Study Suggests Regression in Foundational Reading Skills across Primary Grades

Hello Friends!

I just read this very new and important study conducted by David Page and colleagues. It takes a look at the foundational reading skills of 250 students from 1st-3rd grades at relatively high-SES schools. Reading instruction in the schools was described in this way:

"There was also a strong sense of teacher autonomy across the district accompanied by the belief that both principals and teachers were knowledgeable in quality reading instruction and were doing a great job. This resulted in giving teachers a carte blanch approach to teaching reading. In lieu of explicit foundational skills instruction, there was an ongoing initiative focused on the inclusion of reading instruction that could help students hone their metacognitive thinking."

The research resulted in a number of findings, but I want to share three major ones. Now, before going further, it is important to stress that this was not a longitudinal study that followed the same group of students across three years of time. Instead, it was a cross-sectional study that measured the skills of students in 1st, 2nd, and 3rd grades at beginning, middle, and end of year. So, really we are looking at the growth of three grades of students across one school year and comparing the performance of similar (but not the same!) groups of students across 1st-3rd grades. This is important. Comparing where one group of students are at in first grade to where a different group is at in 3rd grade, even when the groups have similar composition in terms of variables like SES, is just not as conclusive as assessing the same students over a three-year period. So, to me, this makes the findings here "suggestive" rather than conclusive. 

That said, here are the three major findings, in, more-or-less, the authors' words:

1) "Our results find that in this high-SES district, an unexpected 30% of students struggle with reading acquisition."

2) "Our cross-sectional results show that the percentage of students who struggle
with reading remains remarkably consistent across first-, second-, and third-grade. The results suggest as Stanovich (1986) hypothesizes, the Matthew effect in reading where those who struggle stay consistently behind their proficient peers." 

3) "This study contributes to existing research by first revealing that it is possible for students entering first-grade with high-attaining reading skills likely due to SES advantages, to experience declining progress relative to national norms to the point where they have become average by third-grade."

The authors add a bit more detail to this third finding here:

"Figure 2 through 4 and the results to research question three clearly show declining first- to third-grade attainment for sight-word reading (77th to 37th), pseudo-word reading (70th to 37th), and accumaticity (52nd to 23rd). Secondly, Figure 5 reveals that the declines seen in Figure 2 through 4 were not solely attributable to the struggling reader group. For the proficient reading group, sight-word reading drops from the 81st to the 61st percentile, pseudo-word reading drops from the 70th to 58th percentile, and accumaticity declines from the 48th to the 28th percentile. Struggling readers experience large declines in sight- and pseudo-word reading that results in attainment between the 16th to 18th percentile on all three measures."

Again, I want to stress that these were not the same students who actually declined from, say, the 77% to the 37% in decoding from 1st to 3rd grade. However, they were similar first graders and third graders at the same schools. So, I still find this outcome very telling. This district eschewed a strong, consistent, scientifically-based program of reading instruction and instead gave teachers "carte blanche" freedom with regard to reading instruction and tended to de-emphasize instruction in foundational skills from 1st-3rd grades. Looks like a mistake to me. This cross-sectional look indicates that both higher- and lower-performing readers in second grade scored much lower with regard to a national norming sample than did their first-grade peers, and the same thing was true for third graders in comparison to second graders. In the end, the authors conclude,

"These results call into question the perceived notion that students with accomplished foundational reading skills in first-grade are destined to become fluent readers. Also in question is the extent to which scientifically-based research had informed the teaching of foundational skills in the district."

It seems clear to me that the research to guide the practice of reading instruction in primary grades is out there. But, will we chose to listen?


David D. Paige, Grant Smith, William Rupley & Will Wells (2020) Reducing High-Attaining Readers to Middling: The Consequences of Inadequate Foundational Skills Instruction in a High-SES District, Literacy Research and Instruction, DOI: 10.1080/19388071.2020.1780653

A New Study on the Effects of Letter-Sound Instruction in Kindergarten

 Hello Friends! 

