his physicians.

His intellectual and scholarly development, too, had reached that
stage in which foreign study and travel were certain to be most
stimulating and fruitful. He was amazingly apt, and two years of
opportunity meant much more to him than to most men. He
returned to take up the duties of his new office a trained and mature
scholar, at home in the best methods and traditions of German
universities, yet with no sacrifice of his individuality and intellectual
independence.
While in Germany Mr Child studied at Berlin and Göttingen, giving
his time mostly to Germanic philology, then cultivated with
extraordinary vigor and success. The hour was singularly propitious.
In the three or four decades preceding Mr Child's residence in
Europe, Germanic philology (in the wider sense) had passed from
the stage of "romantic" dilettantism into the condition of a well-
organized and strenuous scientific discipline, but the freshness and
vivacity of the first half of the century had not vanished. Scholars,
however severe, looked through the form and strove to comprehend
the spirit. The ideals of erudition and of a large humanity were not
even suspected of incompatibility. The imagination was still invoked
as the guide and illuminator of learning. The bond between antiquity
and mediævalism and between the Middle Ages and our own
century was never lost from sight. It was certainly fortunate for
American scholarship that at precisely this juncture a young man of
Mr Child's ardent love of learning, strong individuality, and broad
intellectual sympathies was brought into close contact with all that
was most quickening in German university life. He attended lectures
on classical antiquity and philosophy, as well as on Germanic
philology; but it was not so much by direct instruction that he
profited as by the inspiration which he derived from the spirit and
the ideals of foreign scholars, young and old. His own greatest
contribution to learning, The English and Scottish Popular Ballads,
may even, in a very real sense, be regarded as the fruit of these
years in Germany. Throughout his life he kept a picture of William
and James Grimm on the mantel over his study fireplace.

Mr Child wrote no "dissertation," and returned to Cambridge without
having attempted to secure a doctor's degree. Never eager for such
distinctions, he had been unwilling to subject himself to the
restrictions on his plan of study which candidacy for the doctorate
would have imposed. Three years after, however, in 1854, he was
surprised and gratified to receive from the University of Göttingen
the degree of Doctor of Philosophy, accompanied by a special tribute
of respect from that institution. Subsequently he received the degree
of LL. D. from Harvard (in 1884) and that of L. H. D. from Columbia
(in 1887); but the Göttingen Ph. D., coming as it did at the outset of
his career, was in a high degree auspicious.
The Boylston Professorship of Rhetoric and Oratory, to which, as has
been already mentioned, Mr Child succeeded on his return to
America toward the end of 1851, was no sinecure. In addition to
academic instruction of the ordinary kind, the duties of the chair
included the superintendence and criticism of a great quantity of
written work, in the nature of essays and set compositions prepared
by students of all degrees of ability. For twenty-five years Mr Child
performed these duties with characteristic punctuality and devotion,
though with increasing distaste for the drudgery which they
involved. Meantime a great change had come over Harvard: it had
developed from a provincial college into a national seminary of
learning, and the introduction of the "elective system"—
corresponding to the "Lernfreiheit" of Germany—had enabled it to
become a university in the proper sense of the word. One result of
the important reform just referred to was the establishment of a
Professorship of English, entirely distinct from the old chair of
Rhetoric. This took place on May 8, 1876, and on the 20th of the
next month Mr Child was transferred to the new professorship. His
duties as an instructor were now thoroughly congenial, and he
continued to perform them with unabated vigor to the end. In the
onerous details of administrative and advisory work, inseparable,
according to our exacting American system, from the position of a
university professor, he was equally faithful and untiring. For thirty
years he acted as secretary of the Library Council, and in all that

time he was absent from but three meetings. As chairman of the
Department of English and of the Division of Modern Languages, and
as a member of many important committees, he was ever prodigal
of time and effort. How steadily he attended to the regular duties of
the class-room, his pupils, for fifty years, are the best witnesses.
They, too, will best understand the satisfaction he felt that, in the
fiftieth year of his teaching, he was not absent from a single lecture.
No man was ever less a formalist; yet the most formal of natures
could not, in the strictest observance of punctilio, have surpassed
the regularity with which he discharged, as it were spontaneously,
the multifarious duties of his position.
Throughout his service as professor of rhetoric, Mr Child, hampered
though he was by the requirements of his laborious office, had
pursued with unquenchable ardor the study of the English language
and literature, particularly in their older forms, and in these subjects
he had become an authority of the first rank long before the
establishment of the English chair enabled him to arrange his
university teaching in accordance with his tastes. Soon after he
returned from Germany he undertook the general editorial
supervision of a series of the 'British Poets,' published at Boston in
1853 and several following years, and extending to some hundred
and fifty volumes. Out of this grew, in one way or another, his three
most important contributions to learning: his edition of Spenser, his
Observations on the Language of Chaucer and Gower, and his
English and Scottish Popular Ballads.
Mr Child's Spenser appeared in 1855.
[3]
Originally intended, as he
says in the preface, as little more than a reprint of the edition
published in 1839 under the superintendence of Mr George Hillard,
the book grew upon his hands until it had become something quite
different from its predecessor. Securing access to old copies of most
of Spenser's poems, Mr Child subjected the text to a careful revision,
which left little to be done in this regard. His Life of Spenser was far
better than any previous biography, and his notes, though brief,
were marked by a philological exactness to which former editions

