Teachers have used the method of repeated readings to build oral
reading fluency in students with and without special needs. A new
fluency building intervention called interval sprinting uses shorter
timing intervals (i.e., sprints) across a passage. This study used an
alternating treatment design to compare repeated readings and interval
sprinting for three participants with or atrisk for reading deficits.
The results show that the participants attained the fluency criterion on
one of the reading practice procedures after approximately the same
number of sessions without consensus as to the procedure. Compared to
readings in the first phase, students demonstrated higher average
initial readings and reached criterion as fast or faster in the second
phase, demonstrating reading transfer.
**********
Olympic speed skaters race over distances from 500 to 10,000 meters
(approximately 0.3 to 6.0 miles) sometimes exceeding on-ice speeds of 30
miles per hour (Vickers, 2006). Elite marathon runners maintain six
minute mile speeds for over two and a half hours with perfect form
(Binder, 1996). These athletes have not only mastered the skills
particular to their sport, but perform their well practiced behaviors
effortlessly and with deft composure. Such rapid and automatic
performances fall under the term behavioral fluency. Individuals
demonstrate behavioral fluency when they show, "that combination of
accuracy plus speed of responding that enables competent individuals to
function efficiently and effectively in their natural environments"
(p. 163).
While the previously mentioned examples take place in the world of
competitive sports, there exists a parallel to academics. Teachers
aspire for their students to perform academic behaviors with grace and
fluidity (i.e., fluency). For example, teachers strive to have students
write dynamic, well-composed manuscripts in English class, diligently
work through complicated Algebra problems, or confront problems with
critical thinking. In the past, the educational system has not fully
embraced fluency in all of its subject matter (Ailington, 1983; Kubina
& Morrison, 2000). But changes in both research and practice has
occurred. Notably, in some subjects like reading fluency interventions
have taken center stage (National Reading Panel, 2000).
As researchers have begun to scientifically scrutinize fluency in
reading, they have made a number of important discoveries. For example,
oral reading fluency, or ORF, provides a quality measure of a
student's overall reading competence (Fuchs, Fuchs, Hosp &
Jenkins, 2001; Shinn & Good, 1992). Poor readers display lower
levels of ORF than good readers. Students with low ORF often stammer
while reading and easily become distracted which adversely affect their
understanding and comprehension of text (Binder, Haughton, & Van
Eyk, 1990; Fuchs et al., 2001). Students who demonstrate high levels of
ORF experience many benefits (Binder, 1996; Fuchs et al.; Kubina &
Morrison, 2000). Fluent readers appear to naturally translate text into
spoken language appearing to glide while reading (Archer, Gleason, &
Vachon, 2003; Fuchs et al.). Because they spend less time decoding,
fluent readers also better understand what they read (Mounsteven, 1990;
National Reading Panel, 2000; Therrien, 2004).
Time plays a major role in ORF. Even though some have
conceptualized ORF as the ability to read with prosody (Schreiber,
1980),Archer et al. (2003) suggest the epitome of ORF breaks down to
"rate plus accuracy" (p. 96). This definition places ORF into
observable and measurable terms: words per minute. Researchers have
found students' one-minute reading samples correlate well to
standardized reading assessment scores (Good, Simmons, &
Kame'enui, 2001), formatively measure a student's reading
progress across time (Deno, 2003), and can assist in the identification
of reading deficits and special education needs (Hosp & Fuchs,
2005). Additionally, one-minute reading timings appear often in a useful
fluency building method called repeated readings.
Because increased levels of ORF require practice, teachers and
researchers need a systematic plan for fluency building (Archer et al.,
2003; Kubina, 2005). Repeated reading represents one method for fluency
building for students with and without special needs (Chard, Vaughn,
& Tyler, 2002; Kuhn & Stahl, 2003; Meyer & Felton, 1999).
Kavale (2005) reported a .76 mean effect size for repeated readings for
students with specific learning disabilities. Repeated reading
interventions have effectively assisted students with speech-language
impairments (e.g., Begeny, Daly, & Valleley, 2006), learning
disabilities (e.g., Nelson, Alber, & Gordy, 2004; Valleley &
Shriver, 2003), and emotional and behavioral disorders (e.g., Scott
& Shearer-Lingo, 2002; Staubitz, Cartledge, Yurick, & Lo, 2005).
Additionally, researchers have used repeated readings effectively in
peer-mediated (e.g., Yurick, Robinson, Cartledge, Lo, & Evans, 2006)
and small group settings (e.g., Kuhn, 2005) to build fluency.
At the basic structure of repeated readings, a student reads the
same instructional-level passage until meeting a pre-set criterion of
words per minute or a specific number of passage reading repetitions
while receiving feedback and error correction (Meyer & Felton, 1999;
Therrien, 2004; Therrien & Kubina, 2006). Once meeting criterion or
number of rereads with the first passage, the student would repeat the
process with another passage and so on (Meyer & Felton; Therrien).
In his meta-analysis of repeated readings for students with learning
disabilities, Therrien reported effect sizes (ES) for critical repeated
reading components: 3-4 rereads (ES= .85-.95), error correction (ES =
1.37), and reading to a criterion (ES = 1.74).
The procedure of repeated readings can take many forms. The student
can reread passages as a teacher or peer reads the passage (e.g., Kuhn,
2005), reread the passage without assistance (e.g., Compan, Iamsupasit,
& Samuels, 2001), or reread the passage in a prosodic reading format
targeting expression and inflection (e.g., Young, Bowers, MacKinnon,
1996). Additionally, timed practice readings can match (e.g., Therrien,
Wickstrom, & Jones, 2006; Valleley & Shriver, 2003) or not match
(e.g., Begeny & Silber, 2006; Staubitz et al., 2005) the test
reading. Regardless of the specific form, the goal of repeated reading
remains the same, the development of fluency.
