Do Cognitive and Non-Cognitive Factors Predict Responses to Reading Fluency Interventions?

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Do Cognitive and Non-Cognitive Factors Predict Responses to Reading Fluency Interventions? © 2024 the Authors Published version Aro, Tuija; Koponen, Tuire; Peura, Pilvi; Räikkönen, Eija; Viholainen, Helena; Aro, Mikko Aro, T., Koponen, T., Peura, P., Räikkönen, E., Viholainen, H., & Aro, M. (2024). Do Cognitive and Non-Cognitive Factors Predict Responses to Reading Fluency Interventions?. Journal of Experimental Education, Early online. https://doi.org/10.1080/00220973.2024.2358501 2024 Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=vjxe20 The Journal of Experimental Education ISSN: (Print) (Online) Journal homepage: www.tandfonline.com/journals/vjxe20 Do Cognitive and Non-Cognitive Factors Predict Responses to Reading Fluency Interventions? Tuija Aro, Tuire Koponen, Pilvi Peura, Eija Räikkönen, Helena Viholainen & Mikko Aro To cite this article: Tuija Aro, Tuire Koponen, Pilvi Peura, Eija Räikkönen, Helena Viholainen & Mikko Aro (29 May 2024): Do Cognitive and Non-Cognitive Factors Predict Responses to Reading Fluency Interventions?, The Journal of Experimental Education, DOI: 10.1080/00220973.2024.2358501 To link to this article: https://doi.org/10.1080/00220973.2024.2358501 © 2024 The Author(s). Published with license by Taylor & Francis Group, LLC Published online: 29 May 2024. Submit your article to this journal View related articles View Crossmark data Do Cognitive and Non-Cognitive Factors Predict Responses to Reading Fluency Interventions? Tuija Aro a,b,c , Tuire Koponen b,c , Pilvi Peura c , Eija R€ aikk€ onen a , Helena Viholainen a , and Mikko Aro a,c a University of Jyv€ askyl€ a, Finland; b Niilo M€ aki Institute, Finland; c Centre of Excellence in Learning Dynamics and Intervention Research (InterLearn), University of Jyv€ askyl€ a, Finland ABSTRACT We investigated whether emotional and motivational factors had predictive effects beyond those of cognitive factors on responses to two reading fluency interventions. Eighty-two dysfluent readers (Grades 3–5) participated in a 12-week school-based fluency intervention, either combined with or without self-efficacy support. Response to the intervention was determined by the fluency gain score and the Reliable Change Index. In the skillfocused intervention, cognitive predictors contributed to the response, and reading-related anxiety and self-efficacy had effects beyond the cognitive predictors. Weaker initial reading skills and older age predicted response in the combined intervention. Thus, children’s personal characteristics may have a greater influence on their responses in a skills-focused intervention than in an intervention that also considers emotional and motivational aspects. KEYWORDS Cognitive predictors; noncognitive predictors; reading fluency intervention; response to intervention In orthographically transparent languages, such as Finnish, Italian, or Spanish, the majority of children develop an accurate decoding skill during their first grade in school (Seymour et al., 2003), and reading fluency becomes a primary challenge around the second grade for those with delayed development. Accordingly, reading disability is mainly manifested as a problem in acquiring an efficient and fluent decoding skill (Aro & Wimmer, 2003). As reading fluency is subsequently needed for reading comprehension (Pikulski & Chard, 2005), effective interventions for children struggling to become fluent readers are of utmost importance. Although reading fluency interventions have been shown as moderately effective at the group level (e.g., Maki & Hammerschmidt-Snidarich, 2022), there is considerable variability in participants’ responses to interventions (e.g., Al Otaiba & Fuchs, 2006). Relatively little effort has been invested in understanding this variability, although a better understanding would help in developing effective interventions for a variety of students. Furthermore, existing studies on intervention responses have focused on cognitive predictors. For example, Fuchs et al. (2021) found that participants with weaker pretreatment phonological awareness showed stronger intervention effects. However, cognitive predictors have been found insufficient in explaining individual variations in responses to fluency interventions (e.g., Stuebing et al., 2015) and there has been an emerging recognition of the importance of the “non-cognitive” factors of school learning, such as motivation, emotions, and beliefs (Farrington et al., 2012). Still, there is a scarcity of intervention CONTACT Tuija Aro [email protected] Department of Psychology, University of Jyv€ askyl€ a, Finland � 2024 The Author(s). Published with license by Taylor & Francis Group, LLC This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/ 4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The terms on which this article has been published allow the posting of the Accepted Manuscript in a repository by the author(s) or with their consent. THE JOURNAL OF EXPERIMENTAL EDUCATION https://doi.org/10.1080/00220973.2024.2358501 research targeting these factors as predictors of response to intervention. Thus, in this study, we studied whether emotional and motivational factors had effects—beyond the effects of cognitive factors—on children’s responses to two reading fluency interventions. In a broad sense, reading fluency can be defined as the ability to read accurately and rapidly, with appropriate expression (Kuhn et al., 2010). In this study, we operationalized reading fluency as the ability to read accurately and with speed. The present study is based on our earlier study reporting reading fluency and reading fluency related self-efficacy (SE) outcomes of elementary school children who participated in two partly different reading fluency interventions (Aro et al., 2018). The first intervention was a traditional reading fluency skill-training intervention including repeated reading in a group and individual computer-based training. The second intervention used a combined approach, embedding reading SE support (e.g., visualization and verbalization of progress, mastery experiences, verbal persuasions, and discussing emotions) in a similar reading fluency skill-training as the other group received. In this previous study, we examined the mean-level effects of the interventions using three reading fluency measures and one reading SE measure as indicators of the outcome (i.e., gain score indicating difference between pre- and post-intervention assessments). The results revealed that the two intervention groups had similar improvements