could not pretend. Altogether, though meant for the general reader
and therefore sparingly annotated, Mr Child's volumes remain, after
forty years, the best edition of Spenser in existence.
The plan of the 'British Poets' originally contemplated an edition of
Chaucer, which Mr Child was to prepare. Becoming convinced,
however, that the time was not ripe for such a work, he abandoned
this project, and to the end of his life he never found time to resume
it. Thomas Wright's print of the Canterbury Tales
[4]
from the
Harleian MS. 7334 had, however, put into his hands a reasonably
faithful reproduction of an old text, and he turned his attention to a
minute study of Chaucer's language. The outcome was the
publication, in the Memoirs of the American Academy of Arts and
Sciences for 1863, of the great treatise to which Mr Child gave the
modest title of Observations on the Language of Chaucer's
Canterbury Tales. It is difficult, at the present day, to imagine the
state of Chaucer philology at the moment when this paper appeared.
Scarcely anything, we may say, was known of Chaucer's grammar
and metre in a sure and scientific way. Indeed, the difficulties to be
solved had not even been clearly formulated. Further, the accessible
mass of evidence on Anglo-Saxon and Middle English was, in
comparison with the stores now at the easy command of every tyro,
almost insignificant. Yet, in this brief treatise, Mr Child not only
defined the problems, but provided for most of them a solution
which the researches of younger scholars have only served to
substantiate. He also gave a perfect model of the method proper to
such inquiries—a method simple, laborious, and exact. The
Observations were subsequently rearranged and condensed, with
Professor Child's permission, by Mr A. J. Ellis for his work On Early
English Pronunciation; but only those who have studied them in their
original form can appreciate their merit fully. "It ought never to be
forgotten," writes Professor Skeat, "that the only full and almost
complete solution of the question of the right scansion of Chaucer's
Canterbury Tales is due to what Mr Ellis rightly terms 'the wonderful
industry, acuteness, and accuracy of Professor Child.'" Had he
produced nothing else, this work, with its pendant, the Observations

on Gower,
[5]
would have assured him a high place among those very
few scholars who have permanently settled important problems of
linguistic science.
Mr Child's crowning work, however, was the edition of the English
and Scottish Popular Ballads, which the reader now has before him.
The history of this is the history of more than half a lifetime.
The idea of the present work grew out of Mr Child's editorial labors
on the series of the 'British Poets,' already referred to. For this he
prepared a collection in eight small volumes (1857-58) called English
and Scottish Ballads.
[6]
This was marked by the beginnings of that
method of comparative study which is carried out to its ultimate
issues in the volumes of the present collection. The book circulated
widely, and was at once admitted to supersede all previous attempts
in the same field. To Mr Child, however, it was but the starting-point
for further researches. He soon formed the plan of a much more
extensive collection on an altogether different model. This was to
include every obtainable version of every extant English or Scottish
ballad, with the fullest possible discussion of related songs or stories
in the "popular" literature of all nations. To this enterprise he
resolved, if need were, to devote the rest of his life. His first care
was to secure trustworthy texts. In his earlier collection he had been
forced to depend almost entirely on printed books. No progress, he
was convinced, could be made till recourse could be had to
manuscripts, and in particular to the Percy MS. Accordingly he
directed his most earnest efforts to securing the publication of the
entire contents of the famous folio. The Percy MS. was at Ecton Hall,
in the possession of the Bishop's descendants, who would permit no
one even to examine it. Two attempts were made by Dr Furnivall, at
Mr Child's instance, to induce the owners to allow the manuscript to
be printed,—one as early as 1860 or 1861, the other in 1864,—but
without avail. A third attempt was more successful, and in 1867-68
the long-secluded folio was made the common property of scholars
in an edition prepared by Professor Hales and Dr Furnivall.
[7]