Behavioral fluency shares the functional definition of fluency
found in many definitions of ORF (e.g., Archer et al., 2003; Binder,
1996; Fuchs et al., 2001). Behavioral fluency, however, also predicts
the associations of critical learning outcomes. One such critical
learning outcome, endurance, states that a learner can engage in a task
at a uniform rate for a given period of time. Furthermore, the learner
can do so in the face of environmental distraction (Binder, 1996; Binder
et al., 1990). When placed in the context of reading, endurance means
students maintain reading accuracy, rate, and continue reading in the
face of distraction. Students who do not have endurance (i.e., yet to
achieve fluency) tend to read slower, for shorter intervals of time, and
may display more off-task behavior (Binder et al.). To build endurance,
researchers from the field of Precision Teaching, from where behavioral
fluency originated, have developed several practice methods. One such
method calls for the teacher to use shorter timings called sprints to
build endurance (Binder; Haughton, 1980).
The practice of endurance building with academic skills has similar
characteristics with athletic endurance building. As an example, Gibala
et al. (2006) have found a way to efficiently promote endurance. As a
result of comparing sprint interval and traditional endurance training
methods, they concluded that sprint interval training produced
biological changes similar to endurance training in a time efficient
manner. Therefore, the results from athletic training suggest a similar
outcome attainable in reading when developing oral reading fluency:
interval sprinting. Can an academic sprint-interval training procedure
produce results similar to, or better than, reading fluency methods like
repeated reading but with the added benefit of efficiency? We present a
new method called interval sprinting (IS) to explore the question of
efficiency and feasibility.
Interval sprinting with reading involves two main steps. First, a
practitioner determines a reading passage for IS practice and equally
divides the passage into a specified number of parts (i.e., number of
words) based on the length of time per time interval (i.e., number of
seconds). Second, while maintaining the parts in context (i.e., passage
remains intact), students would spend their time sprinting (i.e.,
reading aloud) from each of the starting points across the passage.
Therefore students have me opportunity to read me first part of a
passage, then the second part, and so on until the end of the passage.
Because the passage parts remain in context, an important aspect of
repeated readings (Therrien & Kubina, 2007), students may
conceivably read quickly enough to practice some aspects of the passage
twice as often as compared to students reading the passage for one
minute. This additional practice within the same allotment of practice
time leads to the possibility of increased reading efficiency during the
fluency building procedure.
Even with increased time allocated for reading, Heibert and Fisher
(2005) note that teachers should continue to look for the most efficient
fluency building methods available. Can modifications to the repeated
reading procedures produce both effective and efficient results? The
purpose of this study seeks to examine the IS procedure. To do so, we
compared IS with repeated readings to a fluency criterion and asked two
questions: (a) Under which condition (i.e., IS or repeated reading to a
fluency criterion) will students reach criterion first (i.e., 200+
correct words per minute with 2 or less errors) and (b) given equal
reading practice time, which condition will students practice reading
more words
Methods
Subjects and Setting
Two 10-year-old males, Brad and Joseph, and one 8-year-old male,
Jason, participated in this study. From individual parent reports, Brad
had a specific learning disability in reading,
AttentionDeficit/Hyperactivity, auditory processing, and memory
disorders while Joseph had an auditory processing disorder which falls
under the other disability special education classification. Both
received services in resource room settings and had reading goals on
their respective Individualized Education Plans stating they needed
additional reading support. Jason did not receive special education
services; due to difficulties with reading, he received reading
assistance in the form of Title 1.
The experimenter recruited all participants through community
referrals from university faculty in a special education program. The
study took place during summer vacation; therefore all sessions occurred
in each participant's home at either a kitchen (i.e., inside) or
picnic table (i.e., outside). The experimenter conducted all sessions
and met with each participant every available weekday during the course
of the study. Summer activities (e.g., sports camps, family outings,
vacations, etc.) affected session days for certain participants. Each
session lasted approximately 10 to 20 minutes.
Materials
All readings came from Dynamic Indicators of Basic Early Literacy
Skills (Good & Kaminski, 2007). Following reading assessments (noted
below in Assessment/current level of performance), the experimenter
selected four readings from the 6th edition second and fourth grade oral
reading fluency progress monitoring packets (Good & Kaminski). Fry
readability procedures (Fry, 1989) confirmed the grade levels of each
passage prior to inclusion. Each passage contained separate stories
focusing on different topics. Table 1 shows each specific passage
number, the number of words per passage, and the word overlap between
passages.
Calculating word overlap involved a three step process. (1) After
entering every word of each passage into a spreadsheet, a comparison of
words followed. (2) Each instance of a word in two passages counted as
one overlapping word. Multiple words counted as overlapping only as many
times as they overlapped. For example, passage 1 in second grade had the
word 'the' seven times. 'The' appeared 10 times in
passage 10. Therefore 'the' overlapped seven times within
these two passages. (3) Dividing the smallest word total of the two
passages by the number of overlapping words between those two passages
yielded the percent of word overlap. Second grade passages overlapped
35% on average and 24% on average when not including commonly occurring
words (i.e., a, an, and, I, the). Fourth grade passages overlapped 29%
on average and 17% on average when not including commonly occurring
words.
Additional materials included copies of each passage for the
experimenter and participants, a count down timer, No. 2 pencils, and a
video recorder and tripod. Each passage copy used by the experimenter
had word totals at the end of each line while participant copies had no
word count and larger font. To determine timings of reading intervals,
the experimenter used a timer that beeped at the conclusion of each
interval.
Dependent Variables
Measurement of two dependent variables, words read correctly and
incorrectly, took place during this study. Participants read passages
aloud to the experimenter who scored correct and incorrect read words on
a copy of the passage. Correct words consisted of words pronounced
correctly in their appropriate place in the text within three seconds
(Shinn, 1989). Also, self-corrected errors counted as correct.
Mispronunciations, omissions, substitutions, and words not read within
three seconds counted as incorrects (Shinn, 1989). Additionally,
inserting additional words not found in the text counted as incorrects.
