Research Article

Phonological Memory and Naming Speed Predict Response to Intervention in Elementary School-Aged Children with Word-level Reading Difficulties

Mark Uyar, John R Kirby*, Jan J Maclean and Jessie L Eriksen

Queen’s University, Canada

Submission: January 25, 2023; Published: February 21, 2023

*Corresponding author: John R Kirby, Faculty of Education, Queen’s University, 511 Union Street, Kingston, Ontario, Canada

How to cite this article: Mark Uyar, John R Kirby, Jan J Maclean and Jessie L Eriksen. Phonological Memory and Naming Speed Predict Response to Intervention in Elementary School-Aged Children with Word-level Reading Difficulties. Glob J Intellect Dev Disabil. 2023; 11(2): 555810. DOI:10.19080/GJIDD.2023.11.555810

Abstract

Background: We examined the predictors of response to intervention (RTI) in an individually administered program based on the Orton-Gillingham method. The study used a single-group design with intervention and pre- and post-testing.

Method: Prior to intervention, 50 children with word-level reading difficulties aged 5 to 12 were administered measures of phonological awareness, naming speed, and phonological memory. They also were given, both prior to and following intervention, eight reading and spelling measures.

Results: Participants made significant gains in all reading and spelling measures (p < .001). Regression analyses indicated that, after controlling participants’ age, number of sessions, and pre-test scores, pre-intervention phonological memory predicted growth in most reading and spelling outcomes (ps ranging from .05 to .001) and naming speed predicted growth in rate, fluency, and sight word efficiency (ps < .05).

Conclusion: Phonological awareness did not predict growth in any outcome, suggesting that programs which successfully target it eliminate its predictive power. Weaknesses in phonological memory and naming speed may be preventing some children from benefitting from interventions such as this one. Suggestions for mitigating these effects and implications for future research and interventions are discussed.

Keywords: Dyslexia; Orton-Gillingham; Phonological awareness; Phonological memory; Naming speed; Response to intervention

Introduction

Word-level reading difficulties are experienced by many elementary-school aged children, and when serious and persistent are synonymous with dyslexia [1]. Estimates of the prevalence of dyslexia vary considerably, from 3% to 17%, depending on the criteria employed [1]. In this paper we use the terms word-level reading difficulties or disabilities, and dyslexia interchangeably, referring to children who demonstrate word-level reading and spelling achievement that is well below their age- and grade-level expectations. Many children with dyslexia can benefit from remedial instruction [1], but some may require very extensive and intensive intervention. Because schools often may not have the resources needed, many parents engage the help of private clinics or tutors.

Despite a proliferation of research on the mechanisms that underlie reading impairment in dyslexia, relatively little has focused on predictors of response to intervention (RTI). It is important to understand the characteristics of children who are more or less likely to be successful in given programs, to select children for particular interventions and provide clues about how to improve those programs. We focus on elementary school-aged children in one program based partly on the Orton-Gillingham approach and on three predictors: phonological awareness, naming speed, and phonological memory. We use these to predict improvements in word reading accuracy, fluency, reading comprehension, and spelling. We begin by reviewing research on these predictors, then research on predicting RTI. Lastly, we examine the Orton-Gillingham approach.

Predictors of Reading

The predictors that we focus upon are phonological awareness, naming speed, and phonological memory. Because individual differences in these variables predict reading achievement [1-4], these differences may also explain why some children are slower-responders or non-responders to intervention programs [5,6]. It is important to determine which of these reading-related processes most reliably predict RTI in specific intervention programs.

Phonological awareness is the metacognitive understanding that words are composed of individual sounds [1,7], and is assessed with tests requiring participants to manipulate, segment, or blend various speech sounds [4]. Phonological awareness is a key requirement for acquiring the alphabetic principle and phonological decoding [1,2,4], and a mediator between phonological instruction and reading success [8]. Across age groups phonological awareness is a significant predictor of reading outcomes such as spelling and comprehension abilities [9] and most children with reading difficulties score lower on measures of phonological awareness [9,10]. Phonological awareness is strongly associated with the ability to decode unfamiliar words or pseudowords, contributes substantially to word reading accuracy [9], and continues to predict various reading outcomes after controlling other variables such as verbal and nonverbal ability [11]. Many reading intervention programs address phonological awareness deficits through training and practice in segmentation and blending skills, with consequential benefits for reading achievement [9,12-14]. The Orton-Gillingham approach which we describe later is one such program.

The second predictor of reading is naming speed (also known as rapid automatized naming or RAN), the speed with which a person can retrieve and pronounce the names of highly familiar stimuli, such as letters, digits, or objects [15]. Naming speed is a significant predictor of word reading accuracy, fluency, and comprehension [3,7,10,15], though it has a stronger relationship with fluency than with accuracy [3]. Wolff [16] found that naming speed significantly predicted reading fluency, whereas phonological awareness better predicted comprehension and spelling ability. The relationship between naming speed and reading survives controls of verbal IQ, processing speed, phonological awareness, phonological memory, and orthographic processing [3]. Therefore, naming speed is considered an independent factor influencing reading outcomes.

