**3. Results**

#### *3.1. Participant Characteristics*

In the 16 studies reviewed, a total of 1112 students were included—506 second grade, 207 third grade, 120 fourth grade, and 76 fifth grade. Additionally, 57 participants were identified as primary grade students and 139 were identified as upper elementary students but further grade classification was undeterminable, and seven students were enrolled in sixth through eighth grade. The studies included 460 males, 419 females, and 233 students whose gender was unidentified. Ethnicity was reported for 75.5% of the students. Of those reported, 304 were African American, 221 were Caucasian, 208 were Hispanic, 63 were Asian, and 44 were mixed races or other. Across studies, the number of English Language Learners (ELLs) were reported in seven papers [48,50,54,55,57–59], with 295 students identified as ELLs. The number of students serviced through special education was also reported in seven studies [49,54–59]. A total of 140 students were identified as receiving special education services.

#### *3.2. E*ff*ective Oral Reading Fluency Interventions*

The 16 studies included in this review showed that repeated reading with multiple features (e.g., choral reading, verbal cueing, error correction) was most often empirically examined in order to determine the e ffectiveness on improving the oral reading fluency for elementary students with reading di fficulties. Assisted reading, continuous reading, and the neurological impress method were also examined to a lesser extent and often in comparison to a repeated reading condition and control group. Across the 16 studies, one study reported no statistically significant gains for oral reading fluency and e ffect sizes for rate and accuracy measures in the remaining 15 studies varied widely, with negligible to large e ffects found (ES = 0.01 to 1.18). Three studies measured prosody outcomes and all found large, significant e ffects [51,60,61]. E ffect sizes for reading comprehension also varied across papers. Of the nine studies from which e ffect sizes for reading comprehension outcomes could be calculated, one study reported no statistically significant gains [58], and eight studies had negligible to large e ffects on reading comprehension (ES = 0.07–2.59) [47,48,51,54,55,57,59,61].

#### 3.2.1. Repeated Reading

Fourteen (87.5%) of the 16 studies reviewed included a repeated reading intervention [46–48,51–61]; however, none of the studies examined a stand-alone repeated reading procedure. All of the interventions described included supplementary components in addition to the repeated reading procedures. The multiple features seen across studies included (1) a peer coach who was a classmate

identified as more capable and fluent reader, (2) teacher modeling in which the adult served as a model of fluent and prosodic reading, (3) phrase-drill error correction which had the student say the phrase from the text containing each error three times after correct pronunciation was modeled by the instructor after an initial reading, (4) error correction that provided students during or after reading with a missed word and had them repeat the word after the instructor, (5) performance feedback in which the teacher provided information on students' prosody after reading, (6) verbal cueing which provided students with a reminder such as "Read this story the best you can and as quickly as you can.", (7) goal setting which had students set a WCPM goal prior to reading, (8) echo reading which asked the student to mimic the fluent and expressive reading of the instructor, (9) choral reading in which both the teacher and students read the text aloud together in unison, (10) performance of text, (11) listening passage preview which allowed students to listen and follow along while an adult read the text aloud with expression, (12) question answering after reading to promote comprehension, (13) neurological impress method in which the teacher read aloud slightly ahead of the student, and (14) phonics or vocabulary instruction. The number of multiple features included in repeated reading interventions varied widely across studies (See Table 1).

*One-on-one with the teacher.* In seven of the studies [47,48,53,54,56,60,61], students participated in one-on-one sessions of a repeated reading with multiple features intervention with an adult (e.g., teacher, paraprofessional, trained research assistant). With regard to oral reading fluency, two studies were found to have negligible e ffects on students' rate and accuracy by the authors [53,61]. The remaining five studies were found to have small to large e ffect sizes (i.e., 0.38 to 1.18) on standardized measures of oral reading fluency (i.e., Dynamic Indicators of Basic Early Literacy Skills Oral Reading Fluency, Gray Oral Reading Test-Fluency) [47,48,54,56,60]. Four studies included measures of reading comprehension and found moderate to large e ffect sizes (i.e., 0.70–1.47) [47,48,54,61].

Specifically, Begeny, Mitchell, Whitehouse, Harris, and Stage as well as Begeny, Ross, Greene, Mitchell, and Whitehouse provided one-on-one repeated reading intervention three times a week to second graders who were native speakers of English and English Language Learners, respectively [47,48]. In both studies, the intervention produced large e ffect sizes (1.18 and 0.95) on students' oral reading fluency, as well as moderate and large e ffects on reading comprehension (ES = 0.70 and 1.12), respectively.