I just read this very recently published research study from Norway that addresses a very important practical issue in teaching beginning literacy: the pace of instruction of letter-sound relationships in kindergarten. This topic hasn't been subject of much research (although note that Jones and Reutzel [way to go Dean Ray!] did examine it in a quasi-experiment in Utah). The Norwegian study was a basically a "natural" quasi-experiment and involve a very large sample of kids and classrooms. Importantly, the research found that a faster pace of introducing letter sounds in kindergarten led to stronger outcomes in letter knowledge, word reading and spelling and was particularly important for students who enter school with poor letter knowledge. Here are some of the key takeaways in the authors' own words...

Discussion

This study was designed to find out whether a faster pace of letter instruction contributes to the development

of letter knowledge, word reading, and spelling skills in the first year of school, and whether a faster pace of

letter instruction reduces or increases the probability of very low or very high scores on literacy measures. In

general, the findings from the present study were in line with the hypotheses put forward. More specifically,

children in classes with faster letter instruction performed significantly better on all outcome measures and

were less likely to score among the lowest 10% and 20%. Also, a faster pace of letter instruction was

significantly associated with word-reading accuracy in the higher end of the distribution. These findings are

in line with that of Jones and Reutzel (2012) that a faster pace of letter instruction affects children’s

development in letter knowledge. The present study adds to the existing literature by showing that outcomes

in word reading and spelling are also associated with the pace of letter instruction.

Effect of letter-instruction pace on the development of letter knowledge, word reading and

Spelling

The observed effect of letter-instruction pace on letter knowledge supports the hypothesis that

a faster pace gives children better opportunities for sufficient repetition and practice of the

which helps them to decode different words. In addition, especially for sight word efficiency,

obtaining a high score requires keeping representations of several words in one’s memory,

and the ability to do this is best acquired through multiple encounters with words, through

both reading and writing (Adams, 1990).

The explanation for the significant effect seen for spelling seems to follow the same line of

reasoning as for word reading. Young children typically read words better than they spell

them, as spelling requires grapheme retrieval and is more dependent on memory and ample

practice while word reading requires only grapheme recognition (Perfetti, 1997). Still, a faster

pace of letter instruction seems to provide children with better opportunities to develop their

spelling skills as well, possibly as a result of both better knowledge of the letters and more

time to practice.

Letter-instruction pace and the likelihood of tail-end scores for letter knowledge, word

reading and spelling

Children who first start school are typically a very heterogeneous group in terms of literacy, ranging

from children who know only some of the letters to children who are already fluent readers and

writers (Justice et al., 2006; Piasta, 2014; Sigmundsson et al., 2017). Letter knowledge at school entry

is known to be a strong predictor of the development of reading skill (H. Catts, Fey, Tomblin, &

Zhang, 2002; Leppanen et al., 2008; National Early Literacy Panel, 2008; Schatschneider et al., 2004),

and if children who know few letters at school entry have to wait, say, 19 weeks before they are

introduced to a letter that they need in order to read and spell, this will delay their ability to read and

write many words. Our findings indicate that contrary to the highest performing children, the lowest

performing children benefit from a faster introduction of the letters on all outcome measures. These

results partly contradicts the results by Connor et al. (2004) to the effect that the level of literacy

skills (letter recognition, letter knowledge, and word reading) at school entry is a stronger predictor

of literacy development than classroom instruction.

