Professional Documents
Culture Documents
doi: 10.1093/deafed/enaa020
Advance Access Publication Date: 22 July 2020
Theory/Review Manuscript
T H E O RY / R E V I E W M A N U S C R I P T
Abstract
This review systematically identified and compared the technical adequacy (reliability and validity evidence) of reading
curriculum-based measurement (CBM) tasks administered to students who are deaf and hard of hearing (DHH). This review
included all available literature written in English. The nine studies identified used four CBM tasks: signed reading f luency,
silent reading f luency, cloze (write in missing words given blank lines within a passage), and maze (circle the target word
given multiple choice options within a passage). Data obtained from these measures were generally found to be internally
consistent and stable with validity evidence varying across measures. Emerging evidence supports the utility of CBM for
students who are DHH. Further empirical evidence is needed to continue to explore technical properties, identify if student
scores are sensitive to growth over short periods of time, and examine whether CBM data can be used to inform
instructional decision-making to improve student outcomes.
Students who are deaf or hard of hearing (DHH) are a heteroge- underachievement for this subgroup of students who are DHH
neous population, with some students meeting or exceeding lit- (Luft, 2018). Evidenced-based practices are supported by the use
eracy benchmarks and others demonstrating limited proficiency of valid and reliable assessment to measure student response
(Mayer & Trezek, 2018). A large majority of students experience to instruction and to use data to modify the approach when
persistent delays in obtaining proficiency similar to their same- needed (Rose, 2007; Thomas & Marvin, 2016).
aged peers with typical levels of hearing (Allinder & Eccarius, Within the field, there remains a critical need for assess-
1999; Luckner, 2013). Limited literacy proficiency can impede ment tools that can reliably measure a student’s skill level,
students who are DHH from finishing high school (Appelman can generate data that predict performance on comprehensive
et al., 2012), can create barriers in postsecondary education assessment measures, are sensitive to growth over short periods
(Hartmann, 2010), and present challenges with meeting literacy of time, and can inform instructional decision-making for stu-
demands in the workplace (Luft, 2012). dents who are DHH (Thomas & Marvin, 2016). One assessment
Educational legislation calls for the use of data-based method that shows promise to inform instructional planning is
instruction and evidenced-based educational strategies to curriculum-based measurement (CBM; Devenow, 2003; Luckner
efficiently and effectively provide high-quality instruction and & Bowen, 2006; Rose, 2007).
targeted individualized instruction to students who are at risk CBM, as conceptualized by Deno (1985), is an approach
or with disabilities (IDEIA, 2004; ESSA, 2015). As such, there is a that can effectively provide “vital signs” of a student’s overall
need for researchers to develop and practitioners to implement academic performance. CBM was designed to measure students’
evidenced-based practices to address the persistent problem of responsiveness to instruction and inform service delivery.
Received December 18, 2019; revisions received June 13, 2020; accepted June 15, 2020
© The Author(s) 2020. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
398
E. A. Lam et al. 399
Deno (1985) described that CBM should meet the following only communication mode, and that communication modali-
criteria: It should be (a) reliable and valid, (b) simple and efficient, ties may be more complex than a three-category delineation.
(c) easy to understand, and (d) inexpensive. CBM tools are Luckner (2013) called for further research exploring the technical
designed to be quick, efficient, and sensitive to growth over properties of student scores and the potential of using CBM for
short periods of time. progress monitoring. Given the need for viable assessments tools
for students who are DHH, the appropriateness of these tools
Use of CBM with Hearing Populations for other special populations, and the call for future research
to explore the technical features of CBM with students who are
For students with typical levels of hearing, oral reading fluency DHH, we conducted this systematic review.