A third predictor is phonological memory, though it has been less often studied. Phonological memory involves both the phonological loop (passive short-term auditory memory) and working memory [17]. Some studies have found that children with dyslexia perform worse on phonological memory measures [10,17,18] Alloway et al., 2005) such as memory for digits and pseudoword repetition [19]. However, phonological awareness is more strongly associated with reading than is phonological memory [20]. There is also evidence that phonological memory is not strongly related to nonword reading until at least grade one [21], while other studies have failed to detect any effects of phonological memory on word identification, comprehension, or fluency [22,23]. Kastamoniti et al. [24] reviewed the literature on the relationship between phonological memory and reading, concluding that there was no clear relationship.

Response to Intervention

RTI can be defined as a child’s improvement in reading after systematic training [25,26]. Although many programs are successful in remediating reading in some children, other children make much smaller or no gains [6]. In a meta-analysis, Tran et al. found that pre-test measures accounted for over 30% of the variance in post-test scores in school-aged children at-risk for disabilities, suggesting that poor responders can be identified prior to intervention. Early identification is crucial in effective remediation and prevention because older children with reading disabilities are less likely to respond adequately compared to younger children [25].

Several studies support phonological awareness and naming speed as robust predictors of RTI. Frijters et al. [27] found that they predicted performance of 6–9-year-old children with reading disabilities across four treatment programs in word identification, nonword decoding, and fluency. Johnson & Swanson [22] showed they predicted performance among 11–14-year-old children who were responders, low responders, and non-responders after three years of instruction on word identification, fluency, and comprehension. Low phonological awareness and slow naming speed were characteristic of low or non-responders. Nelson et al. [28] found that both phonological awareness and naming speed predicted RTI in their metaanalysis of children at risk for reading disabilities.

There is mixed evidence for phonological memory as a predictor, as some studies have found it modestly predicts RTI [27,28], and others have failed to detect any effect in word identification, comprehension, or fluency [22,23]. Al Otaiba & Fuchs [9] found that in four of seven studies of young children at risk for reading disabilities, unresponsive children had lower scores on phonological memory when compared to responders. Hatcher & Hulme [29] did not find phonological memory to predict RTI targeting word identification in 7-year-old poor readers. However, Torgesen et al. [30] found that phonological memory explained improvement on pseudoword but not real word reading in children with weak phonological skills, between kindergarten and grade 2. This suggests that phonological memory may predict improvement in interventions that specifically train phonological recoding (pseudoword reading).

Orton-gillingham approach

The Orton-Gillingham program is a broad, systematic, multisensory approach to instruction for children with word level reading disabilities [31-33]. It emphasizes sequential, individualized instruction built around explicit synthetic phonics training [34], prioritizing learning letter-sound correspondences before blending these smaller units to form larger units. The Orton-Gillingham approach has many aspects in common with other reading programs, but its most distinctive feature is its requirement for multisensory instruction [33]. The Orton-Gillingham approach encompasses direct instruction in morphology, semantics, and syntax [34].

Despite the popularity of the approach [31], there is relatively little evidence for its effectiveness. Stevens et al.’s [33] meta-analysis indicated a nonsignificant .22 effect size for foundational skills (phonological awareness, decoding, word identification, fluency, and spelling) and a nonsignificant effect of .14 for vocabulary and comprehension measures. Stevens et al. argued that these effect sizes, though nonsignificant, may be of practical importance for struggling readers, and that there is need for more research. With respect to the predictors included in the present study, Orton-Gillingham programs are more focused on skills related to phonological awareness (such as word and pseudoword reading) than on skills related to naming speed or phonological memory. To our knowledge, predictors of RTI for the Orton-Gillingham program have yet to be identified.

The present study

The goal of the present study was to identify predictors of RTI in an intensive and extensive adaptation of the Orton- Gillingham approach. Based on the research reviewed earlier, phonological awareness, naming speed, and phonological memory were measured prior to the intervention. Outcome measures addressed word and pseudoword reading accuracy and speed, reading comprehension, and spelling. This study was carried out with elementary-school-aged children who attended The Reading Clinic in Kingston, Ontario, Canada. This center provides treatment for children with reading difficulties and uses the principles of the Orton-Gillingham model with added fluency training (details of the instructional method are provided in the Method section). We had two research hypotheses. First, despite the lack of strong support for the Orton-Gillingham method [33] but based on our experience with the program studied, we hypothesized that gains would be shown in all reading and spelling measures. Second, based on previous research (see earlier review), we hypothesized that each of phonological awareness, naming speed, and phonological memory would predict RTI on each of the measured outcomes.

Method

Participants

Fifteen girls and 35 boys who attended The Reading Clinic in Kingston were the participants; when they began the program, they ranged in age from 5 years and 9 months to 12 years and 4 months (M = 8.25, SD = 1.3). This program took place in a privately-operated reading clinic in eastern Ontario. Parents learned about the program from psychologists, the clinic web site, flyers, and word of mouth. The students were almost entirely from white, middle-class families. Parents or guardians provided informed consent for their children’s participation, and the study was cleared by the General Research Ethics Board of Queen’s University.