The publication of the Percy MS. not only put a large amount of
trustworthy material at the disposal of Mr Child; it exposed the full
enormity of Bishop Percy's sins against popular tradition. Some
shadow of suspicion inevitably fell on all other ballad collections. It
was more than ever clear to Mr Child that he could not safely take
anything at second hand, and he determined not to print a line of
his projected work till he had exhausted every effort to get hold of
whatever manuscript material might be in existence. His efforts in
this direction continued through many years. A number of
manuscripts were in private hands; of some the whereabouts was
not known; of others the existence was not suspected. But Mr Child
was untiring. He was cordially assisted by various scholars,
antiquaries, and private gentlemen, to whose coöperation ample
testimony is borne in the Advertisements prefixed to the volumes in
the present work. Some manuscripts were secured for the Library of
Harvard University—notably Bishop Percy's Papers, the Kinloch MSS,
and the Harris MS.,
[8]
—and of others careful copies were made,
which became the property of the same library. In all these
operations the indispensable good offices of Mr William Macmath, of
Edinburgh, deserve particular mention. For a long series of years his
services were always at Mr Child's disposal. His self-sacrifice and
generosity appear to have been equalled only by his perseverance
and wonderful accuracy. But for him the manuscript basis of The
English and Scottish Popular Ballads would have been far less strong
than it is.
Gradually, then, the manuscript materials came in, until at last, in
1882, Mr Child felt justified in beginning to print. Other important
documents were, however, discovered or made accessible as time
went on. Especially noteworthy was the great find at Abbotsford
(see the Advertisement to Part VIII). In 1877 Dr David Laing
procured, "not without difficulty," leave to prepare for Mr Child a
copy of the single manuscript of ballads then known to remain in the
library at Abbotsford. This MS., entitled "Scottish Songs," was so
inconsiderable, in proportion to the accumulations which Sir Walter
Scott had made in preparing his Border Minstrelsy, that further

search seemed to be imperatively necessary. In 1890 permission to
make such a search, and to use the results, was given by the
Honorable Mrs Maxwell-Scott. The investigation, made by Mr
Macmath, yielded a rich harvest of ballads, which were utilized in
Parts VII-IX. To dwell upon the details would be endless. The reader
may see a list of the manuscript sources at pp. 397 ff. of the fifth
volume; and, if he will observe how scattered they were, he will
have no difficulty in believing that it required years, labor, and much
delicate negotiation to bring them all together. One manuscript
remained undiscoverable, William Tytler's Brown MS., but there is no
reason to believe that this contained anything of consequence that is
not otherwise known.
[9]
Meanwhile, concurrently with the toil of amassing, collating, and
arranging texts, went on the far more arduous labor of comparative
study of the ballads of all nations; for, in accordance with Mr Child's
plan it was requisite to determine, in the fullest manner, the history
and foreign relations of every piece included in his collection. To this
end he devoted much time and unwearied diligence to forming, in
the Library of the University, a special collection of "Folk-lore,"
particularly of ballads, romances, and Märchen. This priceless
collection, the formation of which must be looked on as one of Mr
Child's most striking services to the university, numbers some 7000
volumes. But these figures by no means represent the richness of
the Library in the departments concerned, or the services of Mr Child
in this particular. Mediæval literature in all its phases was his
province, and thousands of volumes classified in other departments
of the University Library bear testimony to his vigilance in ordering
books, and his astonishing bibliographical knowledge. Very few
books are cited in the present collection which are not to be found
on the shelves of this Library.
In addition, Mr Child made an effort to stimulate the collection of
such remains of the traditional ballad as still live on the lips of the
people in this country and in the British Islands. The harvest was, in
his opinion, rather scanty; yet, if all the versions thus recovered from

tradition were enumerated, the number would not be found
inconsiderable. Enough was done, at all events, to make it clear that
little or nothing of value remains to be recovered in this way.
To readers familiar with such studies, no comment is necessary, and
to those who are unfamiliar with them, no form of statement can
convey even a faint impression of the industry, the learning, the
acumen, and the literary skill which these processes required. In
writing the history of a single ballad, Mr Child was sometimes forced
to examine hundreds of books in perhaps a dozen different
languages. But his industry was unflagging, his sagacity was scarcely
ever at fault, and his linguistic and literary knowledge seemed to
have no bounds. He spared no pains to perfect his work in every
detail, and his success was commensurate with his efforts. In the
Advertisement to the Ninth Part (1894), he was able to report that
the three hundred and five numbers of his collection comprised the
whole extant mass of this traditional material, with the possible
exception of a single ballad.
[10]
In June, 1896, Mr Child concluded his fiftieth year of service as a
teacher in Harvard College. He was at this time hard at work on the
Tenth and final Part, which was to contain a glossary, various
indexes, a bibliography, and an elaborate introduction on the general
subject. For years he had allowed himself scarcely any respite from
work, and, in spite of the uncertain condition of his health,—or
perhaps in consequence of it,—he continued to work at high
pressure throughout the summer. At the end of August he
discovered that he was seriously ill. He died at Boston on the 11th
day of September. He had finished his great work except for the
introduction and the general bibliography. The bibliography was in
preparation by another hand and has since been completed. The
introduction, however, no other scholar had the hardihood to
undertake. A few pages of manuscript,—the last thing written by his
pen,—almost illegible, were found among his papers to show that he
had actually begun the composition of this essay, and many sheets
of excerpts testified to the time he had spent in refreshing his