in reading fluency in mean level which concords with findings of McBreen and Savage (2022) who found no significant differences in reading fluency between groups receiving either cognitive-only or cognitive plus motivational reading intervention. Furthermore, there were large variances of the gain scores and both groups had children demonstrating clinically reliable changes and those not showing such changes. These mean-level analyses did not provide understanding on factors contributing to intervention response, on the contrary, they indicated the need for closer inspection from the perspective of predictive factors. Thus, in the present study, we wanted to gain a more in-depth understanding of the participant-related factors assessed before the interventions influencing the variance in their responses in these two interventions. As previous research on intervention response has targeted mainly cognitive factors, we wanted to understand the extent to which both reading-related cognitive predictive factors and reading-related emotional and motivational predictive factors explained the participants’ responses; more specifically, we studied whether the emotional and motivational factors had effects beyond those of the cognitive ones. Based on the rationales explained in the following sections, we investigated the effects of phonological skills and rapid automatized naming (RAN) as cognitive factors and the effects of reading-related anxiety and reading-related self-effi- cacy as emotional and motivational factors using two operationalizations of the intervention response. Cognitive predictors of response The pre-intervention reading skill level has often been included as one of the predictors of responses to fluency interventions. In general, the predictive effects have been found to be rather small, with varying directions of the effects (for a meta-analysis, see Scholin & Burns, 2012). In some studies, the students with lower skill levels have benefited the most, whereas in others, the students with better skills or milder difficulties have gained the greatest advantages (Scholin & Burns, 2012). The essential role of phonological processing skills in reading development has often been presented in English speaking languages (e.g., Snowling, 2001). Recent evidence from different orthographies (English, French, German, Dutch, and Greek) showed that especially RAN predicts reading fluency across orthographies (Landerl et al., 2019) corroborating earlier evidence and theorization (e.g., Wolf & Bowers, 1999; Kirby et al., 2010). Despite phonological skills and RAN are 2 T. ARO ET AL. suggested to be the core predictors of reading skill, their role in predicting response to intervention is less known. Studies on responses to reading fluency interventions have shown both significant and nonsignificant effects associated with phonological skills and RAN. For example, among Englishspeaking children, phonological awareness was associated with improvement in fluency after an intervention including fluency training (e.g., Barth et al., 2010; Fuchs et al., 2021; Fletcher et al., 2011), while a variety of phonological processing skills did not predict Dutch children’s responses to fluency intervention (Scheltinga et al., 2010). Similarly, other studies have shown that performance in RAN predicts responses to an intervention targeting reading fluency (Scheltinga et al., 2010) and to an intervention including fluency training as a component (e.g., Barth et al., 2010). However, not all intervention studies on reading fluency have found a significant predictive effect associated with RAN. Field et al. (2019) found that RAN did not have a significant predictive role in students’ response to fluency intervention, particularly among those with significant fluency deficits. As phonological skills and RAN have been suggested to predict reading skill development, and their role in explaining intervention outcome is not clear, in this study, we included them (in addition to initial reading skills) to our analyses to obtain more evidence of their predictive roles in fluency intervention outcomes. Emotional and motivational predictors of response It has been known that anxiety symptoms are common among students with learning disabilities (meta-analysis: Nelson & Harwood, 2011), and accordingly, higher anxiety has also been shown to be associated with reading difficulties (Francis et al., 2019; Grills et al., 2022; Grills-Taquechel et al., 2013). Although majority of the studies have focus on other forms of anxiety than readingrelated anxiety, some research has shown that anxiety more focally related to reading is also associated with reading achievement (Mohammadpur & Ghafournia, 2015) and that students with reading difficulties may experience anxiety specifically related to reading (Ramirez et al., 2019) or to academic situations (Elgendi et al., 2021). These findings may be due to anxiety’s relations with poorer working memory performance (Moran, 2016) which may hamper learning. Based on the interference model (for test anxiety see Tobias, 1985) anxious students may experience interference with their concentration, memory functioning, and/or information processing leading to deficient learning. Beside anxiety, also self-beliefs, such as self-concept and SE, have been found to be associated with reading motivation and better reading fluency development (Nevo et al., 2020), but the association between reading self-concept and reading fluency has also been shown to be reciprocal (Quirk et al., 2009). Recent studies have shown that lower reading fluency related SE is associated with poorer word reading (Carroll & Fox, 2016) and with poorer reading fluency development (Peura et al., 2019). In the present study, our focus is on reading-related SE which is an element of reading motivation and assumed to influence on child’s thoughts and feelings in the task situation and thereby on effort and persistence invested in the task, and finally on achievement (e.g., Schunk & Mullen, 2012). Thereby, it may also be a factor influencing response to intervention. Despite the shown associations between reading skill and reading-related anxiety and SE, their relevance for intervention response is less well understood. Recently, Ronimus et al. (2020) found that high SE was associated with better reading fluency development after playing the GraphoGame program, and Vaughn et al. (2022) showed a moderating effect of students’ reading anxiety on their reading fluency outcome after an intervention targeting multiple components of reading; higher anxiety was associated lower intervention effect. In contrast, Grills et al. (2014) reported