Measures

All pre-intervention measures and approximately 70% of post-intervention measures were administered by co-author JJM, who was also one of four instructors in the intervention. Pre-intervention testing took place before the program began, and post-intervention testing occurred within five days following the intervention. The predictor variables comprised the pre-intervention scores for the phonological awareness, phonological memory, and rapid naming subtests of the Comprehensive Test Of Phonological Processing (CTOPP; [35]), age, and the number of sessions the individual child attended. Pre- and post-intervention literacy measures were the Gray Oral Reading Test (GORT-4; [36]), Test Of Word Reading Efficiency (TOWRE; [37]), and Diagnostic Screening Test for Spelling (DST-S; [38]). The number of participants varied from test to test, as shown in Table 1 & 2.

Predictor measures

CTOPP composite scores [35]: Three composite scores were calculated, following the procedures in the test manual. Each is a standard score based on two to four subtests, with a mean of 100 and a standard deviation of 15 in the norming sample. All reliability estimates are from the test manual.

The Phonological Awareness composite measures children’s ability to manipulate and segment speech sounds and is based on three subtests. The first, Phoneme Elision, asks children to repeat words after deleting indicated sounds (e.g., “Say fold without saying the /f/”). There are 20 items, but administration is terminated if children make errors on three consecutive items. In children ages 5-7, the test-retest reliability coefficient as .88, and for children 8-17 years of age, .79. In the 20 item Blending Words subtest, children are asked to indicate what word a series of sounds makes (e.g., t-oi makes toy). Test administration stops if children make three errors in a row. In children ages 5-7, the test-retest reliability coefficient is .88, whereas in children of 8-17 years of age, it is .72. The third subtest, Sound Matching, asks children to point to the image corresponding to the first or last phoneme of a word spoken to them (e.g., “Which word starts with the same sound as pan; pig, cone, or hat?”). The test ends if children miss four of seven items in a row. This test is designed for children 5-7 years of age, for whom the test-retest reliability is .83.

The Phonological Memory composite measures a child’s ability to store phonological information in short-term and working memory and has two subtests. In Memory for Digits, children are asked to repeat sequences of two to eight digits. The test ends if they make three successive errors. In children 5-7 years of age, the test-retest reliability coefficient .74 and between the ages of 8-17, .83. The Nonword Repetition subtest asks children to repeat nonwords that increase in length (e.g., viver, viversoo, viversoomouj). Testing ends if the child misses three consecutive items. For children between the ages of 5 and 7, the test-retest reliability is .68, and .75 for those 8-17 years old.

The Rapid Naming Composite has four subtests in which children are asked to read arrays of digits, letters, colors, or objects as quickly as possible (only children under the age of 8 complete the color and object naming tests, as they may be less familiar with alphanumeric stimuli). The test ends if children make four or more errors on Form A. The children’s final score is based on their time to read Forms A and B. The test-retest reliability coefficients range from .77 to .91 for 5-7-year-olds and .72 to .93 for 8-17-year-olds.

Literacy measures

Eight literacy measures were administered pre- and postintervention. Of these, six were age equivalent scores in months (all subtests of the GORT-4 and the TOWRE) and two were grade equivalent scores (the DST-S subtests). All reliability estimates are from the test manuals.

Gray Oral Reading Test –Version 4 (GORT-4; [36]) is a timed paragraph reading test with two forms containing 14 stories of increasing difficulty, each followed by five multiple choice comprehension questions (Form A was given at pre-test, Form B at post-test). Children are asked to read the stories as quickly and carefully as they can. The test ends when a child misses three of the five questions after any paragraph. Four scores are provided. (a) Reading rate is calculated from the time to read a story, and the test-retest reliability is .95 for Form A or Form B, .91 when switching from Form A to Form B. (b) Reading accuracy assesses the child’s ability to correctly pronounce words in the stories and is calculated through deviations from print. Test-retest reliability is .92, .88, and .91, for Form A, Form B, and switching from Form A to Form B. (c) Rate and Accuracy scores are combined to produce a Fluency score which has test-retest reliabilities of .93, .94, and .91, respectively. (d) The Comprehension score measures the appropriateness of children’s responses to the comprehension questions and has test-retest reliabilities of .86, .85, and .78, respectively.

The Test of Word Reading Efficiency (TOWRE; [37]) has two subtests with two alternate forms. Form A was given at pre-test, Form B at post-test. In Sight Word Efficiency, children read as many words as possible from a list of 104 real words in 45 seconds. If a child hesitates for three seconds on a word, it is counted as an error. Form A test-retest reliability is .97 in 6-9-year-old children and .84 for 10-18-year-old children, whereas the Form B test-retest coefficients are .96 and .92 for children 6-9 and 10-18 respectively. In Phonemic Decoding Efficiency, children are asked to read as many nonwords as possible from a list of 63 nonwords in 45 seconds. If a child hesitates for three seconds on a word, it is counted as an error. Test-retest reliabilities for Form A are .90 and .89 for 6-9 and 10-18-year-olds respectively. Form B test-retest coefficients are .90 and .83 for 6-9- and 10–18-year-olds respectively.