memory as to the opinions of his predecessors, but he had left no
collectanea that could be utilized in supplying the Introduction itself.
He was accustomed to carry much of his material in his memory till
the moment of composition arrived, and this habit accounts for the
fact that there are no jottings of opinions and no sketch of precisely
what line of argument he intended to take.
Mr Child's sudden death was felt as a bitter personal loss, not only
by an unusually large circle of attached friends in both hemispheres,
but by very many scholars who knew him through his works alone.
He was one of the few learned men to whom the old title of
"Master" was justly due and freely accorded. With astonishing
erudition, which nothing seemed to have escaped, he united an
infectious enthusiasm and a power of lucid and fruitful exposition
that made him one of the greatest of teachers, and a warmth and
openness of heart that won the affection of all who knew him. In
most men, however complex their characters, one can distinguish
the qualities of the heart, in some degree, from the qualities of the
head. In Professor Child no such distinction was possible, for all the
elements of his many-sided nature were fused in his marked and
powerful individuality. In his case, the scholar and the man cannot
be separated. His life and his learning were one; his work was the
expression of himself.
As an investigator Professor Child was at once the inspiration and
the despair of his disciples. Nothing could surpass the scientific
exactness of his methods and the unwearied diligence with which he
conducted his researches. No possible source of information could
elude him; no book or manuscript was too voluminous or too
unpromising for him to examine on the chance of its containing
some fact that might correct or supplement his material, even in the
minutest point. Yet these qualities of enthusiastic accuracy and
thoroughness, admirable as they undoubtedly were, by no means
dominated him. They were always at the command of the higher
qualities of his genius,—sagacity, acumen, and a kind of sympathetic
and imaginative power in which he stood almost alone among recent
scholars. No detail of language or tradition or archæology was to

him a mere lifeless fact; it was transmuted into something vital, and
became a part of that universal humanity which always moved him
wherever he found it, whether in the pages of a mediæval chronicle,
or in the stammering accents of a late and vulgarly distorted ballad,
or in the faces of the street boys who begged roses from his garden.
No man ever felt a keener interest in his kind, and no scholar ever
brought this interest into more vivifying contact with the
technicalities of his special studies. The exuberance of this large
humanity pervades his edition of the English and Scottish ballads.
Even in his last years, when the languor of uncertain health
sometimes got the better, for a season, of the spirit with which he
commonly worked, some fresh bit of genuine poetry in a ballad,
some fine trait of pure nature in a stray folk-tale, would, in an
instant, bring back the full flush of that enthusiasm which he must
have felt when the possibilities of his achievement first presented
themselves to his mind in early manhood. For such a nature there
was no old age.
From this ready sympathy came that rare faculty—seldom possessed
by scholars—which made Professor Child peculiarly fit for his
greatest task. Few persons understand the difficulties of ballad
investigation. In no field of literature have the forger and the
manipulator worked with greater vigor and success. From Percy's
day to our own it has been thought an innocent device to publish a
bit of one's own versifying, now and then, as an "old ballad" or an
"ancient song." Often, too, a late stall-copy of a ballad, getting into
oral circulation, has been innocently furnished to collectors as
traditional matter. Mere learning will not guide an editor through
these perplexities. What is needed is, in addition, a complete
understanding of the "popular" genius, a sympathetic recognition of
the traits that characterize oral literature wherever and in whatever
degree they exist. This faculty, which even the folk has not retained,
and which collectors living in ballad-singing and tale-telling times
have often failed to acquire, was vouchsafed by nature herself to
this sedentary scholar. In reality a kind of instinct, it had been so
cultivated by long and loving study of the traditional literature of all