(ORF) is the most commonly used CBM tool. Within the context
of progress monitoring, students are administered the ORF task
as often as once a week with performance scored and charted. Purpose of Review
The general process is as follows: Students are presented with a
passage and prompted to read the text aloud in the time allotted The purpose of this review was to explore the technical adequacy
(e.g., 1 min). The examiner tracks as the student reads, marks of CBM with students who are DHH. According to Fuchs (2004), a
Figure 1. Systematic review screening process documents the systematic search process including the following steps: identification, screening, eligibility, and included
studies.
examined 38 technical reports from the Research Institute on Last, we conducted an interrater agreement (IRA) to deter-
Progress Monitoring (RIPM), and 1 article was identified (Rose, mine the consistency in decision-making. A trained researcher,
2008) and met the inclusion criteria in the full-text screen. We independent of the authors, conducted the review of 20%
identified an additional unpublished master’s thesis (Barkmeier, of the articles at each stage. IRA was as follows: initial
2009) which shared the same sample of participants as Rose database (99% title screen, 100% full screen), ancestral searches
(2008) but analyzed student performance on a different set of (88%, 100%), frequently occurring journals (99%, 100%), and
measures. The master’s thesis was screened, met criteria, and RIPM (100%, 100%). All disagreements were reviewed and
was ultimately included. resolved.
E. A. Lam et al. 401
Inclusion Criteria IRA was conducted with a trained researcher with 20% of
the studies selected for review. In some cases, the first author
To be included in this review, each study needed to (1) assess
reported ranges where the rater reported the summary or com-
the reading performance of students who are DHH using at least
posite value, though both representations of the data were cor-
one CBM reading task, (2) explore the technical characteristics
rect. Inter-rater agreement was 96%; two disagreements were
(validity and reliability) of student scores, and (3) be written in
identified and resolved.
English inclusive of theses, dissertations, and technical reports.
For the CBM requirement, we accepted studies that generally
met the criteria set forth by Deno (1985), except in some cases
the scoring did not meet the criteria of simple and efficient.
Results
For the reliability and validity requirements, we used the def- The literature search yielded nine studies that fell into four
initions set forth by the American Educational Research Asso- categories (signed reading fluency, silent reading fluency, cloze,
ciation (AERA, 1999). As defined, reliability refers to “the con- and maze) described in more depth below. For each study, the
sistency of such measurements when the testing procedure is reliability and validity findings were reviewed. See Table 1 for the
repeated on a population of individuals or groups” (p. 25). Within reliability and Table 2 for the validity findings for each study.
Table 1 Reliability of CBM for students who are deaf or hard of hearing
Allinder & Eccarius (1999) Inter-rater Words read correctly 40 to 100% (M = 78.69%)
Idea units retold 0 to 100% (M = 78.76%)
Internal consistency 1- & 3-min passages .89 to .97
Alternate form 1-min passages .85
3-min passages .94
Lam (2020) Alternate form: Correct words identified .50 to .69 <.01
paper–pencil Correct boundaries .51 to .70 <.0 l
Percent correct .39 to .59 <.01 to .02
Alternate form: e-based Correct boundaries .68 to .75 <.01
Percent correct .49 to .71 <.01
Cloze
LaSasso (1980) Internal consistency Fifth grade passages (six forms) .67 to .82
Kelly & Ewoldt (1984) Inter-rater (sample of 100 Meaningful to passage 82%
responses) Meaningful in sentence 81%
Related to English form 79%
Sign form classification 82%
Maze
Devenow (2003) Alternate form Correct (Phase 2) .60 to .80 <. 001
Corrected (Phase 2) .64 to .82 <. 001
Scan (Phase 2) .45 to .70 <. 001
Note: CMC, correct maze choices; IMC, incorrect maze choices; CMC-IMC, correct maze choices minus incorrect maze choices; CMC-IMC/2, correct maze choices minus
the value of incorrect maze choices divided by a value of 2.