The Diagnostic Screening Test for Spelling (DST; [38]) asks children to spell 78 words of increasing difficulty. The Phonetic scale is for phonologically regular words and the Sight scale is for phonologically irregular words. Form A was given at pre-test, Form B at post-test. Cronbach alpha reliabilities range from .88 to .97 for grades 1-7.

Intervention

General design

All participants took part in the YES! Reading Program at The Reading Clinic in Kingston, Ontario, Canada, which was developed based on the clinic personnel’s experience with a variety of multi-sensory structured language programs, training in the Orton-Gillingham approach, and adherence to best practices from reading research. For example, the program includes fluency exercises that are not part of the traditional Orton-Gillingham approach.

Students were seen daily, five days a week for 55 minutes of 1:1 intervention. Some students who had completed more than 100 sessions were allowed to attend less frequently, two or three times per week. The number of sessions varied from 34 to 429 across students (M = 89.85, SD = 72.02). The intervention was delivered by 4 instructors, one of whom administered all the pre-intervention tests and approximately 70% of the postintervention tests. Instructors were trained at the clinic with 30 hours of instruction and 50 hours of supervised work with students.

Session activities

Each session started with 15-20 minutes of review, followed by 25-30 minutes of new teaching, and ended with 5 minutes of parent consultation. If a student had not reached mastery on prior teaching no new concepts or skills were introduced, and the time was used for further review. The review started with checking the homework from the previous day. Next came a phonogram flash card review requiring the student to give the appropriate phoneme(s), followed by dictation of previously learned phonemes for which the student had to write the corresponding phonogram(s). Depending on the student’s place in the program, the phonogram review also included morphemes (prefixes, suffixes, Latin and Greek bases). Language-related concepts that had been taught (e.g., vowels, open syllables) were reviewed. Then the child read 10-30 review words containing previously taught phonemes and morphemes, spelled 2-5 words, and read and spelled 3-5 irregular words. Unless skills were extremely weak, students participated in 5-7 minutes of fluency exercises, consisting of repeated timed readings of words, sentences and/or paragraphs. The fluency exercises were not introduced until (a) the student was able to decode two-sound words like “in”, “off”, or “up”, (b) short vowel sounds and single consonants had been taught, and (c) the student was able to read these words by blending.

The new teaching started with a new phonogram or morpheme (the “code”), unless a spelling rule, syllable type, or syllabification rule was being taught that day. The student then practiced reading and spelling the code, in isolation, using a variety of multisensory memory strategies. These included use of a keyword (“apple” for short /a/ sound), movement (hand makes chopping movement for /ch/ sound), making up a story (“look how tall the “t” is and the arms it has! It looks just like a telephone pole and has the same sound as the beginning of “telephone” /t/”), and printing it repeatedly while saying the sound. The student then read 5-20 words written in codecontrolled text, in which the new code was presented with very high frequency words, along with previously taught code. Only words containing the new or previously taught code appeared, ensuring that the student had the tools to decode each word and eliminating the value of guessing. Before starting to read the text, the student was instructed to scan through the words, look for those with the new code, circle the new code, and say the sound it makes. The student then read the text with the guidance of the instructor. The instructor then dictated 2-10 words with this new code for the student to spell. Next the student read 2-10 codecontrolled sentences, and a code-controlled paragraph and book if appropriate for their abilities. Spelling of one or two sentences was done if the student showed proficiency and stamina to write 7-10 phonetically regular words and 1-3 irregular words in a list. The session usually ended with a game requiring word or sentence reading using code-controlled text.

Parent consultation, homework, and exiting the program

During the parent consultation, the instructor showed the parent what skill was learned or reviewed that day and the homework activities they were to do that night to reinforce the new/review learning. Homework usually consisted of reviewing the phonogram pack, playing a game using code-controlled words or sentences, reviewing irregular words, and reading a book. Books were initially strictly code-controlled, with conventional literature being added later. The decision to add conventional literature was made by the individual instructor, but the general guideline was to have taught ¾ of the scope and sequence and for the student to be able to read 20 irregular words. Children’s participation in the program ended when their performance reached the average range (i.e., the 25th percentile or better on the Accuracy subtest of the GORT), when they reached the end of the curriculum for their age, when classroom teachers indicated they could keep up with regular classroom instruction, or when parents could no longer afford the time or cost of the program.

Design and analysis

This study employed a single-group longitudinal design. Predictor and pre-intervention literacy measures were administered before the intervention occurred. The intervention varied in length and was followed by administration of the post-intervention literacy measures. Paired-samples t-tests with corrections [39] were used to assess the first hypothesis, concerning change in reading and spelling achievement. The effect size of each change was measured by the difference between pre- and post-intervention means, divided by the standard deviation of the pre-intervention score. The main analyses, addressing the second hypothesis, were hierarchical regression analyses to determine how well the phonological processing measures predicted each literacy outcome after controlling the equivalent pre-test literacy variable, age at the beginning of the program, and number of sessions. Pre-intervention scores were entered in the first step, age and number of sessions in the second step, and one of the phonological processing variables (phonological awareness, phonological memory, or rapid naming) in the third step. Controlling number of sessions in the second step took account of both how much intervention the child received and how much time had elapsed between pre- and post-tests. By controlling the pre-intervention literacy variables in the first steps, the remaining variables predict the part of the post-intervention scores that cannot be predicted by the preintervention scores; thus they are predicting the changes in the literacy scores, i.e., RTI.