that anxiety did not predict intervention response in an intervention study targeting several reading components. Due to the paucity of research, we lack knowledge on reading-related anxiety and reading-related SE as predictors of responses to fluency interventions. THE JOURNAL OF EXPERIMENTAL EDUCATION 3 Combined interventions The growing awareness of the relevance of other than cognitive factors for reading development has resulted in an increase of studies combining such components with reading interventions. In these combined interventions, mostly motivation/self-belief (Lovett et al., 2021; McBreen & Savage, 2022; Toste et al., 2017, 2019; Zentall & Lee, 2012) or anxiety regulation (Francis et al., 2021; Vaughn et al., 2022) has been embedded in reading instruction. These studies have found positive effects on the non-cognitive aspects trained, such as anxiety, reading competence, and success attributions (e.g., Francis et al., 2021; Lovett et al., 2021; Toste et al., 2017) as well as on reading skills, primarily on comprehension (McBreen & Savage, 2022), but also on fluency (Lovett et al., 2021; Toste et al., 2017). For example, Toste et al. (2017) reported that after the intervention, the participants of the intervention including support for motivation were more likely than those in the business-as-usual group to attribute success to internal causes (e.g., effort) than to external causes (e.g., luck). However, despite the above intervention studies having shown positive emotional and motivational mean-level effects of combined interventions, we lack knowledge on who benefit from combined intervention, what kind of individual pre-intervention characteristics explain response to combined intervention and to what extent the same characteristics predict response to solely skill-focused reading fluency intervention. Despite not being the main aim of the study, Vaughn et al. (2022) showed that initial reading anxiety moderated the reading fluency outcome of the intervention targeting both reading and anxiety but not that of the business-as-usual intervention (i.e., no researcher provided treatment). This finding suggests that different pre-intervention characteristics may influence the responses to combined interventions and to interventions comprising purely skill training. Defining response to intervention and using continuous and dichotomic predictors As it has been shown that various definitions of intervention response may result in different findings (e.g., Hughes & Dexter, 2011), both continuous and categorical operationalizations of response were used in the present study. First, by using a continuous reading fluency gain score (i.e., difference between pre- and post-intervention scores) as a measure of response to intervention, we searched for linear relations between the predictors and the responses across the whole distribution. This allowed the participants with the lowest performance levels to demonstrate growth, which could have been obfuscated by using a criterion and/or a norm-referenced cutoff score. Second, we designated the participants as either responders or non-responders (cf. responder status) based on the Reliable Change Index (RCI; Jacobson & Truax, 1991), to provide information on the present sample instead of comparing participants to peers or normative samples. This allowed us to explore individual participants’ likelihood of benefiting from the interventions and to map the characteristics that were prevalent among responders and non-responders (i.e., according to their responder status). As previous studies have suggested, the association between cognitive precursors (i.e., RAN) and academic skill are not necessarily linear (Koponen et al., 2013) and that small to moderate amounts of anxiety can have a motivating role, while excessive amounts of anxiety can result in decreased skill development (see Grills-Taquechel et al., 2012), we wanted to study whether high reading-related anxiety or low reading-related SE, or low performance in RAN or in phonological test affected intervention outcomes. Therefore, beside using continuous pre-intervention predictive variables, we also predicted responses to the interventions with dichotomic variables. To do this, we first divided the participants based on their pre-intervention scores into those showing and those not showing a clear problem in the specific measure. We then studied the association 4 T. ARO ET AL. between their responder status and the presence of clear problems and between their responder status and accumulation of these clear problems. Aim of the present study The previous research findings indicate the potential of providing emotional and/or motivation support together with fluency training, thus backing the stance that combined intervention approaches constitute a significant second step when aiming to tackle the complex concerns related to reading difficulties (Grills et al., 2022). However, more research is needed, especially on whether emotional or motivation factors have predictive effects beyond those of the identified cognitive predictors of intervention outcomes and whether the same factors predict the responses to skill-focused and combined interventions. Thus, the aim of the present study was to analyze to what extent the participants’ pre-intervention cognitive and emotion and motivation related characteristics explain variance in the responses to the two reading fluency interventions, one providing traditional skill-training (FLUENCY) and the other using a combined approach, embedding reading SE support in skill-training (FLUENCY þSE). Our research questions (RQs) were: 1. To what extent was the variance in the reading fluency gain score explained by pre-interven- tion phonological skills and RAN (a) FLUENCY-intervention and (b) FLUENCY þSE-inter- vention? Did the pre-intervention reading-related anxiety and reading-related SE have effects beyond those of the cognitive predictors in (a) FLUENCY-intervention and (b) FLUENCY þSE-intervention? 