E. A. Lam et al. 403
Table 2 Validity of CBM for students who are deaf or hard of hearing
Allinder & Eccarius Achievement test TERA-DHH M no. of words read 1 min .30 ns
(1999) M no. of words read 3 min .21 ns
M no. of idea units retold .36 ns
M no. of words retold .46 p < .05
M no. of unique words retold .47 p < .05
% of content words retold .46 p < .05
Easterbrooks and Achievement test WRMT-R Word comprehension .38 to .46 p < .05
Huston (2008) Passage comprehension .55 to .64 p < .01
Rose, (2008) Fluency test TOSCRF Correct words identified .84 & .90 Not reported
Achievement test MAP Correct words identified .58 to .75 Not reported
Informal Teacher ratings Correct words identified .54 to .85 Not reported
Lam (2020) Achievement WJ-III Passage Correct words identified .33 to .48 <.01 to .05
(paper–pencil CBM) Comprehension Correct boundaries .30 to .45 <.01 to .07
Total wrong −.51 to −.26 <.01 to .12
Percent correct .27 to .49 <.01 to .10
MAP Correct words identified .55 to .71 <.01
Correct boundaries .53 to .72 <.01
Total wrong −.42 to −.71 <.01 to .04
Percent correct .44 to .73 <.01 to .03
Achievement WJ-III Passage Correct boundaries .25 to .34 .04 to .14
(e-based CBM) Comprehension Total wrong −.53 to .05 01 to .76
Percent correct .08 to .50 <.01 to .62
MAP Correct boundaries .37 to .51 <.01 to .07
Total wrong −.59 to −.21 <.01 to .32
Percent correct .28 to .58 <.01 to .18
Cloze
LaSasso (1980) CBM Reading for concepts Passages (third, fifth, & Ranking (easy to NA
series seventh) difficult): fifth,
seventh, third grade
Maze
Chen (2002) Achievement test TOWL-3 (winter, Correct .76 & .88 p < .01
spring) Corrected .77 & .89 p < .01
Scan .82 & .83 p < .01
Accuracy .52 & .62 p < .05 & p < .01
Incorrect .26 & -.17 ns
Informal measure Teacher ratings Correct .79 & .76 p < .01
(winter, spring) Corrected .82 & .74 p < .01
Scan .80 & .74 p < .01
Accuracy .48 & .50 p < .05 & p < .01
Incorrect .19 & .01 ns
Devenow (2003) Achievement test SAT P1-2M correct .72 & .74 p < .05
P1-2M corrected .74 & .75 p < .05
P1-2M scan .56 & .71 p < .05
P1-4M correct .64 & .64 p < .05
P1-4M corrected .64 & .66 p < .05
(Continued)
404 Journal of Deaf Studies and Deaf Education, 2020, Vol. 25, No. 4
Table 2 Continued
Barkmeier (2009) Achievement test MAP Elementary .80 & .91 Not Reported
Middle school .59 to .91 Not Reported
Lam (2020) Achievement WJ-III Passage CMC .33 to .57 <.01 to .05
(paper–pencil CBM) Comprehension IMC −.27 to −.43 <.01 to .10
CMC-IMC .50 to .58 <.01 to .02
CMC-IMC/2 .36 to .58 <.01 to .03
MAP CMC .49 to .65 <.01 to .01
IMC −.50 to −.20 .01 to .33
CMC-IMC .50 to .67 <.01 to .01
CMC-IMC/2 .50 to .67 <.01 to .01
Achievement WJ-III Passage CMC .43 to .50 <.01
(e-based CBM) Comprehension IMC −.50 to −.33 <.01 to .05
CMC-IMC .49 to .52 <.01
CMC-IMC/2 .46 to .51 <.01
MAP CMC .56 to .60 <.01
IMC −.18 to −.04 .40 to .84
CMC-IMC .51 to .61 <.01
CMC-IMC/2 .54 to .61 .01
Note: TERA-DHH, Test of Early Reading Ability—Deaf or Hard of Hearing; WRMT-R, Woodcock Reading Mastery Test—Revised; TOSCRF, Test of Silent Contextual Reading
Fluency; MAP, measures of academic progress; CBM, curriculum-based measurement; WJ-III, Woodcock-Johnson Tests of Achievement—Third Edition; SAT-HI, Stanford
Achievement Test—Hearing Impaired; TOWL-3, Test of Written Language—Third Edition; SAT, Stanford Achievement Test; P1-2M, Phase 1—2 min; P1-4M, Phase 1—4
min; P1, Phase; P2, Phase 2; CMC, correct maze choices; IMC, incorrect maze choices; CMC-IMC, correct maze choices minus incorrect maze choices; CMC-IMC/2, correct
maze choices minus the value of incorrect maze choices divided by a value of 2.