Results

Means and standard deviations are shown in Table 1 for predictor variables and Table 2 for the pre- and post-intervention literacy variables.

Changes in literacy variables

A series of paired samples t-tests was conducted to assess differences between pre- and post-intervention performance on the eight outcome variables (see Table 2), addressing the first hypothesis. All subtest scores increased significantly after intervention (ts > 6.45, ps < .001, suggesting that there had been considerable improvements between the pre- and post-tests. Each of these differences remained significant after applying the Benjamini-Hochberg procedure, controlling for multiple tests, with the false discovery rate of .05. The magnitude of the changes was substantial; in terms of pre-intervention SDs, improvements ranged from over 1.2 SD for Sight Word Efficiency to 1.9 SD for Phonetic Decoding Efficiency.

Predicting RTI

Before conducting the planned analyses, skewness and kurtosis were inspected for each predictor and literacy measure. To correct for distribution violations, transformations according to the guidelines of Tabachnick and Fidell were conducted. The following transformations were performed: the post-test DST-S Phonics score received a square root transformation; the posttest DST-S Sight scores received log (base 10) transformation; and the total number of sessions, pre-test scores for GORT-4 Rate, Accuracy, and Fluency received inverse transformations. In each case the transformations either resolved or improved the distribution problem. All subsequent analyses were performed with these transformed scores.

Regression results (addressing the second hypothesis) are shown in Table 3 for GORT-4 outcomes and Table 4 for TOWRE and DST outcomes. In step 1, pre-intervention scores on the literacy measures significantly predicted post-intervention scores on the same tests, βs ranging from .34 (p < .05) to .52 (p < .001), except for Phonetic Decoding Efficiency, β = .10 (p > .05), which addressed skills that were most directly targeted by the intervention program. In step 2, age was positively associated with improvement on most of the outcomes, but only significantly for GORT Comprehension, β = .36, p < .01. Number of intervention sessions was positively related to improvement on most outcomes, but only significantly for GORT Comprehension, Sight Word Efficiency, and Phonemic Decoding, βs = .33, .35, and .42 respectively, all ps < .05.

The critical findings come from the alternative third steps in the models (Table 3 & 4). Phonological Memory was a significant predictor of growth in seven of the eight outcomes (and marginal for the eighth, p = .067), with βs ranging from .31 to .48 (all ps < .05). Rapid Naming was a significant predictor of growth in three timed measures, Rate, Fluency, and Sight Word Efficiency, βs = .35, .32, and .39 respectively, all ps < .05. Phonological Awareness did not predict growth in any outcome. It should be noted that these effects are after taking account of the pre-intervention scores, age, and number of intervention sessions.

Discussion

primary purpose of this study was to investigate the predictors of RTI in this intervention program. We first address the amount of improvement shown in the program (first hypothesis) and interpret how these gains may have been related to the nature of the program, all within the context of the limitations of the research design. We next discuss our results regarding the predictors of RTI (second hypothesis). Finally, we discuss implications for the design of interventions, limitations of the present study, and future directions.

Gains in literacy achievement

Comparisons of pre- and post-intervention scores indicated that there were significant gains in each literacy variable (see Table 2). The magnitude of these gains was considerable, ranging from 1.2 to 1.9 standard deviations in the pre-intervention scores. It is important to interpret these gains with some caution. Lacking a control group and random assignment to treatment, other factors could be responsible for the children’s success, such as extra teaching in the children’s regular classes or extra attention to literacy at home. Similarly, sources of bias cannot be discounted, because both pre- and post-intervention tests were administered by personnel at the same center where the program was delivered, and it was obvious when measures were pre- or post-tests. Because children spent considerable time in the program, a certain amount of improvement could be expected in any case. However, the gains were large and practically important. For our present purposes, it is primarily important that changes were observed; the RTI predictor analyses addressed what was related to those changes.

Reviews of reading intervention programs for young children have reported effect sizes around .4, comparing children in the intervention to controls [26,40], and Stevens et al. [33] reported a nonsignificant effect of .22 for Orton-Gillingham programs. The changes observed in the present study are substantial but measured in a different way because of the absence of a control group. These gains may have been due to the extensive and intensive nature of the program, to its focus on phonological awareness and phonics which are consistent with current best practices [2,41], to the multi-sensory approach (which has been challenged by Stevens et al. and others), and/or to the addition of fluency and morphology which fit well with current research [42,43]. Unlike many programs, this one lasted for individual children until they had usually made substantial progress or attained grade-level performance (i.e., 25th percentile or better). We emphasize that our research design, without a control group or alternative treatments, does not allow us to attribute the gains to any of these features or to the program itself.

Predictors of RTI

We hypothesized that phonological awareness, phonological memory, and rapid naming would predict improvements or RTI. Previous studies were clear that phonological awareness and rapid naming are associated with reading achievement and RTI [22,27], but the evidence was more ambiguous for phonological memory. Our results (Table 3 & 4) show that phonological memory predicted improvement significantly on almost every measure, ranging from pseudoword reading and spelling to reading comprehension, including both timed and untimed measures. Rapid naming was significantly related to improvements on three timed measures, GORT Rate, GORT Fluency, and TOWRE Sight Word Efficiency. In contrast, phonological awareness was not significantly associated with gains in any outcome; no effect was close to the p = .05 level, and changes in R2 were .04 or less.