2. (a) Were cognitive or emotion and motivation related problems, that is, weak pre-interven- tion performance in the phonological test or in RAN or high scores in anxiety or low score in SE, associated with the responder status defined by RCI in the two intervention groups? (b) Was the accumulation of these problems associated with the responder status in the two intervention groups? Method Procedure and participants This study was part of the Self-Efficacy and Learning Disabilities Intervention (SELDI) study focusing on reading and math fluency interventions in elementary schools. In this article, we report the analyses concerning reading interventions carried out in the school context, applying a quasi-experi- mental design. The 12-week intervention programs started in January 2014. The individual preintervention assessments of the cognitive skills were conducted from November to December after the participant screening and selection. The pre-intervention reading fluency and the readingrelated anxiety and reading-related SE were assessed in January. The post-intervention reading fluency assessment was conducted in May. The Ethical Committee of the University of Jyv€ askyl€ a has approved the study. Participant recruitment The data was collected in four municipalities in central and eastern Finland, where all interested teachers teaching mainstream students in Grades 2–5 were invited to join the study. In total, 20 schools from rural, suburban, or urban areas participated. These schools had 27 special education teachers. They invited the classroom teachers to participate in the study. Seventy-five classroom teachers joined, and they asked permission from their students’ guardians to let the children participate in the study. Participation was voluntary, and informed consents were received from the guardians. Of the 20 participating schools, 14 were selected to provide either FLUENCY þSE- intervention or FLUENCY-intervention (7 schools each) for reading fluency in Grades 3–5; the THE JOURNAL OF EXPERIMENTAL EDUCATION 5 reaming six schools provided only math intervention. The two interventions were not provided in the same school to avoid “contamination.” There were 3–5 students in each intervention group. To select the participants, first, two time-limited group tests of reading fluency (Allu and Luksu tests; see below Measures) and an individually administered reading-aloud test (Text Reading Fluency, see below Measures) was administered. The latter was also used as a measure of the intervention response in this study. Second, all children who performed below the 20th percentile in the group-level tests were further assessed with an individually administered Lukilasse test (see below Measures) to verify the group-assessment results concerning reading problems. Thus, the final inclusion criterion for the intervention study was performance of <7 at a scalescore in the Lukilasse test. The FLUENCY- and the FLUENCY þSE-groups were matched according to the Lukilasse test results. Altogether, 1,098 children (446 from Grade 3, 360 from Grade 4, and 292 from Grade 5) participated in the group assessments. The whole dataset was used to standardize the questionnaire’s sum scores in this study. Of the total number of participants (1,098), 1,016 were not assigned the interventions (52% male), 42 were assigned to the FLUENCY-intervention (64.3% male), and 40 were assigned the FLUENCY þSE-intervention (70% male). The median grade level was 4 in all groups (max 9). No differences between the groups emerged in the Raven’s Colored Progressive Matrices test (Raven et al., 1990) administered in a group in the classroom. See Table 1 for the scores of the intervention groups; the scores of the reference group were M¼30.70 and standard deviation [SD]¼3.90. The demographic information about the intervention groups is also presented in Table 1. Measures Reading Measures for participant selection. Two group-assessments tests (Allu and Luksu) and one individually administered test (Lukilasse) were used in the participant selection process. In Allu, the Table 1. Descriptive information of the sample. Group n a Girls/Boys b Grade 3/4/5 c FLUENCY 40–42 15/27 15/14/13 FLUENCY þSE 38–40 12/28 14/10/16 M SD M SD RCPM (raw score) 29.05 4.63 29.64 4.75 Lukilasse (stand. score) 3.90 1.83 4.61 1.76 RAN (z-score) −.88 1.11 −1.53 1.45 Phonology (stand. score) 6.78 2.61 7.02 3.34 Anxiety (z-score) .79 1.17 .17 1.06 SE (z-score) −.78 1.17 −.54 1.03 % (n) % (n) Mother’s education b Comprehensive school 7.9 (3) 11.1. (3) High school/vocat. school 57.9 (22) 48.1 (12) College/polytechnic/bachelor 34.2 (13) 40.7. (11) Master’s degree 0 (0) 11.1 (3) Finnish as the main home-language b 92.3 (36) 88.9 (24) Diagnoses (e.g., asthma, migraine, DLD) b 2.6 (1) 11.1 (3) a Due to missing data (absence from school during the assessment) number of children varied. b 38–39 participants of the FLUENCY-group and 26–27 of the FLUENCY þSE-group had data on mother’s education, home language, and diagnoses. c Two children had missing data on gender. RCPM ¼Raven’s Colored Progressive Matrices (Raven et al., 1990). Lukilasse ¼Lukilasse Reading subtest (H€ ayrinen et al., 1999) conducted prior to intervention. DLD ¼Developmental Language Disorder. 6 T. ARO ET AL. children read silently and separate the words with a vertical line. According to the manual, the Allu test assesses word reading and has a high scale reliability (Cronbach’s alpha ¼0.97; Lindeman, 1998); it consists of words written in 2–4-word clusters, with no spaces between them (78 clusters in total). The Luksu test is a Finnish adaptation of the Woodcock-Johnson Reading Fluency Test (Woodcock et al., 2001) assessing sentence-level reading. It consists of 70 short and easy statements (e.g., “Strawberries are red.”). Both the scale reliability and the split-half reliability have been reported to be to be good in the normative sample (Cronbach’s alpha ¼0.94 and split-half ¼0.97; Eklund et al., 2013). In the present data (n¼1,075), the Allu and Luksu scores highly correlated, r¼0.78. The Lukilasse test was used for verifying the group-assessment performances and to match the intervention groups. In the Lukilasse the children read aloud a list of words of increasing complexity and length. According to the manual and the normative data, the test has shown good scale reliability for all grades (Cronbach’s alpha range ¼0.94–0.98; H€ ayrinen et al., 1999). In the present data (n¼81), test-retest (pre- and post-intervention assessment) correlation was r¼0.84 (the test was done only for the intervention participants). Measure of pre-intervention reading level and intervention gain. A reading-aloud text-reading test (Text Reading Fluency, Salmi et al., 2011) was administered individually in pre- and postintervention assessments. In the test, each child reads aloud a 120-word text. The score is the number of correctly read words within 1.5 min. In the present data (n¼81), the text-reading score’s correlation with the Lukilasse test was 0.87, and test-retest (pre- and post-intervention assessment) correlation was r¼0.86. The gain score used in this study was the discrepancy between the pre- and the post-intervention text-reading scores. Cognitive and emotion and motivation related predictors Rapid automatized naming (RAN). RAN was assessed