Skipping lines of text did not fall into either a correct or incorrect
tally. Measurement of correct and incorrect words occurred during
one-minute intervals. Calculation of corrects consisted of subtracting
the number of incorrects from the total number words read plus any added
words during each interval.
Independent Variable
The independent variable applied during the study had two different
conditions: interval sprinting (IS) and repeated readings to a fluency
criterion (RRFC).
Interval sprinting condition. Interval sprinting involved asking
the participants to read from six specific points within the passage.
The specific starting points of each sprint passage came from dividing
the first 204 words of a passage into six equal parts, which resulted in
six sprint passages of 34 words each. To maintain overall passage
congruity, the starting point for each sprint passage occurred in steps
of 34 words (i.e., 1st, 35th, 69th, 103rd, 137th, and 171st words).
To begin the condition, the experimenter provided the participant
with an unmarked copy of the passage and told them that they would read
aloud from six different parts of the passage for 10 seconds twice and
to read as fast as they can. The experimenter's copy of the passage
had the six starting points circled, and a running word total at the end
of each line. After setting the count down timer for 10 seconds and
prompting the participant to read from the 1st word, the experimenter
had the participant start reading while recording incorrect words and
marking the final word read. After the timer beeped, the experimenter
prompted the participant to stop reading and quickly tallied and noted
corrects and incorrects without providing feedback* The experimenter
reset the timer for 10 seconds and prompted the participant to read from
the first word of the 1st part again. The experimenter again scored
incorrects and the final word read as the timer expired. The
experimenter again silently tallied and noted corrects and incorrects.
Once the participant completed two readings for the 1st sprint
passage, regardless of performance, the experimenter prompted him to
start reading from the beginning of the 2nd sprint passage (i.e., the
35th word). The participant then read the 2nd sprint passage twice for
10 seconds each with the experimenter again recording corrects and
incorrects without providing feedback* After the second timing for the
2nd sprint passage, the participant moved, with prompting, to the 3rd
sprint passage (i.e., 69th word). Again, the participant read this part
twice for 10 seconds without feedback.
After the second timing of the 3rd passage, the experimenter
provided verbal feedback and error correction over the first six
sprints. Verbal feedback consisted of explaining the number of correct
and incorrect words read per sprint passage reading. Error correction
comprised a model-lead-test procedure (Carnine, Silbert, Kame'enui,
& Tarver, 2004). Every word mispronounced, omitted, substituted,
skipped, or hesitated on for longer than 3 seconds received the same
type of correction. The experimenter pointed to the word on the
participant's copy of the passage and stated, "This word is
--. What word?" The participant had to pronounce the word properly*
When the participant added words into the passage, the experimenter
pointed to the word that occurred just before and just after the added
word and said, "These words are -- and --. What words?" The
participant had to pronounce both words properly. The experimenter
repeated this process for every error made over the course of the first
six sprints.
Following oral feedback and error correction, the experimenter
repeated the above procedure with the remaining three sprint passages.
In other words, the participant read, with prompting and timing from the
experimenter, the 4th, 5th, and 6th sprint passages twice each for 10
seconds. Following verbal feedback and error correction over the final
six sprints, the experimenter had the participant read from the start of
the overall passage for a one minute test.
Repeated readings to a fluency criterion condition. Repeated
readings to a fluency criterion (RRFC) condition involved the
participant reading the targeted passage for one minute twice. To start
this condition, the experimenter provided the participant with an
unmarked copy of the passage and some initial instructions. The
experimenter told the participant that he would read the passage twice
for one minute each time and to read as fast as possible. Setting the
timer for 60 seconds and prompting the participant to start at the 1st
word of the passage, the participant started reading and continued until
the timer beeped. The experimenter recorded incorrects on a copy of the
passage that included a running word total at the end of each line. The
experimenter provided verbal feedback in the form of correct and
incorrect words read and performed error correction identical to the
error correction discussed in the IS condition. The experimenter
conducted identical procedures for the second reading before having the
student complete a one minute test reading of the passage.
Experimental Design
The experimenter used an alternating treatments design (Kennedy,
2005) to evaluate the effects of the IS and RRFC conditions. Because
each session included one instance of IS and one instance of RRFC, each
session started with an alternating procedure. The alternating
treatments design does not require baseline, however, all participants
had one unpracticed initial reading of each passage before starting
practice procedures. The preliminary reading of the first two passages
served as both a reading level assessment and a current level of
performance. The preliminary reading of the second two passages served
as a current level of performance.
One concern of using an alternating treatments design with reading
involves interaction effects. In other words, the reading of one passage
can affect the reading of the other or future passages. In an attempt to
minimize these potential interaction effects, all passages within a
grade level had a percentage of word overlap calculated that did not
exceed 30% when controlling for some common words used in most readings
and all passages contained independent content.
Procedures
Assessment/current level of performance. To assess current level of
performance, participants read two passages. Because this study took
place during the summer, initial passage grade levels mirrored the grade
each participant completed the past school year. Using ranges presented
by Kubina and Starlin (2003), participants reading between 0-49 correct
words on either passage would read two new passages from one grade
earlier. If participants read more than 150 correct words per minute on
either passage they would read two new passages from one grade level
above that passage. The assessment stopped when a participant read
between 50 and 149 correct words in one minute on both passages placing
them in the instructional level for the current passages (Kubina &
Starlin). This final reading counted as the current level of performance
for these two passages. Brad and Joseph, both recently completing the
fourth grade, read fourth grade passages at the instructional level and
Jason, recently completing second grade, read second grade passages at
the instructional level.
Sessions. Prior to the start of the first session of each phase,
the experimenter randomly chose one of the passages for the IS condition
and one of the passages for the RRFC condition. Once assigned, each
passage remained within its condition until the participant read one of
the two passages to fluency criterion. All sessions contained one IS and
one RRFC condition. Therefore, the participant would complete one of the
two conditions and a test before moving to the other condition and test
on the other passage. The starting condition alternated for each
session. At the conclusion of each session, the experimenter provided a
small age-appropriate prize based on participation, not performance
(LeBlanc, Coates, Daneshvar, Charlop-Christy, Morris, & Lancaster,
2003). The experimenter gave the prize following the last test. During
this time, the experimenter and participant cleaned up reading and
recording materials and talked about summer activities. Sessions in the
first phase continued until the participant read one of the two passages
at the fluency criterion (i.e., 200 or more correct words per minute
with two or fewer errors) (Freeman & Haughton, 1993; Kubina &
Starlin, 2003).