nations that it had become wonderfully swift in its operations and
almost infallible. A forged or retouched piece could not deceive him
for a moment; he detected the slightest jar in the genuine ballad
tone. He speaks in one place of certain writers "who would have
been all the better historians for a little reading of romances." He
was himself the better interpreter of the poetry of art for this keen
sympathy with the poetry of nature.
Constant association with the spirit of the folk did its part in
maintaining, under the stress of unremitting study and research,
that freshness and buoyancy of mind which was the wonder of all
who met Professor Child for the first time, and the perpetual delight
of his friends and associates. It is impossible to describe the charm
of his familiar conversation. There was endless variety without effort.
His peculiar humor, taking shape in a thousand felicities of thought
and phrase that fell casually and as it were inevitably from his lips,
exhilarated without reaction or fatigue. His lightest words were full
of fruitful suggestion. Sudden strains of melancholy or high
seriousness were followed, in a moment, by flashes of gaiety almost
boyish. And pervading it all one felt the attraction of his personality
and the goodness of his heart.
Professor Child's humor was not only one of his most striking
characteristics as a man; it was of constant service to his scholarly
researches. Keenly alive to any incongruity in thought or fact, and
the least self-conscious of men, he scrutinized his own nascent
theories with the same humorous shrewdness with which he looked
at the ideas of others. It is impossible to think of him as the sponsor
of some hypotheses which men of equal eminence have advanced
and defended with passion; and, even if his goodness of nature had
not prevented it, his sense of the ridiculous would not have suffered
him to engage in the absurdities of philological polemics. In the
interpretation of literature, his humor stood him in good stead,
keeping his native sensibility under due control, so that it never
degenerated into sentimentalism. It made him a marvelous
interpreter of Chaucer, whose spirit he had caught to a degree
attained by no other scholar or critic.

To younger scholars Professor Child was an influence at once
stimulating and benignant. To confer with him was always to be
stirred to greater effort, but, at the same time, the serenity of his
devotion to learning chastened the petulance of immature ambition
in others. The talk might be quite concrete, even definitely practical,
—it might deal with indifferent matters; but, in some way, there was
an irradiation of the master's nature that dispelled all unworthy
feelings. In the presence of his noble modesty the bustle of self-
assertion was quieted and the petty spirit of pedantic wrangling
could not assert itself. However severe his criticism, there were no
personalities in it. He could not be other than outspoken,—
concealment and shuffling were abhorrent to him,—yet such was his
kindliness that his frankest judgments never wounded; even his
reproofs left no sting. With his large charity was associated, as its
necessary complement in a strong character, a capacity for righteous
indignation. "He is almost the only man I know," said one in his
lifetime, "who thinks no evil." There could be no truer word. Yet
when he was confronted with injury or oppression, none could stand
against the anger of this just man. His unselfishness did not suffer
him to see offences against himself, but wrong done to another
roused him in an instant to protesting action.
Professor Child's publications, despite their magnitude and
importance, are no adequate measure either of his acquirements or
of his influence. He printed nothing about Shakspere, for example,
yet he was the peer of any Shaksperian, past or present, in
knowledge and interpretative power. As a Chaucer scholar he had no
superior, in this country or in Europe: his published work was
confined, as we have seen, to questions of language, but no one
had a wider or closer acquaintance with the whole subject. An
edition of Chaucer from his hand would have been priceless. His
acquaintance with letters was not confined to special authors or
centuries. He was at home in modern European literature and
profoundly versed in that of the Middle Ages. In his immediate
territory,—English,—his knowledge, linguistic and literary, covered all
periods, and was alike exact and thorough. His taste and judgment

were exquisite, and he enlightened every subject which he touched.
As a writer, he was master of a singularly felicitous style, full of
individuality and charm. Had his time not been occupied in other
ways, he would have made the most delightful of essayists.
Fortunately, Professor Child's courses of instruction in the university
—particularly those on Chaucer and Shakspere—gave him an
opportunity to impart to a constantly increasing circle of pupils the
choicest fruits of his life of thought and study. In his later years he
had the satisfaction to see grow up about him a school of young
specialists who can have no higher ambition than to be worthy of
their master. But his teaching was not limited to these,—it included
all sorts and conditions of college students; and none, not even the
idle and incompetent, could fail to catch something of his spirit. One
thing may be safely asserted: no university teacher was ever more
beloved.
And with this may fitly close too slight a tribute to the memory of a
great scholar and a good man. Many things remain unsaid. His
gracious family life, his civic virtues, his patriotism, his bounty to the
poor,—all must be passed by with a bare mention, which yet will
signify much to those who knew him. In all ways he lived worthily,
and he died having attained worthy ends.
G. L. Kittredge .