Jones et al. (2013) highlight the opportunity to repeat and practice the letters more often during the

first year of school as a particular benefit of introducing them faster. Our findings suggest that this is

particularly important for children with poor letter knowledge at school entry – who typically need more

explicit exposure to letters in order to learn them sufficiently well (Jones et al., 2013; Piasta & Wagner,

2010b; Treiman et al., 2007). In classrooms with a slow instructional pace, children who need such explicit

instruction will typically learn those letters that are introduced early on quite well, as they can repeat and

use them more often during the academic year. However, because most children will have automatized

their letter knowledge during the first year, there tends to be rather less explicit letter instruction in later

years. As a result, the children needing such explicit instruction may not have sufficient time to

automatize those letters that are introduced towards the end of the first year. Hence, children already

at risk of reading difficulties are further disadvantaged by a slow pace (Jones et al., 2013). As pointed out

earlier in the discussion, letter knowledge represents one of the steps of the early development of literacy

skills, and children in classes with faster letter instruction are in fact less likely to perform poorly in word

reading and spelling. This is in line with Jones et al. (2013), where it is emphasized that the purpose of

letter knowledge is reading and writing and that a slow pace of letter instruction takes up valuable time,

leaving children with less time to develop their reading and spelling skills.

Kristin Sunde, Bjarte Furnes & Kjersti Lundetræ (2019): Does Introducing the

Letters Faster Boost the Development of Children’s Letter Knowledge, Word Reading and Spelling

in the First Year of School?, Scientific Studies of Reading, DOI: 10.1080/10888438.2019.1615491

Phonics/Decoding Doesn't Need More Support, But Here is Some Anyway...

 Hello Friends!

A couple of my previous posts have dealt with the vital nature of robust phonics/decoding instruction. There is, of course, endless research support for such instruction. However, this recent study from England was called to my attention, and I thought that it was worth sharing. The abstract does a good job of summarizing the basic findings, so I have pasted it in here. What I find particularly noteworthy is the documentation of long-term effects - the impact that developing/not developing proficiency in decoding has on kids' reading comprehension four years down the line. Connie Juel reported similar long-term effects resulting from proficient/lack of proficient decoding back in the 80's. Hopefully, schools will hear this message better in 2020 than they did back then!

Kit S. Double, Joshua A. McGrane, Jamie C. Stiff & Therese N. Hopfenbeck. (2019). The importance of early phonics improvements for predicting later reading comprehension. British Educational Research Journal,45(6), 1220–1234.

The role of phonics instruction in early reading development has been the subject of significant conjecture. Recently, England implemented a phonics screening check to assess the phonetic decoding of 6-year-old students, to ensure that all students master this foundational literacy skill and attain adequate phonemic awareness in the early years of primary schooling. Students who fail this check are obliged to retake the assessment the following year. In this article, we compare the performance of students who initially pass this check (pass) and students who fail the original assessment but pass the retaken assessment (fail–pass), with students who fail both the original and retaken assessments (fail–fail). Using data from the Key Stage 1 assessment of reading and the Progress in International Reading Literacy Study (PIRLS), we examined the reading comprehension performance of these students approximately 1 and 4 years after their first phonics screening. The results suggested that fail–pass students performed substantially better than fail–fail students, even after performance on the initial phonics check was controlled for. While fail–pass students do not appear to entirely catch up with pass students in reading comprehension, their relatively better performance underscores the importance of intervening for those students who are identified as having problems with phonetic decoding to increase their likelihood of success at reading comprehension in later schooling.

Vocabulary Instruction and Google Images

Google Images has made it extremely easy to pair an image with a target vocabulary word. In the MCVIP & VALE research projects, our teachers paired images with all target words on vocab word walls. Here is a typical word wall that features target words with images pulled from Google.

With regard to instruction, an image may be all it takes to get across the meaning of a concrete noun like musket or sandspit in a memorable way. But, with other kinds of words images may be even more valuable as a teaching tool, serving as an excellent way prompt to students to think deeply about or apply a word meaning. Consider the word dismayed. One of our teachers paired this image with this target word...

After providing a kid-friendly definition of dismayed and several examples of use, we could then ask students these type of questions about the image:

• Why do you think that I would show you this picture for the word dismayed?
• If you were in one of the tall buildings, would you be dismayed? Why or why not? If you were on the top of one of the mountains in the background, would you be dismayed? Why or why not?
• Can you give me (or write) a sentence about this image picture that includes the word dismayed?

