2. AUTHORS
Joetta Beaver
With a bachelor of science in elementary education and master's degree in
reading from The Ohio State University, Joetta Beaver worked as an
elementary teacher (K-5) for 30 years, as well as a K-5 Language
Arts/Assessment coordinator and an Early Education teacher-leader. She is
the primary author of DRA2 K-3, co-author of DRA2 4-8 and Developing
Writer's Assessment (DWA), consultant, and speaker.
Mark Carter, PhD
With assessment the focus of much of his professional work, Mark Carter
served as a coordinator of assessment for Upper Arlington Schools (where
he currently teachers fifth grade), conducted numerous seminars, and co-
authored DRA2 4-8, DWA, and Portfolio Assessment in the Reading
Classroom. He received his doctorate of philosophy from The Ohio State
University where he also taught graduate courses in education as an
adjunct professor.
3. OVERVIEW OF THE DRA
The Developmental Reading Assessment is a set of
individually administered criterion-referenced assessments
K-8.
Purpose: Identify students’ reading level based on
accuracy, fluency, and comprehension
Other purposes- Identify students’ strength and
weaknesses at their independent reading level, planning
instruction, monitor reading growth, and preparation for
testing expectations.
Assessment is administered one-on-one requiring students
to read specifically selected leveled assessment texts that
increase with difficulty.
Administered, scored, and interpreted by classroom
teachers.
4. DRA HISTORY & REVISIONS
1988-1997- DRA is researched and developed by Joetta Beaver and
the Upper Arlington School District
1997- DRA K-3 is published by Pearson
1999- Evaluation of the Development of Reading
2002- DRA 4-8
2004- DRA Word Analysis
2005 SRA Second Edition (DRA2), K-3 & 4-8
2006- Evaluation of the Development of Reading
2007- More than 250,000 classrooms use DRA and EDL
2008- Pearson partners with Liberty Source on DRA2 Handheld
Tango Edition
2009- DRA2 Handheld- Tango wins CODIE Award
6. ORAL READING AND FLUENCY
The total number of oral reading errors is
converted to an accuracy score using a
words-per minute (WCPM) metric.
Rating expression, phrasing, rate, and
accuracy on a 4 point scale. This begins
at level 14- the transitional level, grades
1 and 2.
7. COMPREHENSION
Levels 3-16, once the oral reading is over, the student
should take the book and read it again silently. This gives
them another opportunity to check themselves on
comprehension for retelling. Students retell what happens
in the story.
Underline information that the student is able to give, but
which requires prompting.
Note information that the student is able to give, but which
requires prompting, with a TP (teacher prompt).
Follow-up questions follow the summary and if used need
to be tallied to the left. The number of prompts to elicit more
information will be calculated as part of the comprehension
score.
8. WORD ANALYSIS
Assesses phonological awareness,
metalanguage, letter/word recognition,
phonics, and structural analysis in
grades K-3. DRA Word Analysis is
included in the new second edition of
DRA K-3.
12. KORETZ ON VALIDITY
“…Validity, which is the single most important
criterion for evaluating achievement testing.
..but, tests themselves are not valid or invalid.
Rather, it is an inference based on test scores
that is valid or invalid. ..Validity is also a
continuum: inferences are rarely perfect. The
question to ask is how well supported the
conclusion is” (Koretz, 2008, p. 31).
13. VALIDITY CONT.
Messick 1994 would argue that construct validity refers
to the inferences that are drawn about score meaning,
specifically the score interpretation and the
implications for test use (quantitative). This theoretical
framework becomes subject to empirical challenges, a
unified approach of validity
What is the test measuring?
Can it measure what it intends to measure?
14. FOUR TYPES OF VALIDATION
Predictive Concurrent
Criterion-oriented Criterion-Oriented
Validity
Content Construct
15. CRITERION VALIDITY
Predictive validity is were we draw an
inference from test scores to performance.
Concurrent Validity- studied when a test is
proposed a substitute for another, or a test is
shown to correlate with some contemporary
criterion (Cronbach & Meehl, 1955).
16. CONTENT VALIDITY
According to Yu, content validity is when we
draw
inferences from test scores to a larger domain of
items similar to those on the test, sample
population.
This selection of content is usually done by
experts.
Experts tend to lack experience in the field, and
assume that all are experts.
17. CONSTRUCT VALIDITY
According to Hunter and Schmidt (1990), construct validity is a
quantitative question rather than a qualitative distinction such as
"valid" or
"invalid"; it is a matter of degree. Construct validity can be measured
by
the correlation between the intended independent variable
(construct)
and the proxy independent variable (indicator, sign) that is actually
used.
-Yu
18. PEARSON EDUCATION ON
DRA VALIDITY
Pearson refers “…to validity of an assessment,
one looks at the extent to which the assessment
actually measures what it is supposed to
measure.” Questions to be asked when examining
validity include:
Does this assessment truly measure reading
ability?
Can teachers make accurate inferences about
the true reading ability of a student based upon
DRA2 assessment results?
19. PEARSON EDUCATION ON CONTENT
RELATED VALIDITY OF THE DRA
The content validity of a test relates to the adequacy
with which the content is covered in the test.
A “Theoretical Framework and Research,” the DRA2
incorporates reading domains to review and research
good readers with consultants and educators.
Content Validity was built into the DRA and DRA2
assessments during the development process.