Fluency Rubric for Deaf Children (Easterbrooks & Huston, 2008). authors varied in their approaches of scoring student responses
Study participants were first administered the Signed Read- which included requiring word-by-word translation or assessing
ing Fluency Rubric for Deaf Children and then, 0 to 4 months the quality of the signed interpretation of the text. Since the task
later, were administered two subtests (Word Comprehension demands and scoring methods varied widely for signed reading
and Passage Comprehension) from the Woodcock Reading Mas- fluency, it is questionable if the authors were measuring the
tery Test—Revised (WRMT-R; Woodcock, 1987). same construct. As such, these findings suggest a complexity
Inter-rater agreement was strong for the fluency envelope in administering, scoring, and interpreting the performance of
metric (overall visual appearance of the signed interpretation— sign reading fluency with students who are DHH who use a sign-
with or without voice) and moderate for the visual grammar based system.
metric (elements that demonstrate the reader is deriving mean-
ing from the text—in an English-like mode or ASL). Internal
consistency fell in the strong range. Criterion validity evidence Silent Reading Fluency
for the Signed Reading Fluency Rubric for Deaf Children with
Two research teams explored silent reading fluency (Rose, 2008;
the WRMT-R (Woodcock, 1987) spanned the moderate to strong
Lam, 2020). Rose (2008) conducted research at a school for the
ranges.
deaf within the context of a school-wide progress monitoring
In summary, two studies (Allinder & Eccarius, 1999; East-
program. Participants (n = 101, grades 3 to 12) attended a resi-
erbrooks & Huston, 2008) explored the reliability and validity
dential school for the deaf with six students attending the com-
of student performance when students were presented with a
munity public school part-time. All students qualified for special
passage and prompted to read the passage in sign language.
education services due to their hearing loss status (n = 36 mild
Participants within the studies varied in the features of their
to moderate, n = 61 severe to profound, n = 4 range unknown).
expressive sign language communication, which may impact
Twenty-three percent (n = 23) of the sample had additional dis-
how the students processed the task, the degree of difficulty
abilities. Within the school-wide program, Rose (2008) analyzed
of the task, and the type of student response generated. The
student performance on the measures of academic progress
E. A. Lam et al. 405
(MAP; NWEA, 2003), silent reading fluency test (APPROACH 1: ranges. For criterion validity in the paper–pencil conditions, WJ-
(SRFT); Rose and McAnally, 2008), test of silent contextual read- III Passage Comprehension (Woodcock et al., 2001) correlations
ing fluency (TOSCRF; Hammill et al., 2006), and teacher ratings. fell in the weak to moderate range and MAP (NWEA, 2003)
See Maze section for further details on the findings from Bark- correlations in the moderate to strong ranges. For e-based, the
meier (2009), who analyzed student data from the CBM maze, correlations ranged from weak to strong for passage comprehen-
MAP (NWEA, 2003), and teacher ratings. sion subtests and weak to strong for MAP.