FOOTNOTES:
[1] C. E. Norton, 'Francis James Child,' in the Proceedings of the
American Academy of Arts and Sciences, XXXII, 334, 335;
reprinted, with some additions, in the Harvard Graduates'
Magazine, VI, 161-169 (Boston, 1897). I have used this
biographical sketch freely in my brief account of Professor Child's
boyhood.
[2] Four Old Plays | Three Interludes: Thersytes Jack Jugler | and
Heywoods Pardoner and Frere: | and Jocasta a Tragedy | by
Gascoigne and | Kinwelmarsh | with an | Introduction and Notes |
Cambridge | George Nichols | MDCCCXLVIII. The editor's name
does not appear in the title-page, but the Preface is signed with
the initials F. J. C. Jocasta was printed from Steevens's copy of
the first edition of Garcoigne's Posies, which had come into Mr
Child's possession.
[3] The Poetical Works of Edmund Spenser. The text carefully
revised, and illustrated with notes, original and selected, by
Francis J. Child. Boston: Little, Brown, and Company. 1855. 5
vols.
[4] The Canterbury Tales of Geoffrey Chaucer. A new text, with
illustrative notes. Edited by Thomas Wright. London, printed for
the Percy Society, 1847-51. 3 vols.
[5] The paper entitled Observations on the Language of Chaucer
was laid before the American Academy of Arts and Sciences on
June 3, 1862, and was published in the Memoirs of the Academy,
Vol. VIII, pt. ii, 445-502 (Boston, 1863). The second paper,
entitled Observations on the Language of Gower's Confessio
Amantis, was laid before the Academy on January 9, 1866, and
appeared in Memoirs, IX, ii, 265-315 (Boston, 1873). A few copies
of each paper were struck off separately, but these are now very
hard to find. Mr Ellis's rearrangement and amalgamation of the
two papers, which is by no means a good substitute for the
papers themselves, may be found in Part I of his Early English
Pronunciation, London, 1869, pp. 343-97.
[6] English and Scottish Ballads. Selected and edited by Francis
James Child. Boston, 1857-58.

[7] How inseparable were the services of Dr Furnivall and those
of Professor Child in securing this devoutly wished consummation
may be seen by comparing Dr Furnivall's Forewords (I, ix, x), in
which he gives much of the credit to Mr Child, with Mr Child's
Dedication (in vol. I of the present collection), in which he gives
the credit to Dr Furnivall.
[8] Since Mr Child's death the important "Buchan original MS" has
been secured for the Child Memorial Library of the University,—a
collection endowed by friends and pupils of the dead master.
[9] See V, 397 b.
[10] This is 'Young Betrice,' No 5 in William Tytler's lost Brown
MS. (V, 397), which "may possibly be a version of 'Hugh Spencer's
Feats in France'" (see II, 377; III, 275).

1
RIDDLES WISELY EXPOUNDED
A. a. 'A Noble Riddle Wisely Expounded; or, The Maid's
Answer to the Knight's Three Questions,' 4to, Rawlinson,
566, fol. 193, Bodleian Library; Wood, E. 25, fol. 15, Bod.
Lib. b. Pepys, III, 19, No 17, Magdalen College,
Cambridge. c. Douce, II, fol. 168 b, Bod. Lib. d. 'A Riddle
Wittily Expounded,' Pills to Purge Melancholy, IV, 129, ed.
1719. "II, 129, ed. 1712."
B. 'The Three Sisters.' Some Ancient Christmas Carols ...
together with two Ancient Ballads, etc. By Davies Gilbert,
2d ed., p. 65.
C. 'The Unco Knicht's Wowing,' Motherwell's MS., p. 647.
D. Motherwell's MS., p. 142.
The four copies of A differ but very slightly: a, b, c are broadsides,
and d is evidently of that derivation, a and b are of the 17th
century. There is another broadside in the Euing collection, formerly
Halliwell's, No 253. The version in The Borderer's Table Book, VII,
83, was compounded by Dixon from others previously printed.
Riddles, as is well known, play an important part in popular story,
and that from very remote times. No one needs to be reminded of
Samson, Œdipus, Apollonius of Tyre. Riddle-tales, which, if not so
old as the oldest of these, may be carried in all likelihood some
centuries beyond our era, still live in Asiatic and European tradition,
and have their representatives in popular ballads. The largest class
of these tales is that in which one party has to guess another's
riddles, or two rivals compete in giving or guessing, under penalty in
either instance of forfeiting life or some other heavy wager; an