Once a participant reached the fluency criterion, he began the
second phase. The second phase started with an initial, unpracticed
reading of two passages of the same grade level as passages from the
first phase. After randomly assigning the two passages to one of the two
conditions, the experimenter conducted phase two in the exact same
manner as phase one.
Tests. At the conclusion of each condition, the experimenter tested
each participant on the current passage. The experimenter set the timer
for 60 seconds and prompted the participant to begin reading from the
first word of the passage. Once reading, the experimenter recorded
incorrects and noted the word reached when time expired. Without
performing an error correction procedure, the experimenter debriefed the
participant on the number of words read correctly and incorrectly.
Interscorer Agreement
An independent observer provided interscorer agreement (IA) and
procedural integrity (PI) measures for a randomly selected 25% of the
sessions' video recordings. Prior to performing IA and PI, the
observer engaged in training sessions. These sessions consisted of the
independent observer watching video tapes of participants. The
independent observer then measured words read correctly and incorrectly
and compared their scores to a predetermined true value. This training
process continued until the participant maintained 90% agreement with
correct scores.
To calculate IA, the experimenter used a total agreement approach
for both correct and incorrect words (Kennedy, 2005). To calculate total
agreement per observation, the experimenter divided the larger amount of
observed words, read either correctly or incorrectly, by the smaller
amount of observed words, read either correctly or incorrectly
respectively. Overall, average total agreement measured 88.1%.
Procedural Integrity
The same independent observer, who calculated IOA, assessed
procedural integrity on 25% of the sessions. To calculate procedural
integrity, the observer reviewed each session previously used for IA and
completed a checklist verifying the specific steps of the procedure.
Average procedural integrity came to 99.7%.
Social Validity
To measure social validity parents and participants provided
information pertaining to the procedures and outcomes of the study. The
questions focused on involvement in the study. Participants orally
responded to three questions: (a) Did you enjoy participating in this
study? Why or why not, (b) Which reading procedure did you like better,
1 minute (i.e., RRFC) or 10 seconds (i.e., IS)? and (c) Which one did
you feel helped you more, 1 minute (i.e., RRFC) or 10 seconds (i.e.,
IS)? After receiving training in both procedures, parents responded with
written answers to four questions at the conclusion of the study: (a)
Did you feel that your child benefited from their participation? (b) Did
you notice anything different about your child as they progressed
through the study? (c) If possible, would you like to see this type of
"reading" done in their classroom at school? (d) Would you do
this with your child by yourself?
Results
Figures 1,2, and 3 display the correct and incorrect words read per
minute by Brad, Joseph, and Jason, respectively, during both phases. The
solid black dots represent correctly read words and the solid black
squares represent incorrectly read words for each assessment passage of
the repeated reading to a fluency criterion (RRFC) condition. The open
black dots depict correctly read words and the open black squares
represent incorrectly read words for each assessment passage of the
Interval Sprinting (IS) condition.
Unpracticed Initial Readings. The first two readings of the 1st
phase served two purposes: determining appropriate passage grade level
and an unpracticed initial reading score. Each participant scored in the
instructional range on grade level passages (i.e., Brad and Joseph, 4th
grade, and Jason, 2nd grade). Prior to the 2nd phase, two data points
show each participant's unpracticed initial readings of the next
two passages. A comparison between these two sets of data points for
each participant reveals two important effects. First, each student
scored approximately the same both within and between phases showing
similarities in difficulty across the passages. Second, the average
words correct on unpracticed initial readings increased by approximately
four words from the 1st to 2nd phase.
First Phase. Two of the three participants first met the fluency
criterion under the IS condition. Jason read 211 correct with 1 error
during the 14th session and Joseph read 209 correct with 1 error during
the 9th session. Meeting criterion first under the RRFC condition, Brad
read 229 correct with 2 errors during the 13th session. While both
Joseph and Jason did not exceed 200 correct words per minute until the
session they met criterion, Brad surpassed the 200 correct per minute
word mark during the 4th session. Although, he still read with an
average of 13 incorrects per minute with a range of 7-17. Not including
the session where he met criterion, Brad read 200 words per minute or
more during seven sessions under the RRFC condition and three times
under the IS condition. However, he did not drop below six errors in
either condition until meeting criterion during the 13th session. Brad
continued to make the same errors regardless of error correction and
feedback. Similar to the Staubitz et al. (2005) study, a change to the
procedures appeared necessary. While Staubitz et al. lowered the timing
from 60 to 20 seconds to assist a participant in reaching the fluency
criterion; Brad could not have a lowered time interval and maintain the
integrity of comparisons with other participants. Therefore, the
experimenter made two changes to the procedures for Brad after the 11th
session.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
For the first change, the experimenter adjusted Brad's
accuracy criterion from 99% (i.e., 200 or more correct and 2 or less
incorrect) to 98% correct provided he read more than 200 words correct
per minute. For the second change, the experimenter highlighted 10 high
frequency errors on each of the two passages. Brad then read from these
highlighted passages during the first 2 minutes of each condition.
During the test in each condition, however, the experimenter removed the
highlighted copy and had Brad read from a clean copy of each passage.
The experimenter made these changes prior to the 11th session. Brad
reached the fluency criterion three sessions later.