Rose (2008) presented the SRFT (in which students read mod- In summary, Rose (2008) and Lam (2020) explored the relia-
ified passages in which the story was presented in all upper bility and criterion-related validity of silent reading fluency. The
case letters with no spaces or punctuation and students put a results across the studies varied with the results of Rose (2008)
slash between the boundaries of words). This measure used the generally reporting higher correlations for reliability and validity
formatting structure of the TOSCRF (Hammill et al., 2006) with as compared to Lam (2020). Even when comparing the studies
content derived from Reading Milestones (Quigley et al., 2001) and using only the paper–pencil condition and the same metrics,
Reading Bridge (Quigley et al., 2003). For the SRFT and TOSCRF the differences remained. It appears as if sample differences in
tasks, the participant’s score was calculated by summing the size, demographics, measures selected, and length of the tasks
correct number of words identified. The SRFT and the TOSCRF, presented may have impacted differences in scores. Due to these
In summary, two studies (LaSasso, 1980; Kelly & Ewoldt, 1984) Measurement, 1995) yielded the strongest correlations when
explored the reliability and criterion validity of Cloze. Overall, administered in the 2-min time condition.
reliability was moderate to strong (LaSasso, 1980) and relativity As noted earlier, Rose (2008) and Barkmeier (2009) shared the
high (Kelly & Ewoldt, 1984). To evaluate validity, both authors same sample of participants but Rose (2008) analyzed silent
selected a different method of analysis. For LaSasso (1980) the reading fluency results, whereas Barkmeier (2009) explored
passages did not rank from easy to hard based on student maze. Maze passages, derived from the basic academic skill
performance, and for Kelly and Ewoldt (1984), there was higher sample (Espin et al., 1989), were administered three times during
agreement in the consistency of decisions for the SAT-HI than the school year. The score was the number of incorrect words
the story retell. subtracted from the total number of words correct. Perfor-
mance was compared to the MAP (NWEA, 2003) and teacher
ratings.
For reliability, student performance varied across forms and
Maze grade levels, with the highest correlations noted for elementary
Four studies (Chen, 2002; Devenow, 2003; Barkmeier, 2009; Lam, aged students using testing form D. For criterion-related validity,
2020) explored the technical adequacy of the maze, which is a strong correlations were present for both MAP (NWEA, 2003) and
of the student’s overall reading ability? Would a student’s role of oral reading fluency when used with students who are
performance on a CBM task be similar to their performance DHH who use oral communication.
on an achievement task (e.g., high CBM performance and high
achievement performance, would the inverse be true)? Below we
describe the reliability and validity evidence for each of the four Silent Reading Fluency
measures.
Silent reading fluency was measured in two of the nine studies
(Rose, 2008; Lam, 2020) in this review. In Rose (2008) reliability
Signed Reading Fluency was high. Additionally, for students with typical levels of hearing,
correlations between the TOSCRF (the measure to which the
For signed reading fluency variation was present in how the
SRFT was based) and criterion measures ranged from .57 to .80
authors approached viewed, interpreted, and scored student
(Hammill et al., 2006), which is generally similar to the findings
signed responses (Allinder & Eccarius, 1999; Easterbrooks & Hus-
of Rose (2008) using the SRFT with students who are DHH. In
ton, 2008). Across studies correlations were strong for internal
contrast, Lam (2020) reported lower reliability coefficients for
consistency with more notable variation for other reliability
test retest and validity coefficients that were rarely sufficient.
types. Validity evidence between signed reading fluency and an
students. This pattern of more modest correlations in the upper Additionally, students who are DHH who use spoken English
grades is consistent with the pattern of correlations for students to communicate were underrepresented in this review. Since
with typical levels of hearing (Marston, 1989; Wayman et al., ORF is the most common measure for students with typical
2007). hearing, the consideration of appropriateness of students who
These findings provided preliminary validity evidence that are DHH who use spoken English is likely. Available research
student data on the maze may serve as a general outcome suggests that students who are DHH are diverse in how they
measure of students’ overall reading competency. In general, acquire reading skills and the cognitive processes they use when
when students performed well on maze, their overall perfor- engaging in the reading process (Herman et al., 2019). Reading
mance on achievement tests was also high, with the inverse tasks, including ORF, provides limited information as to how the
finding present as well. In addition to the empirical evidence, student is cognitively processing the text and assumes auditory
theoretically maze may be a promising option as it does not and phonological fluency in English (Luft, 2019). Understanding
require oral or signed production and may allow for variation of these factors is needed when administrating and interpreting
in cognitive processing (Marschark, 2006; Marschark et al., 2011). the findings of student performance using ORF and may limit
Since only four studies explored maze, additional evidence are comparison of performance to students with typical levels of
needed to confirm these findings. hearing.