Through these kind of questions, images become great contexts for students to apply new words and consider the nuances of their meanings.

Finally, just committing to finding a Google Image for each of your target words ensures that you are planning ahead for your vocabulary instruction. Such planning always leads to crisper, clearer introductions and discussions of new word meanings. 

The Impact of Word Reading and Vocabulary on Middle School Reading Comprehension

 Hello Literacy Friends!

As I continued to update my knowledge for my upcoming Literacy Difficulties course, I came across a very recent article that examines the relative impact of word reading and vocabulary knowledge on the reading comprehension of adequate and struggling adolescent readers. I found that the article does an excellent job laying out these relationships and thus gives real insight into the causes of differences in reading comprehension among adolescents. I will try to unpack the findings in an efficient manner for you all here.

First, the study included 796 middle schoolers (average age of 14.1), with 491 in the adequate comprehension group and 305 in the struggling comprehension group (below 30% on the Gates comprehension test). These students were assessed for vocabulary knowledge, word reading, reading efficiency (fluency), and inferencing. The researchers analyzed the direct and indirect effect of word reading and vocab knowledge on reading comprehension. (Indirect effects included the effect of each these two components on reading efficiency and inferencing multiplied by the respective effect of reading efficiency and inferencing on reading comprehension.) In the end, the authors could then consider the total effects of word reading and vocab on reading comp for a) the total sample, b) the adequate comprehenders, and c) the struggling comprehenders. So, here, in a nutshell, are the outcomes:

1)    Word reading was a significant predictor of comprehension for the entire sample (effect .21); however, word reading had a large effect on comprehension for struggling comprehenders and a negligible effect for adequate comprehenders. This supports the notion that, in the authors’ words, “there is a threshold of word reading skill that has to be met before its relation to reading comprehension subsides. It is possible that the struggling comprehenders did not have the requisite word reading skills necessary to easily extract meaning from text.”

2)    Vocabulary was a greater significant predictor of comprehension for the entire sample than word reading (effect .60). Vocabulary had a larger effect on comprehension for adequate comprehenders (.62) than struggling comprehenders (.18). The authors summed up these results in this way, “While vocabulary favored the adequate group as a predictor, it was also a statistically significant predictor for the struggling comprehenders. Thus, vocabulary plays an important role for struggling and adequate readers alike.”

3)    Vocabulary knowledge significantly positively impacted the adequate readers’ reading efficiency and inferencing. In simple terms, students with higher vocabulary scores were slightly more fluent readers and better at making inferences.

4)    Without getting into the more complex details, the outcomes related to the struggling comprehenders led the authors to conclude… “instruction in word reading and vocabulary may lead to greater gains for struggling readers than instruction aimed at increasing fluency. Indeed, this view is supported in the synthesis of effective interventions for struggling adolescent readers by Edmonds et al., who found that word reading instruction led to greater and more consistent gains than fluency instruction.”

I would summarize things this way. It is clear that word reading difficulties have a strong impact on students’ comprehension. Basically, until students can read the words well, other components matter less: the students are simply hung up on reading the words. On the other hand, once students reach a high level of word reading proficiency, knowing the meanings of the individual words greatly impacts their reading comprehension, both directly and because it makes them better at inferencing and slightly more fluent. 

My takeaway is that elementary schools should prioritize both of these key components! I believe that it quite feasible for elementary schools to equip nearly all students with strong word reading skills, and, if we can, we should. But, just as significantly, vocabulary knowledge was a significant predictor of comprehension for all students and particularly for the adequate comprehenders. Teachers often ask me, “We rightfully think a lot about our lower-performing students. But, how can we push those kids right in the middle to a higher level? Well, this study shows that if we want students to be especially strong comprehenders in middle school, we need to be teaching vocabulary comprehensively and consistently across the elementary grades. So, at this point, I’ll put in a plug for my earlier CLC post on VALE vocabulary instruction. Check it out below! 