The data paths of correct words read per minute for Brad (Figure 1)
overlapped during only one session with RRFC condition outperforming the
IS condition showing a strong separation. The data paths of correct
words per minute for both Jason (Figure 2) and Joseph (Figure 3) showed
very little separation. Errors for Brad not only appear variable, but
also overlap 40% of the time (Figure 1), while errors for Joseph
decreased steadily in both conditions (Figure 2). Jason's data show
an increasing trend of errors within the RRFC condition while errors in
the IS condition maintained low levels (Figure 3). Towards the end of
the 1st phase, Jason made more errors as he read the RRFC passage
faster.
On average, each participant had 22.6 minutes of reading practice
for the two conditions (i.e., two minutes per condition, two conditions
per session, ranging from 9-13 sessions for an average of 11.3
sessions). During the RRFC condition, the participants read an average
of 169 (range 151-203) correct and 4 (range 2-9) incorrect words per
minute. The participants read an average of 186 (range 169-215) correct
and 7 (range 4-11) incorrect words per minute during the IS condition.
Participants read an average of 17 more words correctly and 3 more
incorrectly in the IS condition in an identical amount of reading time.
Second Phase. Unlike the 1st phase, two of three participants met
fluency criteria on the passage in the RRFC condition first. Brad met
the fluency criterion during the 22nd overall session, or the 8th
session within phase 2, with a score of 206 correct and 4 incorrect per
minute (i.e., 98% correct with more than 200 correct words per minute).
Jason took 9 sessions, or the 24th overall session, to read 212 correct
with 2 errors. Joseph, reaching criterion in the IS condition first,
read 200 correct with 1 error during the 9th, or 20th overall, session.
When comparing number of sessions, each participant met criteria as fast
as or faster than the 1st phase by an average of approximately three
sessions.
Similar to the 1st phase, both Joseph and Jason read 200 or more
words correctly only in the session they met criterion. Unlike the 1st
phase, Brad read more than 200 correct words per minute twice other than
the session he met criteria. Even though Brad had fewer incorrects
during both conditions in the 2nd phase, he again started to make
consistent errors regardless of error correction. To minimize the
frustration Brad experienced during the 1st phase and maintain
uniformity, the experimenter highlighted five words consistently read
incorrectly on both passages prior to the 20th session and maintained
the same accuracy criterion (i.e., 98% correct).
The data paths during the 2nd phase showed more separation across
participants compared to the 1st phase. For Brad and Joseph, the reading
scores during the IS condition outperformed the RRFC condition in all
but two sessions each showing a moderate separation. Jason's data
show the opposite. His data paths (Figure 3) display a strong separation
as he read more correct words each session during the RRFC condition.
Brad (Figure 1) and Jason (Figure 3) demonstrated stable, variable
errors across the phase. After a jump in error frequency, Joseph's
data again showed the downward trend apparent in the previous phase
(Figure 2). All participants decreased average errors per minute from 5
to 3.4 during the RRFC from the 1st to 2nd phase. Participant averages
during the IS condition showed a slight increase (i.e., 4.2 to 4.5) in
errors from the 1st to 2nd phase
On average, each participant had 17.3 minutes of reading practice
for the two conditions (i.e., 2 minutes per condition, two conditions
per session, ranging from 8-9 sessions for an average of 8.6 sessions).
During the RRFC condition, the participants read an average of 148
(range 143-158) correct and 4 (range 2-6) incorrect words per minute.
The participants read an average of 177 (range 170-183) correct and 6
(range 3-6) incorrect words per minute during the IS condition.
Participants did read less total words per minute from the 1st to 2nd
phase (i.e., RRFC, 169 to 148, and IS, 186 to 177). However,
participants averaged more words during both IS condition compared to
both RRFC conditions. Participants read 17 more correct and 3 more
incorrect during the 1st phase compared to 29 more correct and 2 more
incorrect during the 2nd phase.
Social Validity All participants stated that they enjoyed taking
part in the study. Some of the responses indicated that they liked
reading to the experimenter and reading quickly and accurately. Each of
the participants preferred the RRFC condition to the IS condition and
only one of the three participants said the IS condition helped him
more. The other two felt the RRFC condition helped them more.
Each parent reported feeling their child benefited from the study
and noticed some specific changes in their behavior. The parents
commented that their children seemed more motivated to read, didn't
get bored or distracted when reading, and had fun. Also, each parent
responded that they would use both procedures at home and would like to
see them used at school.
In addition to the questionnaire, some parents provided unprompted
information relevant to the study. One parent explained that his child
now reads television captions aloud accepting error correction. As the
study progressed, another parent expressed that her child started to
read aloud daily. Another parent noted how his child used to dislike the
"buzz" from the timer and could not stand timed readings. He
went on to say that his child now liked reading with a timer and telling
them how many words he read.
Discussion
This study compared repeated readings to a fluency criterion (RRFC)
with interval sprinting (IS). The experimental question asked which of
the two procedures facilitated attainment of the fluency criterion of
200 or more correct words per minute with 2 or less errors. The data
suggest little difference between the methods as participants met
criterion under both procedures an equal number of times. Additionally,
the data paths contained within the alternating treatment design, while
showing strong separation during isolated phases, did not produce
consistent differences across participants presenting unclear
differences between the two procedures.
One possible reason for the similar results may derive from on the
practice-totest timings. Participants reading in the RRFC condition
practiced and tested on the same one-minute metric. In order to draw
balanced comparisons, participants reading in the IS also tested for
one-minute never having the opportunity to practice the passage for one
minute. Therefore, their one-minute assessment scores may have
represented the fastest they could have read that session for one minute
regardless of condition given three factors. First, participants read
grade-level equivalent passages. Second, participants read approximately
the same number of words per minute each test session. Third, both
procedures reported effective gains.
The almost equal, yet effective results mirror the findings from
other studies reporting repeated readings to a fluency criterion gains
(e.g., Staubitz et al., 2005; Therrien & Kubina, 2007; Yurick et
al., 2006). These results systematically replicate and extend the
literature base for RRFC and also provide preliminary support for the
effectiveness of the IS procedure.