References Fuchs, L. S., Fuchs, D., & Hamlett, C. L. (1989). Monitoring reading
growth using student recalls: Effects of two teacher feed-
Allinder, R. M., & Eccarius, M. A. (1999). Exploring the technical
back systems. Journal of Educational Research, 83, 103–111. doi:
adequacy of curriculum-based measurement in reading for
10.1080/00220671.1989.10885938
children who use manually coded English. Exceptional Chil-
Fuchs, L. S., & Vaughn, S. (2012). Responsiveness-to-intervention:
dren, 65(2), 271–283. doi: 10.1177/001440299906500210
A decade later. Journal of Learning Disabilities, 45(3), 195–203.
Appelman, K. I., Callahan, J. O., Mayer, M. H., Luetke, B. S., &
doi: 10.1177/0022219412442150
Stryker, D. S. (2012). Education, employment, and inde-
Hammill, D. D., & Larsen, S. C. (1996). Test of Written Language (3rd
pendent living of young adults who are deaf and hard of
Ed.). Austin, TX: Pro-Ed.
hearing. American Annals of the Deaf , 157(3), 264–275. doi:
Hammill, D. D., Wiederholt, J. L., & Allen, E. A. (2006). TOSCRF: Test
10.1353/aad.2012.1619
of Silent Contextual Reading Fluency, Examiner’s manual. PRO-
American Educational Research Association, American Psycho-
ED.
logical Association, & National Council on Measurement in
Harcourt Brace Educational Measurement (1995). Stanford
Education (1999). Standards for educational and psychological
Achievement Test (9th Ed. Ed.Form S (Stanford 9)). Pearson.
testing. Washington D.C: American Educational Research
Hartmann, E. A. (2010). Evaluating employment outcomes of adults
hearing students. Journal of Deaf Studies and Deaf Education, Rose, S., McAnally, P., Barkmeier, L., Virnig, S., & Long, J. (2008).
23(2), 148–163. doi: 10.1093/deafed/enx057. Silent reading fluency test: reliability, validity, and sensitivity to
Luft, P. (2019). Strengths-based reading assessment for deaf and growth for students who are deaf and hard of hearing at the
hard-of-hearing students. Psychology in the Schools, 57(3), elementary, middle school, and high school levels. (Report No. 9).
375–393. doi: 10.1002/pits.22277. Minneapolis, MN: University of Minnesota.
Luckner, J. L. (2013). Using the dynamic indicators of basic literacy Sandberg, K. L., & Reschly, A. L. (2011). English learners: Chal-
skills with students who are deaf or hard of hearing: Per- lenges in assessment and the promise of curriculum-based
spectives of a panel of experts. American Annals of the Deaf , measurement. Remedial and Special Education, 32(2), 144–154.
158(1), 7–19. doi: 10.1353/aad.2013.0012. doi: 10.1177/0741932510361260.
Luckner, J., & Bowen, S. (2006). Assessment practices of profes- Shin, J., & McMaster, K. (2019). Relations between CBM (oral
sionals serving students who are deaf or hard of hearing: reading and maze) and reading comprehension on state
An initial investigation. American Annals of the Deaf , 151(4), achievement tests: A meta-analysis. Journal of School Psychol-
410–417. doi: 10.1353/aad.2006.0046. ogy, 73, 131–149. doi: 10.1016/j.jsp.2019.03.005.
Marschark, M. (2006). Intellectual functioning of deaf adults and Talbott, E., Maggin, D. M., Van Acker, E. Y., & Kumm, S.
children: Answers and questions. European Journal of Cogni- (2017). Quality indicators for reviews of research in