Oslund, E., Clemens, N., Simmons, D.. & Simmons, D. (2018). The direct and indirect effects of word reading and vocabulary on adolescents’ reading comprehension: Comparing struggling and adequate comprehenders. Reading and Writing, 31, 355-379.

Kindergarten Reading Intervention: The Impact of Intensity and Instructional Design

In prep for my upcoming Literacy Difficulties class, I recently read a research article on reading intervention in kindergarten that struck me as very important, and thus I thought that I would attempt to briefly summarize the findings here. As I guy who has said for many years, “It is not just about teaching the right stuff, but also doing it at the right level of intensity,” I loved the design of the study. Specifically, the researchers compared outcomes for students who participated in three differently designed kindergarten interventions. The three interventions were designed to compare the effects of two key variables – time and “design specificity.” So, the three treatments were… 1) 30 minutes of HIGH-design-specificity, code-based instruction, 2) 15 minutes of HIGH-design-specificity, code-based instruction + 15 minutes of read aloud, vocab and narrative structure instruction, and 3) 30 minutes of MODERATE-design-specificity, code-based instruction. Based on these interventions, the researchers could compare groups 1-2 to examine the effects of additional instructional time on key code-based objectives (e.g., phonemic awareness, phonics, beginning decoding/spelling) and groups 1-3 to examine the effects of design specificity. There were many assessments used and thus many comparisons. I will try to do justice to the findings very briefly…

1)    Students in the 30-minute HIGH treatment scored significantly greater growth than students in the 15-minute HIGH + 15-minute story reading in letter naming and writing, decoding, word reading and spelling. The two groups performed equally on phoneme segmenting. Importantly, the 30-minute HIGH treatment was differentially more effective for students who entered K with the lowest alphabet knowledge (naming 3 or fewer letter names – 70% of the total sample). In short, for the students with the lowest alphabetic knowledge, 30-minute of highly specifically designed code-based instruction was much more effective than 15. 15 minutes of such instruction was equally effective for students who were at the “top” of the bottom 25% in alphabetic knowledge. In simple terms, kids who start really low in terms of early literacy knowledge need not just carefully targeted instruction but also “the right intensity” (time on task) of this instruction. 

2)    Students in the 30-minute HIGH treatment scored significantly greater growth than students in the 30-minute MODERATE treatment in letter naming and writing, decoding, word reading and spelling. Similar to the above comparison, the 30-minute HIGH treatment was differentially more effective for students who entered K with the lowest alphabet knowledge (naming 3 or fewer letter names – 70% of the total sample). In short, for the students with the lowest alphabetic knowledge, 30 minutes of highly specifically designed code-based instruction was much more effective than 30 minutes of moderately specifically designed code-based instruction. So, once again, to maximize growth of kids who start really low in terms of early literacy knowledge, it is critical that the instruction not just target the “right stuff” but also have excellent design.  

Overall, there were a lot of effect sizes involved. So, practically speaking, I will just say this to give a sense of the magnitude of the gains: The kids in the 30-High treatment ended up at the 67% on Dibels decoding whereas those in the other conditions ended up at about the 40%. And, a little info on “specifically designed instruction:” this variable involved elements such as explicitness of big ideas, scaffolding, and strategic review.

The bottom line here seems quite clear: Just generally “doing the right stuff” doesn’t cut it with our neediest students; maximizing the growth of these kiddos requires carefully designed code-based instruction and A LOT OF IT!

Simmons, D., Kame’enui, E., Harn, B., Coyne, M., Edwards, L., & Thomas, C. (2007). The effects of instructional emphasis and

specificity on early reading and vocabulary development of kindergarten children. Journal of Learning Disabilities, 40, 331-347.

Implementing Multi-Tiered Support Systems in Reading in the Messy World of Real Schools

Hello Friends!