During the 2nd phase, both methods contributed to higher average
initial reading scores and all participants reached criterion in the
same or fewer number of sessions to criterion in the 2nd phase. This
finding provides support for the assertion that RRFC and IS lead to
transfer in other passages (Faulkner & Levy, 1994). Samuels (1979)
first observed this phenomenon with repeated reading: higher initial
reading scores on successive passages and shorter amounts of time to
criterion. Rashotte and Torgesen (1985) found transfer occurred once 50%
of the words overlapped between passages. Faulkner and Levy (1994) noted
that for poor readers content or words needed to overlap with overlap
defined as approximately 72% of shared words. The passages in the
current study had no content overlap and relatively low word overlap
(i.e., approximately 30%), yet the participants displayed meaningful
transfer during the 2nd phase. Reading transfer for these participants
may have resulted from one or a combination of factors within the study:
High fluency criterion (i.e., 200+ words per minute), practicing at a
high criterion for correct words per minute with a low error criterion
(i.e., 2 or less errors), and/or the synergy of two effective fluency
procedures.
The majority of reading fluency studies had students typically read
at rates lower than 200 or more correct words per minute. For example,
Rashotte and Torgesen (1985) reported mean word per minute increases
from 54 to 90 in one condition and 57 to 90 in the other condition over
four readings. Faulkner and Levy (1994) did not display first passage
times, but reported reading times for the transfer passage. Looking at
poor readers in 6th and 3rd grade, mean word per minute rates calculated
to a range of 72 to 100 for poor readers in 3rd and 6th grades across
all conditions. Participants reaching fluency criterions of 200 or more
correct words per minute may need neither word nor content overlap to
display transfer.
The rate of 200 words per minute used in the current study,
especially for 2nd and 4th graders, exceeds most of the rates published
in previous research. For example, in their review of repeated readings,
Meyer and Felton (1999) reported desirable oral reading word rates for
2nd graders at 85 to 115 words per minute and 120 to 150 for 5th
graders. Hasbrouck and Tindal (2005) published similar mean scores of 89
words per minute for 2nd graders and 125 for 4th graders. With respect
to the score of 200 words per minute, Hansbrouk and Tindal reported that
rate as a 90th percentile reading score for 6th graders while Meyer and
Felton state 200 words per minute as a mean high school silent reading
rate. This study showed that participants with or at risk for reading
disabilities reached an oral reading rate of 200 words per minute after
only an average of 33 minutes of practice during the 1st phase and 24
minutes of practice on during the 2nd phase with grade level material.
Therefore, setting a high fluency criterion not only seems feasible, but
also demonstrated the positive effect of transfer to new reading
passages with low word overlap and no related content.
Another factor possibly contributing to transfer involves
participants consistently practicing to not only a higher frequency of
words read per minute, but the inclusion of a very high accuracy
criterion. Previous researchers (Staubitz et al., 2005; Yurick et al.,
2006) had students reach rates of 145 words per minute with 10 or fewer
errors for 4th graders to a range of 145-180 words per minute with 10 or
fewer errors for 5th graders. Therefore, at minimum, these students met
criterion with 93-95% accuracy. In the current study, 2nd and 4th
graders reached criterion with minimum 98-99% accuracy. Obtaining
fluency, or what Binder (1996) calls true mastery or behavioral fluency,
with a high oral reading rate may help positively shape reading
performance. These new patterns of reading performance, once learned,
may have resulted in a more careful and sharpened approach to decoding
text thereby promoting students' reading transfer between unlinked
passages.
The third factor connected to reading transfer may have resulted
from the synergy of the two effective fluency building procedures.
Considering the effectiveness of the previously mentioned two factors,
each student may have benefited from experiencing each condition every
session. This also may explain why neither method distinguished itself
through the alternating treatment design.
Focus on Interval Sprinting
All three participants demonstrated endurance with oral reading
during the interval sprinting procedure. Each participant practiced
reading for a series of 10-second timings then read for one minute. The
one-minute assessment constituted an unpracticed, much longer reading
time. In previous examples (Bourie, 1980; Desjardins, 1981) and dis
cussions (Binder, 1996) of endurance building, "sprinting"
referred to achieving high rates of accurate behavior for short
intervals of time, then lengthening the intervals (Binder et al, 1990).
During the IS condition, participants never practiced for longer than 10
seconds consecutively per timing yet displayed one-minute test scores
equal to those during the RRFC condition. By distributing the 10-second
sprints throughout the passage, participants showed that they could read
as well for one minute with the added benefit of coveting more words
during practice.
The additional word practice resulted from participants routinely
reading above rates of 34 words per 10 seconds or more than 204 words
per minute. This, as mentioned earlier, allowed students to make more
efficient use of the allocated practice time. The inclusion of
additional IS practice passages across time may have shown the results
of this reading productivity. Using the average increase of words during
the two phases (23) and extrapolating over 150 school days, students
could practice 3,450 more words (i.e., almost 14 typed, double-spaced
pages) given the same practice time. In the context of cumulative
practice, especially under guided practiced conditions, students may
find future performance affected by practicing a few additional words
daily (i.e., more correct reading leads to more correct reading).
To experimentally balance the IS and RRFC conditions, the
experimenter limited feedback within the IS procedure to two specific
places, after the 6th and 12th sprints. When applied in practical
situations, practitioners would not withhold error correction and
feedback until these points. With the IS procedure, teachers have the
opportunity to provide 12 distinct periods of feedback and error
correction (i.e., after each sprint) compared to the 2 used in the
present experiment. Such feedback could conceivably reduce student
errors more quickly, fostering high quality practice sessions.
Social Validity
Each student stated they preferred the RRFC over the IS condition.
The students' comments fall in line with the matching law
(Hernstein, 1997; Martens, 1992). The matching law states that when
confronted with a choice, people will chose those behaviors that meet
with the most reinforcement (Martens). Therefore, fluent readers may
like to read passages and for longer periods of time, relatively
speaking. These instances of reading allow the reader to meet with more
success (i.e., reinforcement). One-minute timings consist of more
continuous opportunities to read as compared to 10 seconds of reading.