One possible explanation for the lack of effect of phonological awareness is that the intervention program targeted phonological awareness and its application in phonics in a very explicit and extensive manner. Hatcher and Hulme [29] demonstrated that phonological awareness was only a predictor of RTI when phonological analysis skills were taught implicitly rather than explicitly. The explicit instruction that children received in the current study may have been sufficient to improve their phonological awareness so that it was no longer a limiting factor, and the explicit and extensive attention to phonetic decoding may have mitigated the adverse effects of lower phonological awareness. Because phonological awareness was not assessed after the intervention program, it was not possible to determine to what extent phonological awareness had improved.

Although the intervention program addressed fluency, and participants improved substantially in all fluency-related measures, rapid naming predicted improvement in GORT Rate, GORT Fluency, and TOWRE Sight Word Efficiency. This suggests that the underlying difficulty in rapid naming had not been eliminated or its effects mitigated fully. As Kirby et al. [3] suggested, the underlying weakness in rapid naming may be very difficult to remediate, but its consequences for word reading may be lessened, for instance through extensive practice. As demonstrated in the current results, some children required extraordinary numbers of sessions, as many as 400, to make substantial progress (see Table 2 and Intervention section). This extensive practice is somewhat analogous to the practice of re-reading which has been shown to be successful. The present results support the contention of Amtmann et al. [5] that slow rapid naming is a “marker” of children who respond less favorably to intervention and therefore need more extensive intervention.

Phonological memory was strongly related to RTI for all reading and spelling outcomes (though marginally for GORT Comprehension), consistent with the findings of Frijters et al. [27] and Nelson et al. [28]. A plausible explanation of this lies in the nature of the intervention, which taught rules linking print to sequences of sounds. Individual sounds, both phonemes and larger units, need to be retained in sequence before being integrated into whole words, and for many words sequences of alternative sounds need to be held and evaluated, due to the ambiguous relationships between graphemes and phonology in English. This limitation is most obvious in word and pseudoword reading and spelling, but extends to comprehension, in which sequences of words must be retained long enough to compute sentence meaning [2]. Torgesen et al. [30] found that phonological memory explained improvement in pseudoword reading more than in real word reading; the present intervention may have been extensive enough to allow those pseudoword skills to transfer to real word reading and then to comprehension.

Implications for practice, limitations, and future directions

The current study demonstrates that children’s phonological memory and rapid naming predict gains in reading achievement in an intensive intervention of the Orton-Gillingham type. This raises questions about how to adapt such programs to address the problems of children with slow naming speed and low phonological memory. One approach would be to try to improve the two underlying abilities directly and encourage their transfer to literacy tasks. This may be difficult. For instance, there is little evidence that naming speed can be improved through intervention [3]. Memory training could be a way to address phonological memory problems. However, whereas some research found benefits for reading [44,45], others found little transfer from working memory training [46].

It may be more effective to lessen the consequences of slow naming speed and low phonological memory by teaching children, through extensive practice, to work with larger units of print, with orthographic chunks rather than single graphemes. Kirby et al. [3] made this argument about naming speed, but this technique should also counter some of the effects of low phonological memory, by decreasing the number of units to be held simultaneously. Working with larger units of text is almost certainly what typically-developing readers do to increase their reading speed as they automate word recognition. Larger orthographic units include morphemes, which have stable spellings [47], and other predictable letter sequences (e.g., -at, -ight). This may require considerable practice in the absence of efficient decoding processes, but may be effective [48-50]. Future RTI studies should include measures of orthographic and morphological awareness.

The current study should be interpreted with prudence due to several limitations. The lack of a control group means that we cannot attribute the literacy gains to the program or any feature of it with confidence. Further research is needed to identify whether particular components were more important than others, and whether gains continue to be apparent compared to appropriate controls. However, these results suggest that the program was effective and that it merits investigation in further studies. A larger sample size would have allowed greater power and the use of additional control variables, such as verbal and nonverbal ability. Unlike many previous studies, the program studied here was privately operated, thus many of the children were likely of higher socioeconomic status rather than representing a typical spectrum of family backgrounds; further studies with more diverse samples are warranted. Predictor variable effects may differ depending on age, and due to modest statistical power, interaction terms could not be reasonably tested. Age-matched wait-list controls should be used to compare the progress of children in treatment to children of the same reading age who are typically progressing, and children with dyslexia who are progressing more slowly. Finally, post-test measures of phonological awareness, naming speed, and phonological memory would clarify whether reading outcome improvements are entirely based on improvements in these processes.

In conclusion, this study points to the need for further studies of the predictors of RTI in reading intervention programs, and for the application of such studies’ results to improve those programs. The results demonstrate the importance of phonological memory and rapid naming, suggesting that these processes be considered when designing programs and assessments. Increased practice time, increased intensity and duration, and more focus on larger orthographic units could lead to improvements in tailoring interventions to students’ needs.