20. PEARSON CRITERION RELATED
VALIDITY ON THE DRA
Criterion-related validity refers to the extent to which a
measure predicts performance on some other significant
measures, (called a criterion) other than the test itself.
Criterion validity may be broken down into two components:
concurrent and predictive.
Concurrent validity correlates the DRA to many other
reading tests:
Gray’s Oral Reading Test-4th Edition
GORT-4; Wiederholt & Bryant, 2001
DIBELS Oral Reading Fluency Test-6th Edition
Correlations Between DRA2 and Teacher Ratings
21. DRA REVIEW,
NATALIE RATHVON, PH. D.
The following evidence of validation is based
upon the review of the DRA completed by:
Natalie Rathvon, Ph. D., Assistant Clinical
Professor, George Washington University,
Washington DC, Private Practice Psychologist
and School Consultant, Bethesda, MD
(August 2006)
22. DRA CONTENT VALIDITY
In a review by Natalie Rathvon, PH.D.
Oral Fluency, running record- derived from only Clay’s Observational Survey (Clay,
1993).
Teacher surveys (return rates were 46%), conducted (ns of 80 to 175) revealed
that DRA provided teachers with information describing reading behaviors and
identifying instructional goals
There were also concerns about adequacy and accuracy of the comprehension
assessment and the accuracy of text leveling prior to 2003 before the Lexile
framework evaluated the readability in the DRA text.
Concerns about who develop and reviewed the assessment. There is no evidence
that external reviewers participated in the development, revision, or validation
process.
Rathvon states, “Means, standard deviations, and standard errors of
measurement should be presented for accuracy, rate, and comprehension scores
for field test students reading adjacent text levels to document level-to-level
progression.”
23. CONSTRUCT VALIDITY EVIDENCE
Results from Louisiana statewide DRA administrations for spring
of 2000 through 2002 for students in grades 1 through 3 (ns =
4,162 to 74,761) show an increase in DRA levels across grades,
as well as changes in DRA level for a matched sample of student
(n = 32.739) over a three year period. This indicates that the
skills being measured are developmental.
The DRA as evidence can detect changes in reading levels.
As evidenced in two studies evaluating the relationship between
Lexile Scale measures and DRA running-record format is a valid
method of assessing reading comprehension.
24. SUMMARY OF WHAT DRA IS:
An attractive reading battery modeled after an informal
reading inventory based Clay’s Observational Survey
(Clay, 1993)
Authentic Texts
Instructionally relevant measures of fluency and
comprehension
Provides meaningful results for classroom teachers,
parents, and other stakeholders
Provides encouraging evidence that the use of DRA
predicts future reading achievement for primary grade
students.
25. DRA CRITERION RELATED
VALIDITY
No concurrent validity evidence is presented documenting the
relationship between the DRA and standardized or criterion-
referenced tests of reading, vocabulary, language, or other
relevant domains for students in kindergarten or grades 4 through
8.
There is a need for studies examining the extent to which
individual students obtain identical performance levels on the
DRA and validated reading measures are especially needed.
No information is provided to document the relationship between
the DRA Word AnalysNo concurrent validity evidence is presented
for any of the DRA assessments in terms of relationship between
DRA performance and contextually relevant performance
measures, such as teacher ratings of student achievement or
classroom grades.
26. SUMMARY OF WHAT DRA IS:
Responsive to intervention for primary grade students
An assessment model that has raised teacher
awareness of student reading levels corresponding
them with appropriate texts.
Teacher reviewed and survey based on classroom
practice (return rates were 46%), conducted (ns of 80
to 175) (Rathvon, 2006)
Provided evidence that the Lexile Scale measures and
DRA running record format is a valid method of
assessing reading comprehension.
27. SUMMARY OF WHAT DRA IS NOT:
Informal reading inventories lack in reliability and validity (Invenizzi et
al,; Spector, 2005)
Provide evidence of text equivalence within levels
Provide evidence for overall reading level for half the grade levels.
Have consistent process of text selection, scoring, and
administering- vulnerable to teacher inconsistencies and judgments
(improved since Lexile model)
Provide enough evidence of criterion-related validity for older
students
Provide concurrent validity evidence documenting the relationship
between the DRA and standardized or criterion-referenced tests of
reading, vocabulary, or language in kindergarten and grades 4-8
Provided to document the relationship between the DRA Word
Analysis and any criterion measure.
28. SUMMARY OF WHAT DRA IS NOT:
Provide sufficient evidence that teachers can select texts aligned
with students’ actual reading level (or achieve acceptable levels of
scorer consistency and accuracy)
Provide evidence of demographic groups
Include external reviewers in the development, revision, and
validitaion of any DRA series
Provide complete field testing reporting
Provide theoretical rationale or empirical data supporting the
omission of a standard task to estimate student reading level.
Provide standard means , standard deviations, and standard
errors of measurement ensuring accuracy
29. WHAT DOES ALL OF THIS MEAN?
Learning about the validity of the Developmental Reading
Assessment was difficult. I have yet to administer one, but
would like to go through the process. There is no empirical
evidence that consistently supports the validity of the DRA.
There are far too many variables, and opportunities for human
behavior to alter results and effect the variability.
However, the difference in how teachers approach the
diagnostics of the reading levels of students, and the
awareness of getting leveled texts in the classroom has
changed dramatically over the past few years. The changes
in reading instruction based on results of the DRA (though not
valid) has changed reading instruction in our district.