with the alpha-numerical subtests of the Test of Rapid Serial Naming (Finnish version) (Ahonen et al., 1999). The mean of the times taken to name the letter and the digits was used as the score. For the Finnish version a good split-half reliability coefficient (0.80) has been reported (e.g., Torppa et al., 2017). As RAN is a time limited test, counting Cronbach’s alpha separately either for letters or for digits is not meaningful. However, their correlation in the present data (n¼81) was 0.84 indicating good reliability. Phonological skills. Two phonological processing tasks in the Phonological Awareness subtest of Nepsy-II (Korkman et al., 2007) were used. In the Word Segment Recognition task, the child is asked to identify words from word segments. The Phonological Segmentation is a test of elision. The child is asked to repeat a word and then to create a new word by omitting a syllable or a phoneme or by substituting one phoneme in a word with another. In the manual, the reliability coefficient has been reported to be very high (0.96; Korkman et al., 2008). As the items-level data was not available for the present sample, Cronbach’s alpha could not be counted. Reading-related anxiety. Reading-related anxiety was assessed with three statements on anxiety or tension arousal in situations involving reading: “I become anxious when I know I have to read aloud,” “I become anxious when I start reading,” and “I feel tension in my body when I have to read.”. The items were adopted from a subscale initially designed for measuring affective and physical state sources of self-efficacy, which was based closely on the ideas of Usher and Pajares (2009; see Bandura, 1997). The factor structure of the reading anxiety measure has been shown to be satisfactory (high factor loadings) and the structure to be invariant across grade levels (see Peura et al., 2021). In this study we used a sum score of the three reading anxiety items. The instructions and the items were read aloud one by one to ensure that all children could answer THE JOURNAL OF EXPERIMENTAL EDUCATION 7 Factors associated with the gains score Our first main finding was that the pre-intervention predictors jointly explained a greater variance in the text-reading gain score (45%) of the FLUENCY-group than that of the FLUENCY þSE-group (20%). Although the contributions were only marginally significant, both the phonology and RAN as well as the reading-related anxiety and SE contributed considerably more to the gain score (both around 12%) in the FLUENCY-group compared to the FLUENCY þSE-group (both around 1%). This suggests that the cognitive predictors only contribute to the response in the FLUENCY-group; similarly, the reading-related anxiety and SE have effects beyond the variance explained by the cognitive predictors only in the FLUENCY- group. A more detailed examination of the results showed that the pre-intervention reading had an opposite effect on the responses in the two intervention groups and that phonological skills contributed to the gain only in the FLUENCY-intervention. Specifically, weak initial reading skills and lower phonology scores predicted a minor response in the FLUENCY-group. This implies that children with weaker basic skills do not benefit from an intervention targeting solely their skills. Furthermore, grade contributed to the gain only in the FLUENCY þSE-intervention, which, together with the contribution of the poor initial reading skills, indicated that the older the children and the poorer their initial reading skills, the greater their gain. It can be tentatively surmised that students who continued to struggle with reading fluency over a long time benefited from an intervention that also targeted their emotions and self-beliefs. This finding needs to be verified with other samples since previous studies on combined interventions have not analyzed the effects of grade (Lovett et al., 2021; McBreen & Savage, 2022; Toste et al., 2017; Vaughn et al., 2022) and previous research on fluency interventions has not found grade effects nor observed that early intervention is effective (Maki & Hammerschmidt-Snidarich, 2022). Reading anxiety slightly contributed to the gain in the FLUENCY-group. Furthermore, although there was no significant contribution to the variance in the reading fluency gain score, Figure 1. Mean of problems in phonology and/or RAN, in anxiety and/or SE, and total problems in the non-responder and responder groups in the two intervention groups. 14 T. ARO ET AL. the correlations indicated that reading-related SE was also associated with the gain in the FLUENCY-group only. This further suggests that in addition to the children’s cognitive characteristics, their reading-related anxiety and SE may contribute to their responses differently, depending on the content of the intervention, highlighting the need for further research on emotional and motivational factors and different types of interventions. It should be noted that despite the rather high variance explained by the predictors in the FLUENCY-group, much of the variance remained unexplained. Our tentative results urge for more research that aims to understand the contributions of cognitive and emotions and motivation to intervention outcomes. Ideally, in future research, in addition to these factors, also environmental and interactional factors (see Bazen et al., 2023), as well as participants’ engagement and involvement in activity, should be considered in the design. However, such a study requires a much larger sample than the present one. Pre-intervention problems and the response status Our second main finding was that problems in the predictors, especially poor phonology skills and low SE, were associated with belonging to the non-responders in the FLUENCY-group. This was not the case in the FLUENCY þSE-group. The effects of cognitive and emotion and motivation related problems accumulated, so that having several problems in pre-intervention measures increased the probability of being a non-responder only in the FLUENCY-group. In the FLUENCY þSE-group, these problems were equally common among the responders and the non-responders. Thus, corresponding to the first main finding, this suggests that having clear problems in pre-intervention measures is associated with not showing response to an intervention targeting solely reading fluency. In the FLUENCY þSE-intervention, the students received encouraging and concrete feedback and emotional support. It might be that such support helped them better overcome both cognitive and emotion/motivation-related barriers. The findings obtained using a categorical operationalization of the intervention