example of which is the English ballad, modern in form, of 'King
John and the Abbot of Canterbury.' In a second class, a suitor can
win a lady's hand only by guessing riddles, as in our 'Captain
Wedderburn's Courtship' and 'Proud Lady Margaret.' There is
sometimes a penalty of loss of life for the unsuccessful, but not in
these ballads. Thirdly, there is the tale (perhaps an offshoot of an
early form of the first) of The Clever Lass, who wins a husband, and
sometimes a crown, by guessing riddles, solving difficult but
practicable problems, or matching and evading impossibilities; and of
this class versions A and B of the present ballad and A-H of the
following are specimens.
Ballads like our 1, A, B, 2, A-H, are very common in German. Of
the former variety are the following:
A. 'Räthsellied,' Büsching, Wöchentliche Nachrichten, I, 65, from the
neighborhood of Stuttgart. The same, Erlach, III, 37; Wunderhorn,
IV, 139; Liederhort, p. 338, No 153; Erk u. Irmer, H. 5, p. 32, No 29;
Mittler, No 1307 (omits the last stanza); Zuccalmaglio, II, 574, No
317 [with change in st. 11]. A knight meets a maid on the road,
dismounts, and says, "I will ask you a riddle; if you guess it, you
shall be my wife." She answers, "Your riddle shall soon be guessed; I
will do my best to be your wife;" guesses eight pairs of riddles, is
taken up behind him, and they ride off. B. 'Räthsel um Räthsel,'
Wunderhorn, II, 407 [429, 418] == Erlach, I, 439. Zuccalmaglio, II,
572, No 316, rearranges, but adds nothing. Mittler, No 1306, inserts
three stanzas (7, 9, 10). This version begins: "Maid, I will give you
some riddles, and if you guess them will marry you." There are
seven pairs, and, these guessed, the man says, "I can't give you
riddles; let's marry;" to which she gives no coy assent: but this
conclusion is said not to be genuine (Liederhort, p. 341, note). C.
'Räthsellied,' Erk, Neue Sammlung, Heft 3, p. 64, No 57, and
Liederhort, 340, No 153
a
two Brandenburg versions, nearly agreeing,
one with six, the other with five, pairs of riddles. A proper conclusion
not having been obtained, the former was completed by the two last
stanzas of B, which are suspicious. C begins like B. D.
'Räthselfragen,' Peter, Volksthümliches aus Österreichisch-Schlesien,

I, 272, No 83. A knight rides by where two maids are sitting, one of
whom salutes him, the other not. He says to the former, "I will put
you three questions, and if you can answer them will marry you." He
asks three, then six more, then three, and then two, and, all being
answered, bids her, since she is so witty, build a house on a needle's
point, and put in as many windows as there are stars in the sky;
which she parries with, "When all streams flow together, and all
trees shall fruit, and all thorns bear roses, then come for your
answer." E. 'Räthsellied,' Tschischka u. Schottky, Oesterreichische
Volkslieder, 2d ed., p. 28, begins like B, C, has only three pairs of
riddles, and ends with the same task of building a house on a
needle's point. F. 'Räthsellied,' Hocker, Volkslieder von der Mosel, in
Wolf's Zeits. für deutsche Myth., I, 251, from Trier, begins with the
usual promise, has five pairs of riddles, and no conclusion. G.
'Räthsel,' Ditfurth, Fränkische V. L., II, 110, No 146, has the same
beginning, six pairs of riddles, and no conclusion.
Some of the riddles occur in nearly all the versions, some in only one
or two, and there is now and then a variation also in the answers.
Those which are most frequent are:
Which is the maid without a tress? A-D, G.
And which is the tower without a crest? A-D, F, G.
(Maid-child in the cradle; tower of Babel.)
Which is the water without any sand? A, B, C, F, G.
And which is the king without any land? A, B, C, F, G.
(Water in the eyes; king in cards.)
Where is no dust in all the road? A-G.
Where is no leaf in all the wood? A-G.
(The milky way, or a river; a fir-wood.)
Which is the fire that never burnt? A, C-G.
And which is the sword without a point? C-G.
(A painted fire; a broken sword.)
Which is the house without a mouse? C-G.
Which is the beggar without a louse? C-G.
(A snail's house; a painted beggar.)
[11]

A ballad translated in Ralston's Songs of the Russian People, p. 356,
from Buslaef's Historical Sketches of National Literature and Art, I,
31, resembles very closely German A. A merchant's son drives by a
garden where a girl is gathering flowers. He salutes her; she returns
her thanks. Then the ballad proceeds:
'Shall I ask thee riddles, beauteous maiden?
Six wise riddles shall I ask thee?'
'Ask them, ask them, merchant's son,
Prithee ask the six wise riddles.'
'Well then, maiden, what is higher than the forest?
Also, what is brighter than the light?
Also, maiden, what is thicker than the forest?
Also, maiden, what is there that's rootless?
Also, maiden, what is never silent?
Also, what is there past finding out?'
'I will answer, merchant's son, will answer,
All the six wise riddles will I answer.
Higher than the forest is the moon;
Brighter than the light the ruddy sun;
Thicker than the forest are the stars;
Rootless is, O merchant's son, a stone;
Never silent, merchant's son, the sea;
And God's will is past all finding out.'
'Thou hast guessed, O maiden fair, guessed rightly,
All the six wise riddles hast thou answered;
Therefore now to me shalt thou be wedded,
Therefore, maiden, shalt thou be the merchant's
wife.'
[12]
Among the Gaels, both Scotch and Irish, a ballad of the same
description is extremely well known. Apparently only the questions
are preserved in verse, and the connection with the story made by a
prose comment. Of these questions there is an Irish form, dated
1738, which purports to be copied from a manuscript of the twelfth
century. Fionn would marry no lady whom he could pose. Graidhne,