References

1. Fletcher JM, Lyon RG, Fuchs LS, Barnes MA (2019) Learning disabilities: From identification to intervention (2^nd edn). New York, NY: Guilford Press.
2. Castles A, Rastle K, Nation K (2018) Ending the reading wars: Reading acquisition from novice to expert. Psychological Science in the Public Interest 19(1): 5-51.
3. Kirby JR, Georgiou GK, Martinussen R, Parrila R (2010) Naming speed and reading: From prediction to instruction. Reading Research Quarterly 45(3): 341-362.
4. Melby-Lervåg M, Halaas Lyster SA, Hulme C (2012) Phonological skills and their role in learning to read: A meta-analytic review. Psychological Bulletin 138(2): 322-352
5. Amtmann D, Abbott RD, Berninger VW (2008) Identifying and predicting classes of response to explicit phonological spelling instruction during independent composing. Journal of Learning Disabilities 41(3): 218-234.
6. Tran L, Sanchez T, Arellano B, Swanson HL (2011) A meta-analysis of the RTI literature for children at risk for reading disabilities. Journal of Learning Disabilities 44(3): 285-295.
7. Parrila RK, Protopapas A (2017) Dyslexia and word reading problems. In: K Cain, DL Compton, RK Parrila (Eds.), Theories of reading development. Amsterdam, NL: John Benjamins Publishing Company, pp. 335-358.
8. Hulme C, Bowyer-Crane C, Carroll JM, Duff FJ, Snowling MJ (2012) The causal role of phoneme awareness and letter-sound knowledge in learning to read: Combining intervention studies with mediation analyses. Psychological Science 23(6): 572-577.
9. Al Otaiba S, Fuchs D (2002) Characteristics of children who are unresponsive to early literacy intervention: A review of the literature. Remedial and Special Education 23(5): 300-316.
10. Kudo MF, Lussier CM, Swanson HL (2015) Reading disabilities in children: A selective meta-analysis of the cognitive literature. Research in Developmental Disabilities 40(1): 51-62.
11. Kirby JR, Parrila R, Pfeiffer SL (2003) Naming speed and phonological awareness as predictors of reading development. Journal of Educational Psychology 95(3): 452-464.
12. Ehri L, Nunes S, Willows D, Schuster B, Yghaoub-Zadeh Z, et al. (2001) Phonemic awareness instruction helps children learn to read: Evidence from the National Reading Panel meta-analysis. Reading Research Quarterly 36(3): 250-287.
13. Lovett MW, Lacerenza L, Borden SL, Frijters JC, Steinbach KA, et al. (2000a) Components of effective remediation for developmental reading disabilities: Combining phonological and strategy-based instruction to improve outcomes. Journal of Educational Psychology 92(2): 263-283.
14. Lovett MW, Steinbach KA, Frijters JC (2000b) Remediating the core deficits of developmental reading disability: A double-deficit perspective. Journal of Learning Disabilities 33(4): 334-358.
15. Wolf M, Bowers PG (1999) The double-deficit hypothesis for the developmental dyslexias. Journal of Educational Psychology 91(3): 415-438.
16. Wolff (2014) RAN as a predictor of reading skills, and vice versa: Results from a randomized reading intervention. Annals of Dyslexia 64(2): 151-165.
17. Swanson HL, Zheng X, Jerman O (2009) Working memory, short-term memory, and reading disabilities: A selective meta-analysis of the literature. Journal of Learning Disabilities 42(3): 260-287.
18. Menghini D, Finzi A, Carlesimo GA, Vicari S (2011) Working memory impairment in children with developmental dyslexia: Is it just a phonological deficit? Developmental Neuropsychology 36(2): 199-213.
19. Clark BN, McRoberts GW, Van Dyke AJ, Shankweiler PD, Braze D (2012) Immediate memory for pseudowords and phonological awareness are associated in adults and pre-reading children. Clinical Linguistics & Phonetics 26(7): 577-596.
20. Boets B, De Smedt B, Cleuren L, Vandewalle E, Wouters J, et al. (2010) Towards a further characterization of phonological and literacy problems in Dutch-speaking children with dyslexia. British Journal of Developmental Psychology 28(1): 5-31.
21. Nithart C, Demont E, Metz‐Lutz M, Majerus S, Poncelet M, et al. (2011) Early contribution of phonological awareness and later influence of phonological memory throughout reading acquisition. Journal of Research in Reading 34(3): 346-363.
22. Johnson DED, Swanson HL (2011) Cognitive characteristics of treatment-resistant children with reading disabilities: A retroactive study. Journal of Psychoeducational Assessment 29(2): 137-149.
23. Scheltinga F, van der Leij A, Struiksma C (2010) Predictors of Response to Intervention of word reading fluency in Dutch. Journal of Learning Disabilities 43(3): 212-228.
24. Kastamoniti A, Tsattalios K, Christodoulides P, Zakopoulou V (2018) The role of phonological memory in reading acquisition and dyslexia: A systematic literature review. European Journal of Special Education Research 3(4): 278-323.