Additionally when presented with 12 tasks (i.e., 10-second trials)
versus 2 tasks (i.e., one-minute trials), it seems better to choose the
lesser one.
Study Limitations
The following limitations in the study require attention. First,
the varying order of practice sessions and calculation of word overlap
helped control for some confounds associated with an alternating
treatment design, but not all. The presentation of only one of the two
practice procedures per session alternatively would have helped control
for carryover effects. Second, the omission of a control condition
presents another limitation. The possible separation between data paths
for a third condition (i.e., reading a passage silently or aloud for 2
minutes) and the other two conditions may have better highlighted IS and
RRFC gains. Also because two participants read the same four passages,
simply varying the presentation order of those four passages may have
further demonstrated experimental control.
A third limitation for this study involves the format of materials
during the IS condition. The word total numbering system on each
experimenter copy represented a running total of words rather than a per
line measure. This greatly assisted tallying one-minute reading scores,
but did not help when counting 10-second sprints. The experimenter had
to manually count words read during sprints, artificially increasing the
time between sprints or adding overall time to the IS procedure. Simply
changing the numbering system would have decreased pauses.
Future Research Directions
Based on the results of this study, the interval sprinting
procedure warrants further examination. One such examination could
consist of exploring the relationship between reading transfer and high
fluency criterion measures used with the IS procedure as the only
condition. Maintaining a criterion of 200 correct and 0-2 incorrect
words per minute, research may also test the IS procedure with both
older and younger children with varying disabilities. Passage type may
also vary. For example, this study displayed the effect of the IS
procedure with stand-alone passages. Future research can also examine
whether more interval sprints per passage (i.e., three to four rather
than two) may facilitate students reaching the criterion more quickly
and increasing transfer.
Conclusions
Almost 25 years ago, Allington (1983) suggested that building oral
reading fluency helps create "effective and efficient readers"
(p. 561). Methods that build fluency should meet the same effective and
efficient criteria. The data presented in this study suggest interval
sprinting leads to the attainment of oral reading fluency in children
with or at-risk for reading disabilities. Using 10-second practice
trials, interval sprinting not only mirrored the results when
participants used repeated readings to a fluency criterion, but
participants displayed more reading production given the same reading
time. With the ever increasing importance placed on effective and
efficient fluency building procedures for reading, interval sprinting
may prove a valuable addition to the continuum of fluency-building
practice procedures.
References
Allington, R. L. (1983). Fluency: The neglected reading goal. The
Reading Teacher, 36, 556561.
Archer, A. L., Gleason, M. M., & Vachon, V. L. (2003). Decoding
and fluency: Foundation skills for struggling older readers. Learning
Disability Quarterly, 26, 89-101.
Begeny, J. C., & Silber, J. M. (2006). An examination of group
based treatment packages for increasing elementary aged students'
reading fluency. Psychology in the Schools, 43, 183195.
Binder, C. (1996). Behavioral fluency: Evolution of a new paradigm.
The Behavior Analyst, 19, 163-197.
Binder, C., Haughton, E., & Van Eyk, D. (1990) Increasing
endurance by building fluency: Precision teaching attention span.
Teaching Exceptional Children 22(3), 24-27.
Bourie, C. (1980). Micro-sessions. Data Sharing Newsletter, 28, 2.
Carnine, D. W., Silbert, J., Kame'enui, E. J., & Tarver,
S. G. (2004). Direct Instruction Reading (4th ed.). Upper Saddle River,
NJ: Pearson Merrill Prentice Hall.
Chard, D. J., Vaughn, S., & Tyler, B. (2002). A synthesis of
research on effective interventions for building reading fluency with
elementary students with learning disabilities. Journal of Learning
Disabilities, 35, 386-406.
Compan, B., Iamsupasit, S., & Samuels, J. (2001). Effect of
repeated reading and self-directed behavior on reading skills and
generalization of reading skills of third-grade Hill tribe students
(Report No. CS-014-494). (ERIC Document Reproduction Service No.
ED458531).
Deno, S. L. (2003) Developments in curriculum-based measurement.
Journal of Special Education, 37, 184-192.
Desjardins, A. (1981). On endurance and ten-second timings. Data
Sharing Newsletter, 34, 3.
Faulkner, H. J., & Levy, B. A. (1994). How text difficulty and
reader skill interact to produce differential reliance on word and
content overlap in reading transfer. Journal of
Experimental Child Psychology, 58, 1-24.
Fry, Edward. (1989). Readability formulas: Maligned but valid.
Journal of Reading, 32, 292-297.
Fuchs, L. S., Fuchs, D., Hosp, M. K., & Jenkins, J. R. (2001).
Oral reading fluency as an indicator of reading competence: A
theoretical, empirical, and historical analysis. Scientific Studies in
Reading, 5, 239-256.
Freeman, G., & Haughton, E. (1993). Building reading fluency
across the curriculum. Journal of Precision Teaching, 10(2), 29-30.
Gibala, M. J., Little, J. P., van Essen, M., Wilkin, G. P.,
Burgomaster, K. A., Safdar, A., Raha, S., & Tarnopolsky, M. A.
(2006). Short-term interval versus traditional endurance training:
Similar initial adaptations in human muscle and exercise performance.
Journal of Physiology, 575(3), 901-911.
Good, R. H., & Kaminski, R. A. (Eds.). (2007). Dynamic
Indicators of Basic Early Literacy Skills (6th ed.). Eugene, OR:
Institute for the Development of Educational Achievement. Available:
http://dibels.uorgon.edu/
Good, R. H., III, Simmons, D. C., & Kame'enui, E. J.
(2001). The importance and decisionmaking utility of a continuum of
fluency-based indicators of foundational reading skills for third-grade
high-stakes outcomes. Scientific Studies of Reading, 5, 257-288.