Young, Mohr, and Rasinski, examined the e ffects of a hybrid intervention combining repeated reading and the Neurological Impress Method (NIM) with third through fifth graders [60]. Students in the hybrid intervention condition demonstrated a moderately significant increase in rate and accuracy (ES = 0.68). In contrast, Young, Pearce, Gomez, Christensen, Pletcher, and Fleming tested the same hybrid intervention on a group of first- through third-grade students [61]. While negligible e ffects on oral reading fluency were found in this sample (ES = 0.07), moderate to large e ffects were found for reading comprehension (ES = 0.77 for comprehension questions, ES = 1.47 for passage retell). In addition to the intervention group and the no intervention control group, Young et al. also included a NIM-only group [61]. The students in the NIM only group were not provided with the opportunity to reread the text. Similar to the hybrid intervention condition, no e ffects on students' rate and accuracy were found by the authors. Further, the authors reported that the NIM-only condition had a nonsignificant e ffect on the passage retell aspect of reading comprehension but improved students' ability to answer comprehension questions after reading, as evidenced by a large, statistically significant effect size (ES = 0.93).

O'Connor, White, and Swanson provided second- and fourth-grade students with a repeated reading intervention that included error correction from the teacher [54]. The intervention, which lasted 14 weeks, produced a moderate e ffect (ES = 0.53) on students' oral reading fluency and a moderate to large e ffect on reading comprehension (ES = 1.09 on Woodcock Reading Mastery Test-Passage Comprehension and ES = 0.75 on Gray Oral Reading Test-Comprehension).

Therrien, Wickstrom, and Jones used a repeated reading intervention with fourth- through eighth-grade students which included a series of instructional steps such as verbal cueing and error correction [56]. The intervention was found to produce a small e ffect size (ES = 0.38) on students' oral reading fluency.

Finally, Martens et al. examined an after-school fluency training program in which second and third graders received a repeated reading intervention provided by graduate and undergraduate students [53]. Based on grade level CBM probes, negligible e ffects on students' oral reading fluency were found (ES = 0.14 for 2nd graders on 2nd grade probe and 0.06 for 3rd graders on 3rd grade probe). While the authors investigated rate and accuracy two days after the intervention ended, this was only completed for the intervention group and consequently, not enough information was provided to be able to calculate e ffect sizes associated with between-group di fferences.

*One-on-one with a peer coach*. Three studies examined the use of a repeated reading intervention in which students worked one-on-one with a peer coach [46,52,55]. Two of the three studies produced large e ffects on oral reading fluency outcomes [50,54].

In both Algozzine, Marr, Kavel, and Dugan and Marr, Algozzine, Nicholson, and Dugan [46,52], students worked together for 10–12 min, three times a week for 36 weeks. While both Algozzine et al. and Marr et al. found students in the intervention group had statistically significant gains in oral reading fluency from pre-test to post-test, statistically significant gains were also found for students in the no-intervention control group although at a lower rate. However, in both Algozzine et al. and Marr et al., repeated reading with a peer coach was found to produce large e ffect sizes on students' oral reading fluency (e ffect size = 1.06 and 1.12, respectively). While reading comprehension growth in both studies was measured through an assessment created and used by the participating school district, data was only reported for the intervention group. Therefore, no e ffect sizes associated with between-group di fferences could be calculated for reading comprehension in these studies.

Similarly, Saenz, Fuchs, and Fuchs investigated Peer-Assisted Learning Strategies (PALS), with 20 special education and 33 low-achieving third- through sixth-grade students from bilingual classrooms [55]. Low-achieving students in this study were those in the lowest quartile of the class rank based on classroom observations, previous scores on state competence exams, and district informal reading inventories. In the PALS intervention, students worked in pairs in order to complete three readings of a text. With regard to oral reading fluency, a small e ffect (ES = 0.33) was found for ELL students with learning disabilities who participated in the PALS program three times a week for 15 weeks. Conversely, the same intervention demonstrated a negligible e ffect (0.01) for ELL students who were identified as low achieving. Further, in terms of reading comprehension, both ELL students with learning disabilities as well as ELL students who were identified as low achieving had large, statistically significant gains (ES = 1.15 for ELL+LD and 0.83 for ELL+ low achieving).

*Small group with a teacher*. Four studies had students working with a teacher in small groups of peers with similar needs and negligible to small e ffects were found with regard to oral reading fluency and reading comprehension outcomes [51,57–59]. Kuhn examined the e ffects on oral reading fluency and reading comprehension of (1) a repeated reading intervention with multiple features, (2) a listening-only intervention, (3) a continuous reading intervention, and (4) a no-intervention control group [51]. After six weeks of intervention, Kuhn concluded that students in the repeated reading condition outperformed students in the listening-only condition as well as the control group, although calculated e ffect sizes were trivial (ES = 0.17 and 0.14, respectively). Additionally, non-significant gains in reading comprehension were found in the repeated reading condition, the listening-only condition, and the control group.

Vadasy and Sanders examined the same repeated reading intervention, QuickReads, with secondand third-grade students, including English Language Learners and special education students, across multiple sites [58,59]. The only di fference between the studies was that Vadasy and Sanders first examined the e ffects of the intervention delivered by a trained paraprofessional and later examined the e ffects of the intervention delivered by a classroom teacher [58,59].