response are in line with those obtained using a continuous gain score as an indication of the response. Hughes and Dexter (2011) have cautioned researchers that different definitions of intervention response may result in various findings and conclusions. In the present study, both continuous and categorical approaches indicated that pre-intervention characteristics were associated with the response to the skill-focused intervention. However, although the use of cutoff score is always arbitrary to some extent, if we had used solely the continuous approach, we would not have detected the finding suggesting the association between the accumulation of rather clear problems and the weaker response to the skill-focused intervention. Limitations We recognize that these results are preliminary, and that further research is needed to substantiate them with larger samples, enabling a larger statistical power. Although our sample accurately mirrors characteristics of the Finnish school-aged population with difficulties in gaining fluent reading skill, to achieve better statistical generalization, future studies using larger as well as culturally and orthographically diverse samples are needed. Not using a randomized controlled design can be perceived as a limitation, although our choice of a quasi-experimental design was driven by our aim to achieve high ecological validity, with interventions provided as part of the school routines. Thus, our justifiable claim that these findings are generalizable to the everyday school context and special educational practices in Finland can be regarded as a strength, even if the requirements of an ideal intervention design were not met. Being unable to match the groups in terms of the predictive variables also hindered our interpretation of some findings, such as the influence of reading-related anxiety. To verify that the response to skill-focused training is more THE JOURNAL OF EXPERIMENTAL EDUCATION 15 sensitive to participants’ emotional characteristics compared to an intervention combined with emotional/SE support, future studies that will also match participants according to other than reading-related factors are needed. Conclusions Despite the shortcomings, some conclusions can be drawn for both future research and pedagogical practices. Based on the result indicating that the overall variance in the responses explained by the predictors was larger in the intervention targeting solely reading fluency, it can be cautiously concluded that in such an intervention, children’s personal characteristics may have a greater impact on their responses than in an intervention that also considers their emotions and motivation-related beliefs. However, more research is needed to corroborate the finding. Furthermore, it can be concluded—with caution—that in a skill-focused intervention, high anxiety and low SE are relevant predictors of the intervention response. However, non-linear associations require further research, but it is plausible that some factors may become relevant only when clear problems occur, that is, after a certain threshold is met, preventing, or precluding, learning. Our findings have two main implications. First, research should not ignore the significance of emotions and motivational factors for the intervention response; thus, they deserve explicit attention in intervention research on reading fluency problems. Second, the pedagogical implications of our findings are that children with several cognitive and possible emotional and motivational problems should be offered an intervention that (in addition to targeting reading skills) considers their emotions and motivation-related beliefs, for instance, by giving them systematic feedback and encouragement, as well as recognizing the emotions and beliefs related to learning difficulties and practicing the exercises to hone the required skills. Disclosure statement No potential conflict of interest was reported by the author(s). Funding The Centre of Excellence InterLearn is funded by the Academy of Finland’s Center of Excellence Programme (2022–2029) (Grant Agreement Nos. 346120 and 346119). This study was supported by the Academy of Finland (No. 264415 and No. 264344 for 2013–2015). ORCID Tuija Aro http://orcid.org/0000-0003-0004-3062 Tuire Koponen http://orcid.org/0000-0003-0039-1016 Pilvi Peura http://orcid.org/0000-0003-0915-2732 Eija R€aikk€onen http://orcid.org/0000-0003-4450-9178 Helena Viholainen http://orcid.org/0000-0002-8164-9180 Mikko Aro http://orcid.org/0000-0002-0545-0591 References Ahonen, T., Tuovinen, S., & Lepp€ asaari, T. (1999). Nopean Sarjallisen Nime€ amisen Testi [The test of rapid serial naming]. Niilo M€aki Instituutti & Haukkarannan koulu. Al Otaiba, S., & Fuchs, D. (2006). Who are the young children for whom best practices in reading are ineffective? An experimental and longitudinal study. Journal of Learning Disabilities, 39(5), 414–431. https://doi.org/10.1177/ 00222194060390050401 16 T. ARO ET AL. Aro, T., Viholainen, H., Koponen, T., Peura, P., R€aikk€onen, E., Salmi, P., Sorvo, R., & Aro, M. (2018). Can reading fluency and self-efficacy of reading fluency be enhanced with an intervention targeting the sources of self-effi- cacy?. Learning and Individual Differences, 67, 53–66. 10.1016/j.lindif.2018.06.009 Aro, M., & Wimmer, H. (2003). Learning to read: English in comparison to six more regular orthographies. Applied PsychoLinguistics, 24(4), 621–635. https://doi.org/10.1017/S0142716403000316 Bandura, A. (1997). Self-efficacy: The exercise of control. W.H. Freeman and Company. Bandura, A. (2006). Guide for constructing self-efficacy scales. In F. Pajares, & T. Urdan (Eds.). Self-efficacy beliefs of adolescents (pp. 307–337) Information Age Publishing. Barth, A., Denton, C., Stuebing, K., Fletcher, J., Cirino, P., Francis, D., & Vaughn, S. (2010). A test of the cerebellar hypothesis of dyslexia in adequate and inadequate responders to reading intervention. Journal of the International Neuropsychological Society: JINS, 16(3), 526–536. https://doi.org/10.1017/S1355617710000135 Bazen, L., de Bree, E., van den Boer, M., & de Jong, P. (2023). Perceived negative consequences of dyslexia: The influence of person and environmental factors. Annals of Dyslexia, 214–234. https://doi.org/10.1007/s11881-022- 00274-0 Borland, E., N€agga, K., Nilsson, P. M., Minthon, L., Nilsson, E. D., & Palmqvist, S. (2017). The Montreal Cognitive Assessment: Normative data from a large Swedish population-based cohort. Journal of Alzheimer’s Disease, 59(3), 893–901. https://doi.org/10.3233/JAD-170203 Breznitz, Z., & Bloch, B. (2010). Reading acceleration program (RAP) [Computer Software]. The University