In preparing for my annual spring term Literacy Difficulties course, I came across two excellent, recent pieces on implementing MTSS in reading that I have not used in the past (refs at bottom). Both pieces issued from a successful collaboration between a State Department of Education and researchers at a state university that focused on supporting four very high-needs schools in developing, implementing, and refining school-wide MTSS systems in reading. Significantly, the project produced the kinds of outcomes that most schools hunger for: Two years of implementation of the MTSS systems produced a +0.5 effect size in students’ reading achievement, meaning that students who received two years of supports accelerated their reading performance by 19 percentile points.

The pieces, which overlap quite a bit, begin with the premise that most schools know, in general, what needs to be done in terms of implementing MTSS in reading but “underestimate

the supports that schools need to build systems and infrastructure to implement and sustain these practices.” The authors then go into detail in terms of how the project schools “worked to overcome the complexities inherent in implementing multitiered reading supports in high-priority schools,” providing "examples of how schools involved with the K–3 reading initiative delved into the details to move past barriers and build the systems and infrastructure to implement a comprehensive MTSS framework fully, with fidelity and consistency.” I found that pieces were highly engaging, filled with practical insights, and extremely challenging, as they describe many specific practices that enabled the schools to be successful but that, quite honestly, most schools typically just don’t have in place.

It would take quite a bit of verbiage to adequately summarize the content of the pieces, and I was unable to upload a chart from the Coyne chapter that would have done it for me. So, here is the briefest of summaries. Both pieces identify a set of four common stumbling blocks, that they present in the form of practical statements, that often trip up schools’ attempts to implement MTSS in reading in robust ways: 

• "We have an MTSS plan, but it doesn't guide our day-to-day reading practice."
• "We have identified a common approach to Tier 1, but it doesn't seem like there is consistency in reading instruction across teachers and classrooms."
• "We have useful reading data from our students, but it feels like we are not able to use it to make meaningful instructional decisions for all our students."
•  "We have students who need intensive small-group instruction, but now what?" 

The pieces then go on to present examples of exactly how the four schools overcame these obstacles, detailing the establishment of literacy leadership teams and the teams' long- and short-term objectives; decisions regarding planning, prioritizing, and meeting; and tools for utilizing data. Again, it is beyond the scope of this post to lay out these "solutions" in detail. But, if these stumbling blocks sound familiar and the idea of actual practical solutions is thought-provoking, email me and I can send the chapter/articles to you in PDF. They will provide much more specific description of the practices and tools that the project schools employed to address these issues.

Overall, I would leave you with these following critical points:

1)    Most schools vastly underestimate the challenges involved in implementing a robust MTSS system in reading and greatly overestimate the degree to which MTSS practices are implemented fully and with fidelity.

2)    Partial implementation of MTSS systems in reading likely will not improve student outcomes, particularly for students with, or at risk, for learning disabilities (Balu et al., 2015). "Half way" just isn't good enough, and ends up being a waste of vitally important resources. It really takes “everything and the kitchen sink” to produce the outcomes we desire.

3)    There are schools out there – in extremely challenging settings – that are “getting it right.” If they can do it, you can do it! And, if you can do it, you should do it!

4)    These two pieces will give you plenty of real, practical guidance in constructing a robust MTSS system in reading. I would also be happy to “join the discussion” if your school is genuinely interested in undertaking this process.

Coyne,M. D., Oldham, A., Leonard, K., Burns, D.,&Gage,N. (2016).Delving into the details:
Implementing multitiered K–3 reading supports in high-priority schools. In B. Foorman
(Ed.), Challenges to implementing effective reading intervention in schools. New Directions for
Child and Adolescent Development, 154, 67–85.

Leonard, K., Coyne,M. D., Oldham, A., Leonard, K., Burns, D.,& Gillis, M.B. (2019). Implementing MTSS in Beginning Reading: Tools and Systems to Support Schools and Teachers. Learning Disabilities Research & Practice, 34(2), 110–117