In both studies [58,59], the intervention procedures first had the instructor complete a brief letter–sound practice with the students due to the large number of students in the study with low word reading skills. Next, throughout the repeated reading procedure, instructors provided students with error correction strategies for students' miscues as needed (e.g., encouraging the student to sound out the word, phoneme by phoneme, and then blend the sounds; helping the student to segmen<sup>t</sup> a multisyllable word and then put the parts together). Both studies found the repeated reading intervention, QuickReads, to produce a small effect size on students' oral reading fluency when delivered by a trained paraprofessional (ES = 0.41 and 0.32, respectively) and a moderate effect when delivered by a classroom teacher (ES = 0.53) [58,59]. While Vadasy and Sanders found that the intervention employed by a paraprofessional did not produce a statistically significant difference in reading comprehension over the control group [58], the researchers determined that QuickReads delivered by a classroom teacher produced a small effect (ES = 0.36) [59].

Lastly, Vadasy and Sanders also examined a repeated reading intervention which included 199 fourth- and fifth-grade students with reading difficulties [57]. However, differing from their other studies [58,59], this intervention began and ended sessions with vocabulary instruction rather than letter–sound practice and, consequently, no error correction strategies were provided to students as they read. At the conclusion of the study, the researchers found that there were no significant treatment effects for oral reading fluency. Further, the average oral reading fluency rate of students in the intervention group remained in the lowest quartile, based on Hasbrouck and Tindal's norms for fourth and fifth grade [65]. However, a small effect (0.35) was found for reading comprehension [57].

#### 3.2.2. Listening While Reading

Two studies examined the effects of an intervention which had students individually listen along to an audiobook while reading [49,50]. Esteves and Whitten's study investigated the assisted reading intervention with 20 upper elementary students, all of whom were serviced through special education with individualized education plan goals in the area of reading [49]. The researchers found that the students in the intervention group demonstrated a larger increase in the number of words read correctly from pre-test to post-test as compared to students in the no-intervention control group. A moderate effect on students' rate and accuracy was found (ES = 0.64).

Friedland, Gilman, Johnson, and Demeke also investigated the effect of listening while reading using audiobooks on a group of 46 third-grade students—almost half of which were English language learners [50]. After four weeks of intervention, Friedland and colleagues concluded that students in the treatment group demonstrated a greater improvement in the number of words read correctly per minute than students in the control group. However, a negligible effect size of 0.07 was obtained from our calculations. Interestingly, neither study of listening while reading included a measure of students' reading comprehension. Thus, we were not able to examine whether this method of fluency intervention also increased students' reading comprehension.

#### 3.2.3. Continuous Reading

A continuous reading intervention was examined in a one-on-one setting by O'Connor et al. and a small group setting by Kuhn [51,54]. Rather than rereading the same text multiple times, students in the continuous reading condition read continuously for the entirety of the intervention session and completed a single reading of a different text during each intervention session. Both studies found that continuous reading produced moderate gains in students' oral reading fluency rate and accuracy as well as moderate-to-large gains in reading comprehension

In O'Connor et al.'s study [54], second- and fourth-grade students in the continuous reading condition read a text at their instructional level (88%–94% accuracy) one-on-one with the teacher. The researchers investigated this continuous reading condition against a one-on-one repeated reading intervention as well as a no-intervention control group. Students in both intervention conditions read from the same text, only differing on whether they read the same page repeatedly or read multiple pages without repeating. Based on oral reading fluency rate scores at the end of the study, the researchers found that students in the continuous reading condition significantly outperformed

students in the no intervention control group, with moderate e ffect sizes found (ES = 0.58). Further, a moderate-to-large e ffect (ES = 0.71 on the Woodcock Reading Mastery Test-Passage Comprehension and ES = 0.95 on the Gray Oral Reading Test-Comprehension) was also found for reading comprehension. Comparing the repeated-reading-intervention and no intervention groups, moderate and large e ffect sizes for di fferences in oral reading fluency and reading comprehension were obtained, respectively (ES = 0.53 and 1.93), favoring the repeated reading condition. It is important to note, however, that students in the repeated-reading-intervention did not di ffer significantly from students in the continuous-reading-intervention on measures of oral reading fluency or reading comprehension.

Similarly, Kuhn included a continuous reading condition in her study of oral reading fluency interventions with 24 second-grade students [51]. Di ffering from O'Connor et al.'s study [54], students in Kuhn's continuous reading intervention engaged in small-group echo or choral reading of a text with the adult tutor. At the end of the study, Kuhn reported that the students in the continuous reading intervention group outperformed students in the listening-only condition as well as the no-intervention control group for fluency, with calculated e ffect sizes of 0.58 and 0.26 respectively. However, mirroring O'Connor et al. [54], there was no statistically significant di fference between students in the continuous reading and repeated reading conditions for oral reading fluency. Further, only students in the continuous reading condition demonstrated greater improvement in reading comprehension in comparison to the control group, and a large e ffect was obtained in the calculation of between-group di fferences (ES = 2.59).