"daughter of the king of the fifth of Ullin," answered everything he
asked, and became his wife. Altogether there are thirty-two
questions in the several versions. Among them are: What is blacker
than the raven? (There is death.) What is whiter than the snow?
(There is the truth.) 'Fionn's Questions,' Campbell's Popular Tales of
the West Highlands, III, 36; 'Fionn's Conversation with Ailbhe,'Heroic
Gaelic Ballads, by the same, pp. 150, 151.
The familiar ballad-knight of A, B is converted in C into an "unco
knicht," who is the devil, a departure from the proper story which is
found also in 2 J. The conclusion of C,
As soon as she the fiend did name,
He flew awa in a blazing flame,
reminds us of the behavior of trolls and nixes under like
circumstances, but here the naming amounts to a detection of the
Unco Knicht's quiddity, acts as an exorcism, and simply obliges the
fiend to go off in his real character. D belongs with C: it was given
by the reciter as a colloquy between the devil and a maiden.
The earlier affinities of this ballad can be better shown in connection
with No 2.
Translated, after B and A, in Grundtvig's Engelske og skotske
Folkeviser, p. 181: Herder, Volkslieder, I, 95, after A d.
A
a. Broadside in the Rawlinson collection, 4to, 566, fol. 193, Wood, E.
25, fol. 15. b. Pepys, III, 19, No 17. c. Douce, II, fol. 168 b. d. Pills
to Purge Melancholy, IV, 130, ed. 1719.
1
There was a lady of the North Country,

Lay the bent to the bonny broom
And she had lovely daughters three.
Fa la la la, fa la la la ra re
2
There was a knight of noble worth
Which also lived in the North.
3
The knight, of courage stout and brave,
A wife he did desire to have.
4
He knocked at the ladie's gate
One evening when it was late.
5
The eldest sister let him in,
And pin'd the door with a silver pin.
6
The second sister she made his bed,
And laid soft pillows under his head.
7
The youngest daughter that same night,
She went to bed to this young knight.
8
And in the morning, when it was day,
These words unto him she did say:
9
'Now you have had your will,' quoth she,
'I pray, sir knight, will you marry me?'
10

The young brave knight to her replyed,
'Thy suit, fair maid, shall not be deny'd.
11
'If thou canst answer me questions three,
This very day will I marry thee.'
12
'Kind sir, in love, O then,' quoth she,
'Tell me what your [three] questions be.'
13
'O what is longer than the way,
Or what is deeper than the sea?
14
'Or what is louder than the horn,
Or what is sharper than a thorn?
15
'Or what is greener than the grass,
Or what is worse then a woman was?'
16
'O love is longer than the way,
And hell is deeper than the sea.
17
'And thunder is louder than the horn,
And hunger is sharper than a thorn.
18
'And poyson is greener than the grass,
And the Devil is worse than woman was.'
19
When she these questions answered had,

Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com

Assessment And Intervention Advances In Learning And Behavioral Disabilities Thomas E Scruggs

About This Presentation

Slide Content

Tags

Categories

Download

Quick Actions

Statistics

Related Slideshows

Assessment And Intervention Advances In Learning And Behavioral Disabilities Thomas E Scruggs

About This Presentation

Slide Content

Slide 1

Slide 2

Slide 3

Slide 6

Slide 7

Slide 8

Slide 9

Slide 10

Slide 11

Slide 12

Slide 13

Slide 14

Slide 15

Slide 16

Slide 17

Slide 18

Slide 19

Slide 20

Slide 21

Slide 22

Slide 23

Slide 24

Slide 25

Slide 26

Slide 27

Slide 28

Slide 29

Slide 30

Slide 31

Slide 32

Slide 33

Slide 34

Slide 35

Slide 36

Slide 37

Slide 38

Slide 39

Slide 40

Slide 41

Slide 42

Slide 43

Slide 44

Slide 45

Slide 46

Slide 47

Slide 48

Slide 49

Slide 50

Slide 51

Slide 52

Slide 53

Slide 54

Slide 55

Slide 57

Slide 58

Slide 59

Slide 60

Slide 61

Slide 62

Slide 63

Slide 64

Slide 65

Slide 66

Slide 67

Slide 68

Slide 69

Slide 70

Slide 71

Slide 72

Slide 73

Slide 74

Slide 75

Slide 76

Slide 77

Slide 78

Slide 79

Slide 80