25. Vaughn S, Capin P, Scammacca N, Roberts G, Cirino P, Fletcher JM (2020) The critical role of word reading as a predictor of response to intervention. Journal of Learning Disabilities 53(6): 415-427.
26. Wanzek J, Stevens EA, Williams KJ, Scammacca N, Vaughn S, et al. (2018) Current evidence on the effects of intensive early reading interventions. Journal of Learning Disabilities 51(6): 612-624.
27. Frijters JC, Lovett MW, Steinbach KA, Wolf M, Sevcik RA, et al. (2011) Neurocognitive predictors of reading outcomes for children with reading disabilities. Journal of Reading Disabilities 44(2): 150-166.
28. Nelson JR, Benner GJ, Gonzalez J (2003) Learner characteristics that influence the treatment effectiveness of early literacy interventions: A meta-analytic review. Learning Disabilities Research & Practice 18(4): 255-267.
29. Hatcher PJ, Hulme C (1999) Phonemes, rhymes, and intelligence as predictors to children’s responsiveness to remedial reading instruction: Evidence from a longitudinal intervention study. Journal of Experimental Child Psychology 72(2): 130-153.
30. Torgesen JK, Wagner RK, Rashotte CA, Rose E, Lindamood P, et al. (1999) Preventing reading failure in young children with phonological processing disabilities: Group and individual responses to instruction. Journal of Educational Psychology 91(4): 579-593.
31. Rose TE, Zirkel P (2007) Orton-Gillingham methodology for students with reading disabilities. Journal of Special Education 41(3): 171-185.
32. Sayeski KL, Earle GA, Davis R, Calamari J (2019) Orton Gillingham: Who, what, and how. Teaching Exceptional Children 51(3): 240-249.
33. Stevens EA, Austin C, Moore C, Scammacca N, Boucher AN, et al. (2021) Current state of the evidence: Examining the effects of Orton-Gillingham reading interventions for students with or at risk for word-level reading disabilities. Exceptional Children 87(4): 397-417.
34. Ritchey KD, Goeke JL (2006) Orton-Gillingham and Orton-Gillingham–based reading instruction: A review of the literature. Journal of Special Education 40(3): 171-183.
35. Wagner RK, Torgesen JK, Rashotte CA (1999) Comprehensive Test of Phonological Processing. Austin, TX: PRO-ED, Inc.
36. Wiederholt JL, Bryant BR (2001) Gray Oral Reading Tests – Fourth Edition. Austin, TX: PRO-ED, Inc.
37. Wagner RK, Torgesen JK, Rashotte CA (1999) Test of Word Reading Efficiency. Austin, TX: PRO-ED, Inc.
38. Gnagey TD (1982) Diagnostic Screening Test: Spelling – Third Edition. Slosson Educational Publications, Inc.
39. Benjamini Y, Hochberg Y (2000) On the adaptive control of the false discovery rate in multiple testing with independent statistics. Journal of Educational and Behavioral Statistics 25(1): 60-83.
40. Gersten R, Haymond K, Newman-Gonchar R, Dimino J, Jayanthi M (2020) Meta-analysis of the impact of reading interventions for students in the primary grades. Journal of Research on Educational Effectiveness 13(2): 401-427.
41. McArthur G, Sheehan Y, Badcock NA, Francis DA, Wang HC, et al. (2018) Phonics training for English-speaking poor readers. Cochrane Database of Systematic Reviews 2018(11).
42. Bowers PN, Kirby JR, Deaco SH (2010) The effects of morphological instruction on literacy skills: A systematic review of the literature. Review of Educational Research 80(2): 144-179.
43. Goodwin AP, Ahn S (2010) A meta-analysis of morphological interventions: Effects on literacy achievement of children with literacy difficulties. Annals of Dyslexia 60(2): 183-208.
44. Dahlin KIE (2011) Effects of working memory training on reading in children with special needs. Reading and Writing: An Interdisciplinary Journal 24(4): 479-491.
45. Loosli SV, Buschkuehl M, Perrig WJ, Jaeggi SM (2012) Working memory training improves reading processes in typically developing children. Child Neuropsychology 18(1): 62-78.
46. Melby-Lervåg M, Hulme C (2013) Is working memory training effective? A meta-analytic review. Developmental Psychology 49(2): 270-291.
47. Kirby JR, Bowers PN (2017) Morphological instruction and literacy: Binding phonological, orthographic, and semantic features of words. In: K Cain, DL Compton, RK Parrila (Eds.), Theories of reading development. Amsterdam, NL: John Benjamins Publishing Company, pp. 437-462.
48. Elbro C, Arnbak E (1996) The role of morpheme recognition and morphological awareness in dyslexia. Annals of Dyslexia, 46(1): 209-240.
49. Deacon SH, Parrila R, Kirby JR (2008) A review of the evidence on morphological processing in dyslexics and poor readers: A strength or weakness? In: G Reid A, Fawcett F Manis, L Siegel (Eds.), The Sage Handbook of Dyslexia. UK: Sage Publications, pp. 212-237.
50. Wolf M, Barzillai M, Gottwald S, Miller L, Spencer K, et al. (2009) The RAVE-O intervention: Connecting neuroscience to the classroom. Mind, Brain, and Education 3(2): 84-93.