of Haifa, the Edmond J. Safra Brain Research Center for the Study of Learning Disabilities. Carroll, J. M., & Fox, A. C. (2016). Reading self-efficacy predicts word reading but not comprehension in both girls and boys. Frontiers in Psychology, 7, 2056. https://doi.org/10.3389/fpsyg.2016.02056 Cohen, J. E. (1988). Statistical power analysis for the behavioral sciences. LEA. Eklund, K., Salmi, P., Polet, J., & Aro, M. (2013). Tuen tarpeesta tunnistamiseen. Lukemisen ja kirjoittamisen arviointi. Toinen luokka. Tekninen opas [A screening tool of reading and spelling for 2nd grade]. Technical manual]. Niilo M€ aki Institute. Elgendi, M. M., Stewart, S. H., MacKay, E. J., & Deacon, S. H. (2021). Two aspects of psychological functioning in undergraduates with a history of reading difficulties: Anxiety and self-efficacy. Annals of Dyslexia, 71(1), 84– 102. https://doi.org/10.1007/s11881-021-00223-3 Farrington, C. A., Roderick, M., Allensworth, E., Nagaoka, J., Keyes, T. S., Johnson, D. W., & Beechum, N. O. (2012). Teaching adolescents to become learners. The role of non-cognitive factors in shaping school performance: A critical literature review. University of Chicago Consortium on Chicago School Research. Field, S. A., Begeny, J. C., & Kyung Kim, E. (2019). Exploring the relationship between cognitive characteristics and responsiveness to a tier 3 reading fluency intervention. Reading & Writing Quarterly, 35(4), 374–391. https://doi.org/10.1080/10573569.2018.1553082 Fletcher, J., Stuebing, K., Barth, A., Denton, C., Cirino, P., Francis, D., & Vaughn, S. (2011). Cognitive correlates of inadequate response to reading intervention. School Psychology Review, 40(1), 3–22. https://doi.org/10.1080/ 02796015.2011.12087725 Francis, D., Caruana, N., Hudson, J., & McArthur, G. (2019). The association between poor reading and internalizing problems: A systematic review and meta-analysis. Clinical Psychology Review, 67, 45–60. https://doi.org/10. 1016/j.cpr.2018.09.002 Francis, D., Hudson, J., Kohnen, S., Mobach, L., & McArthur, G. (2021). The effect of an integrated reading and anxiety intervention for poor readers with anxiety. PeerJ. 9, e10987. https://doi.org/10.7717/peerj.10987 Fuchs, D., Cho, E., Toste, J., Fuchs, L., Gilbert, J., McMaster, K., … Thompson, A. (2021). A quasi-experimental evaluation of two versions of first-grade PALS: One with and one without repeated reading. Exceptional Children, 87(2), 141–162. https://doi.org/10.1177/0014402920921828 Grills, A., Fletcher, J., Vaughn, S., Barth, A., Denton, C., & Stuebing, K. (2014). Anxiety and response to reading intervention among first grade students. In. Child & Youth Care Forum, 43(4), 417–431. https://doi.org/10.1007/ s10566-014-9244-3 Grills, A., Fletcher, J., Vaughn, S., & Bowman, C. (2022). Internalizing symptoms and reading difficulties among early elementary school students. Child Psychiatry and Human Development, 54(4), 1064–1074. https://doi.org/ 10.1007/s10578-022-01315-w Grills-Taquechel, A. E., Fletcher, J. M., Vaughn, S. R., Denton, C. A., & Taylor, P. (2013). Anxiety and inattention as predictors of achievement in early elementary school children. Anxiety, Stress, and Coping, 26(4), 391–410. https://doi.org/10.1080/10615806.2012.691969 Grills-Taquechel, A. E., Fletcher, J. M., Vaughn, S. R., & Stuebing, K. K. (2012). Anxiety and reading difficulties in early elementary school: Evidence for unidirectional-or bi-directional relations? Child Psychiatry and Human Development, 43(1), 35–47. https://doi.org/10.1007/s10578-011-0246-1 H€ayrinen, T., Serenius-Sirve, S., & Korkman, M. (1999). Lukilasse [Lukilasse Graded Achievement Package for Comprehensive School-Age Children]. Psykologien Kustannus. THE JOURNAL OF EXPERIMENTAL EDUCATION 17 Hudson, A., Koh, P., Moore, K., & Binks-Cantrell, E. (2020). Fluency interventions for elementary students with reading difficulties: A synthesis of research from 2000–2019. Education Sciences, 10(3), 52. https://doi.org/10. 3390/educsci10030052 Hughes, C., & Dexter, D. (2011). Response to intervention: A research-based summary. Theory into Practice, 50(1), 4–11. https://doi.org/10.1080/00405841.2011.534909 Jacobson, N., & Truax, P. (1991). Clinical significance: A statistical approach to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology, 59(1), 12–19. https://doi.org/10.1037/0022- 006x.59.1.12 Kirby, J. R., Georgiou, G. K., Martinussen, R., & Parrila, R. (2010). Naming speed and reading: From prediction to instruction. Reading Research Quarterly, 45(3), 341–362. https://doi.org/10.1598/RRQ.45.3.4 Korkman, M., Kirk, U., & Kemp, S. (2007). NEPSY-II. A developmental neuropsychological assessment. Harcourt Assessment. Korkman, M., Kirk, U., & Kemp, S. L. (2008). NEPSYIIK€ asikirja II. Kehittely, K€ aytt€ o ja Psykometriset Tiedot [NEPSY-II. Manual II. Development, use and psychometric information]. Psykologinen Kustannus Oy. Kuhn, M. R., Schwanenflugel, P. J., & Meisinger, E. B. (2010). Aligning theory and assessment of reading fluency: Automaticity, prosody, and definitions of fluency. Reading Research Quarterly, 45(2), 230–251. https://doi.org/ 10.1598/RRQ.45.2.4 Landerl, K., Freudenthaler, H. H., Heene, M., de Jong, P., Desroches, A., Manolitsis, G., … Georgiou, G. K. (2019). Phonological awareness and rapid automatized naming as longitudinal predictors of reading in five alphabetic orthographies with varying degrees of consistency. Scientific Studies of Reading, 23(3), 220–234. https://doi.org/10.1080/10888438.2018.1510936 Lindeman, J. (1998). ALLU: Ala-asteen Lukutesti [ALLU: Reading test for elementary school]. University of Turku. Center for Learning Research. Lovett, M., Frijters, J., Steinbach, K., Sevcik, R., & Morris, R. (2021). Effective intervention for adolescents with reading disabilities: Combining reading and motivational remediation to improve outcomes. Journal of Educational Psychology, 113(4), 656–689. https://doi.org/10.1037/edu0000639 Maki, K., & Hammerschmidt-Snidarich, S. (2022). Reading fluency intervention dosage: A novel meta-analysis and research synthesis. Journal of School Psychology, 92, 148–165. https://doi.org/10.1016/j.jsp.2022.03.008 McBreen, M., & Savage, R. (2022). The impact of a cognitive and motivational reading intervention on the reading achievement and motivation of students at-risk for reading difficulties. Learning Disability Quarterly, 45(3), 199–211. https://doi.org/10.1177/0731948720958128 Mohammadpur, B., & Ghafournia, N. (2015). An elaboration on the effect of reading anxiety on reading achievement. English Language Teaching, 8(7), 206–215. https://doi.org/10.5539/elt.v8n7p206 Moran, T. P. (2016). Anxiety and working memory capacity: A