Hasbrouck, J., & Tindal, G. (2005). Oral Reading Fluency: 90
Years of Measurement (Technical Report No. 33). Eugene, Oregon:
University of Oregon, College of Education, Behavioral Research
and
Teaching. http://brt.uoregon.edu/tech reports .htm
Haughton, E.C. (1980) Practicing practices: Learning by activity.
Journal of Precision Teaching, 1(3), 3-20.
Herrnstein, R. J. (1997). The matching law. Cambridge MA: Harvard
University Press.
Hiebert, E. H., & Fisher, C. W. (2005). A review of the
national reading panel's studies on fluency: The role of text.
Elementary School Journal, 105, 443-460.
Hosp, M. K., & Fuchs, L. S. (2005). Using CBM as an indicator
of decoding, word reading, and comprehension: Do the relations change
with grade? School Psychology Review, 34, 9-26.
Johnson, J. M., & Pennypacker, H. S. (1993). Strategies and
Tactics of Behavioral Research (2nd ed.). Hillsdale, NJ: Lawrence
Erlbaum Associates.
Johnson, K. R., & Layng, T. V. J. (1992). Breaking the
Structuralist Barrier. American Psychologist, 47, 1475-1490.
Kavale K. A. (2005). Effective intervention for students with
specific learning disability: The nature of special education. Learning
Disabilities, 13, 127-138.
Kennedy, C. H. (2005). Single-case designs for educational
research. Boston: Pearson Allyn and Bacon.
Kubina, R. M. (2005). Developing reading fluency through a
systematic practice procedure. Reading & Writing Quarterly, 21,
185-192.
Kubina, R. M., & Morrison, R. S. (2000). Fluency in education.
Behavior and Social Issues, 10, 83-99.
Kubina, R. M., & Starlin, C. (2003). Reading with precision.
European Journal of Behavior Analysis, 4(1 & 2), 13-22.
Kuhn, M. R. (2005). A comparative study of small group fluency
instruction. Reading Psychology, 26(2), 127-146.
Kuhn, M. R., & Stahl, S. A. (2003). Fluency: A review of
developmental and remedial practices. Journal of Educational Psychology,
95, 3-21.
LeBlanc, L. A., Coates, A. M., Daneshvar, S., Charlop-Christy, M.
H., Morris, C., & Lancaster, B. M. (2003). Using video modeling and
reinforcement to teach perspective-taking skills to children with
autism. Journal of Applied Behavior Analysis, 36, 253-257.
Martens, B. K. (1992). Contingency and choice: The implications of
matching theory for classroom instruction. Journal of Behavioral
Education, 2, 121-137.
Meyer, M. S., & Felton, R. H. (1999). Repeated reading to
enhance fluency: Old approaches and new directions. Annals of Dyslexia,
49, 283-306.
Mounsteven, J. (1990). Speed reading: A technique for developing
fluent readers. Teaching Exceptional Children, 22(3), 69-71.
National Reading Panel (2000). Teaching children to read: An
evidence-based assessment of the scientific research literature on
reading and its implications for reading instruction. Washington, DC:
National Institute of Child Health and Human Development.
Nelson, J. S., Alber, S. R., & Gordy, A. (2004). Effects of
systematic error correction and repeated readings on the reading
accuracy and proficiency of second graders with disabilities. Education
& Treatment of Children, 27, 186198.
Rashotte, C. A., & Torgesen, J. K. (1985). Repeated reading and
reading fluency in learning disabled children. Reading Research
QuarterIv, 20, 180-188.
Samuels, S. J. (1979). The method of repeated readings. The Reading
Teacher, 32, 756-760.
Schreiber, P.A. (1980). On the acquisition of reading fluency.
Journal of Reading Behavior, 12, 177-186.
Scott, T. M., & Shearer-Lingo, A. (2002). The effects of
reading fluency instruction on the academic and behavioral success of
middle school students in a self-contained EBD classroom. Preventing
School Failure, 46, 167-173.
Shinn, M. R. (1989). Curriculum-based measurement: Assessing
special children. New York: Guilford.
Shinn, M. R., & Good, R. H. (1992). Curriculumbased measurement
of oral reading fluency: A confirmatory analysis of its relation to
reading. School Psychology Review, 21, 459-479.
Staubitz, J. E., Cartledge, G., Yurick, A. L., & Lo, Y. (2005).
Repeated reading for students with emotional or behavioral disorders:
Peer- and trainer-mediated instruction. Behavioral Disorders, 31, 51-64.
Therrien, W. J. (2004). Fluency and comprehension gains as a result
of repeated reading: A meta- analysis. Remedial and Special Education,
25, 252-261.
Therrien, W. J., & Kubina, R. M. (2006). Developing reading
fluency with repeated reading. Intervention in School and Clinic, 41,
156-160.
Therrien, W. J. & Kubina, R. M. (2007). The importance of
context in repeated reading. Reading Improvement, 44, 179-188.
Therrien, W. J., Wickstrom, K., & Jones, K. (2006). Effect of a
combined repeated reading and question generation intervention on
reading achievement. Learning Disabilities Research & Practice, 21,
89-97.
Valleley, R. J., & Shriver, M. D. (2003). An examination of the
effects of repeated readings with secondary students. Journal of
Behavioral Education, 12, 55-76.
Vickers, J. N. (2006). Gaze of Olympic speedskaters skating at full
spend on a regulation oval: Perception-action coupling in a dynamic
performance environment. Cognitive Process, 7 (Suppl.1), S 102-S105.
Young, A. R., Bowers, E G., & MacKinnon, G. E. (1996). Effects
of prosodic modeling and repeated reading on poor readers' fluency
and comprehension. Applied Psycholinguistics, 17, 59-84.
Yurick, A. L., Robinson, P. D., Cartledge, G., Lo, Y., & Evans,
T. L. (2006). Using peer-mediated repeated readings as a
fluency-building activity for urban learners. Education & Treatment
of Children, 29, 469-506.
DOUGLAS E. KOSTEWICZ
University of Pittsburgh
RICHARD M. KUBINA JR.
The Pennsylvania State University