meta-analysis and narrative review. Psychological Bulletin, 142(8), 831–864. https://doi.org/10.1037/bul0000051 Nelson, J. M., & Harwood, H. (2011). Learning disabilities and anxiety: A meta-analysis. Journal of Learning Disabilities, 44(1), 3–17. https://doi.org/10.1177/0022219409359939 Nevo, E., Vaknin-Nusbaum, V., Brande, S., & Gambrell, L. (2020). Oral reading fluency, reading motivation and reading comprehension among second graders. Reading and Writing, 33, 1945–1970. https://doi.org/10.1007/ s11145-020-10025-5 Peura, P., Aro, T., R€aikk€onen, E., Viholainen, H., Koponen, T., Usher, E. L., & Aro, M. (2021). Trajectories of change in reading self-efficacy: A longitudinal analysis of self-efficacy and its sources. Contemporary Educational Psychology, 64, 101947. https://doi.org/10.1016/j.cedpsych.2021.101947 Peura, P., Aro, T., Viholainen, H., R€ aikk€ onen, E., Usher, E., Sorvo, R., & Aro, M. (2019). Reading self-efficacy and reading fluency development among primary school children: Does specificity of self-efficacy matter? Learning and Individual Differences, 73, 67–78. https://doi.org/10.1016/j.lindif.2019.05.007 Pikulski, J., & Chard, D. (2005). Fluency-bridge between decoding and reading comprehension. The Reading Teacher, 58(6), 510–519. https://doi.org/10.1598/RT.58.6.2 Quirk, M., Schwanenflugel, P., & Webb, M.-Y. (2009). A short-term longitudinal study of the relationship between motivation to read and reading fluency skill in second grade. Journal of Literacy Research: JLR, 41(2), 196–227. https://doi.org/10.1080/10862960902908467 Ramirez, G., Fries, L., Gunderson, E., Schaeffer, M., Maloney, E., Beilock, S., & Levine, S. (2019). Reading anxiety: An early affective impediment to children’s success in reading. Journal of Cognition and Development, 20(1), 15–34. https://doi.org/10.1080/15248372.2018.1526175 Raven, J., Court, J., & Raven, J. (1990). Manual for Raven’s progressive matrices and vocabulary scales – Section 2: Coloured progressive matrices. Oxford Psychologists Press. Richardson, U., & Lyytinen, H. (2014). The GraphoGame method: The theoretical and methodological background of the technology-enhanced learning environment for learning to read. Human Technology, 10(1), 39–60. https://doi.org/10.17011/ht/urn.201405281859 18 T. ARO ET AL. Ronimus, M., Eklund, K., Westerholm, J., Ketonen, R., & Lyytinen, H. (2020). A mobile game as a support tool for children with severe difficulties in reading and spelling. Journal of Computer Assisted Learning, 36(6), 1011– 1025. https://doi.org/10.1111/jcal.12456 Salmi, P., Eklund, K., J€arvisalo, E., & Aro, M. (2011). LukiMat-Oppimisen arviointi: Lukemisen ja kirjoittamisen tuen tarpeen tunnistamisen v€alineet 2. luokalle [A screening tool of reading and spelling for 2nd grade]. Jyv€ askyl€ a. Niilo M€ aki Institute. 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. https://doi.org/10.1177/0022219409345015 Scholin, S., & Burns, M. (2012). Relationship between pre-intervention data and post-intervention reading fluency and growth: A meta-analysis of assessment data for individual students. Psychology in the Schools, 49(4), 385– 398. https://doi.org/10.1002/pits.21599 Schunk, D. H., & Mullen, C. A. (2012). Self-efficacy as an engaged learner. In Handbook of research on student engagement (pp. 219–235) Springer US. Seymour, P., Aro, M., & Erskine, J. (2003). Foundation literacy acquisition in European orthographies. British Journal of Psychology (London, England: 1953), 94(Pt 2), 143–174. https://doi.org/10.1348/000712603321661859 Snellings, P., van der Leij, A., de Jong, P., & Blok, H. (2009). Enhancing the reading fluency and comprehension of children with reading disabilities in an orthographically transparent language. Journal of Learning Disabilities, 42(4), 291–305. https://doi.org/10.1177/0022219408331038 Snowling, M. J. (2001). From language to reading and dyslexia 1. Dyslexia (Chichester, England), 7(1), 37–46. https://doi.org/10.1002/dys.185 Stuebing, K., Barth, A., Trahan, L., Reddy, R., Miciak, J., & Fletcher, J. (2015). Are child cognitive characteristics strong predictors of responses to intervention? A meta-analysis. Review of Educational Research, 85(3), 395–429. https://doi.org/10.3102/0034654314555996 Tobias, S. (1985). Test anxiety: Interference, defective skills, and cognitive capacity. Educational Psychologist, 20(3), 135–142. https://doi.org/10.1207/s15326985ep2003_3 Torppa, M., Georgiou, G., Niemi, P., Lerkkanen, M.-K., & Poikkeus, A.-M. (2017). The precursors of double dissociation between reading and spelling in a transparent orthography. Annals of Dyslexia, 67(1), 42–62. https:// doi.org/10.1007/s11881-016-0131-5 Toste, J., Capin, P., Vaughn, S., Roberts, G., & Kearns, D. (2017). Multisyllabic word-reading instruction with and without motivational beliefs training for struggling readers in the upper elementary grades: A pilot investigation. The Elementary School Journal, 117(4), 593–615. https://doi.org/10.1086/691684 Toste, J., Capin, P., Williams, K., Cho, E., & Vaughn, S. (2019). Replication of an experimental study investigating the efficacy of a multisyllabic word reading intervention with and without motivational beliefs training for struggling readers. Journal of Learning Disabilities, 52(1), 45–58. https://doi.org/10.1177/0022219418775114 Usher, E. L., & Pajares, F. (2009). Sources of self-efficacy in mathematics: A validation study. Contemporary Educational Psychology, 34(1), 89–101. https://doi.org/10.1016/j.cedpsych.2008.09.002 Vaughn, S., Grills, A., Capin, P., Roberts, G., Fall, A., & Daniel, J. (2022). Examining the effects of integrating anxiety management instruction within a reading intervention for upper elementary students with reading difficulties. Journal of Learning Disabilities, 55(5), 408–426. https://doi.org/10.1177/00222194211053225 Wolf, M., & Bowers, P. G. (1999). The double-deficit hypothesis for the developmental dyslexias. Journal of Educational Psychology, 91(3), 415–438. https://doi.org/10.1037/0022-0663.91.3.415 Woodcock, R., McGrew, K., & Mather, N. (2001). Woodcock-Johnson tests of cognitive abilities and tests of achievement (3rd ed.) Riverside. Zentall, S., & Lee, J. (2012). A reading motivation intervention with differential outcomes for students at risk for reading disabilities, ADHD, and typical comparisons: “Clever is and clever does”. Learning Disability Quarterly, 35(4), 248–259. https://doi.org/10.1177/0731948712438556 Koponen, T., Salmi, P., Eklund, K., & Aro, T. (2013). Counting and RAN: Predictors of arithmetic calculation and reading fluency. Journal of Educational Psychology, 105(1), 162. https://doi.org/10.1037/a0029285 THE JOURNAL OF EXPERIMENTAL EDUCATION 19