Modeling in Fluency Instruction

Teacher Question

At our school, we test students oral reading fluency three times a year (aimswebPlus), and we teach fluency in both our Tier 1 and Tier 2 programs. I believe that I have a good understanding of how to teach fluency using repeated reading. However, one step in that process that I’m unsure of is modeling. How much fluency modeling should a teacher do and how is that best accomplished?

Shanahan responds:
Since the National Reading Panel (2000) determined fluency instruction to be beneficial, I’ve been queried often about it. Those questions have focused on text difficulty, amounts of fluency practice, role of speed, inclusion of comprehension questions, how to pair students, and so on. I don’t remember ever before getting a question about modeling.

Some aspects of literacy instruction are obvious sites for modeling or demonstrating. Teachers often profitably provide such support when teaching comprehension strategies or when guiding kids to write or print letters and words.

It is very reasonable pedagogy to show someone how to do something and then to give them a chance to emulate what you demonstrated. Then, of course, the teacher must carefully observe those attempts at replicating the model.

If students miss the mark by too much, then the whole thing should be repeated – with the teacher again demonstrating and explaining the skill. When there is a flaw in the kids’ attempts, that problem should be addressed in the reiteration.

When kids do a reasonably good job duplicating a teacher’s model – not perfect, perhaps, but close enough, I would eschew further modeling. An emphasis on guided practice would be the better route forward in that case.

That all makes sense – and, yet the how to of all that might not be so clear when it comes to text reading fluency. What is it that a teacher is supposed to demonstrate and how do you know when to offer more modeling and when to emphasize guided practice instead?

Books and articles on fluency usually recommend modeling (Algozzine, Marr, Kavel, & Dugan, 2009; Archer & Hughes, 2011; Donaldson, 2011; Harrison, 2011; Josephs, 2010; McBride, 2016) though they rarely provide much in the way of specifics. Their descriptions of fluency teaching reveal some apparent but unacknowledged differences in opinion about both the purposes and the methods for fluency modeling. In some cases, the advice is aimed at communicating general ideas about fluency – like phrasing or reading rate. Other authorities promote a more close mimicking of the reading of specific sentences, paragraphs, or pages. 

Are practices like “reading while listening” or “choral reading” forms of modeling or types of guided practice? The literature doesn’t acknowledge the importance of that distinction.

I know of no recent studies of fluency modeling, but several such studies were meta-analyzed a couple decades ago (Chard, Vaughn, & Tyler, 2002). Basically, the researchers concluded that modeling provided clear immediate benefits – kids comprehended the texts better when they had already been modeled or they read more words right. None of the studies provided convincing evidence of higher reading achievement or fluency improvement that generalized to other passages due to modeling.

Given all that, this is one of those times when experience is the truest guide.

General Fluency Modeling

Early in the school year, when introducing how fluency will be dealt with in your classroom, it’s a good idea for teacher and kids to get on the same fluency wavelength.

Teachers who agree with this often depend on modeling to accomplish it. They read some short portions of text aloud, pointing out fluency features that they want the kids to notice. 

“I read the author’s words. I didn’t change the words or put in other words.” Or “I stopped at the periods and slowed down when I came to a comma.” With those interjections, they may reread a portion of the text to be sure the students noticed.

That seems like reasonably good instructional practice to me, though you might want to change that up a bit to increase engagement.

Personally, I’d include some negative demonstration.

Be careful here… negative models can have unintended consequences, though I’ve never seen that happen with fluency training.

I’ll read a text very slowly and ask students to evaluate my effort. They usually love to point out my problems and to tell me how to read the text. Perhaps, I’ll try it again, bragging that I’m not one of those readers who plods along too slow. This time, I’m likely to race through the text at the speed of light, so fast as to be unintelligible. If the kids were happy to point out my initial problem, now they are beside themselves. “That’s too fast, Dr. S, go slower!” they boom.

We discuss the need to read text at about the speed that we talk, neither slower nor faster. I explain why this matters: we’re practicing fluency to improve our ability to translate print into language that we can understand.

Kids don’t need much explanation for why it matters that we read the author’s words. That’s obvious even to 5- and 6-year-olds. What students often are unaware of is the importance of staying honest with yourself in this regard during silent reading. One of the benefits of oral reading practice is it forces kids to try to decode each word. That, in fact, is why oral reading practice is so good for fluency; teacher and students can hear the problems and the improvements.

But that doesn’t mean that fluency is about oral reading. Readers must be fluent in silent reading, too.

They might try every word when reading aloud, but that may not be the case when reading silently. There it is all too easy to elide and skip those unknown words or to hurry along without correcting any miscues. I might raise this issue by making some oral reading errors in my demonstration, showing the kids how I go back and fix those mistakes. Then we discuss how that needs to work during silent reading.

There are also issues of pausing, punctuation, and typography. In each case, I model both disfluency and fluency. I think the contrasting models makes the features more accessible.

This kind of introductory modeling is worth a serious investment of time – I wouldn’t hesitate to spend 30 minutes on that. I tend to do this all at once, but it can be broken into a series of shorter lessons, addressing each criterion, one at time, over several days.

Usually, such sessions end with the kids formulating a list of “dos and don’ts” for a bulletin board or we include those criteria in a partner reading evaluation form.

I know there are fluency teaching schemes that provide a short version of this with each lesson, usually with little explanation or discussion of the model (e.g. Mize, et al., 2024). I don’t agree with that approach.

Such ongoing, brief, non-explained modeling provides nothing likely to help kids read better.

Likewise, I see teachers schedule daily “model reading” lessons. What they mean by this is that they are reading a chapter book aloud to the class. Kids can learn things from being read to (vocabulary growth is the most obvious payoff). But I’d be very surprised if such reading had any impact on fluency beyond the first weeks of kindergarten.

After one or two of those modeling and explanation sessions, this aspect of modeling would be reduced to occasional brief reviews of the criteria recorded on that bulletin board.

Ongoing Modeling

Otherwise, modeling – if effective – must be briefer, more targeted, and of a somewhat different character.

What does that look like?

Ongoing modeling is more about guiding students to succeed with specific texts. Choral reading or reading while listening are examples of this kind of thing – and as such, these activities should be followed by oral reading practice of those same texts by the individual participants.

I recommend that teachers not carry out these activities with especially long chunks of text. The benefits of this kind of modeling depends heavily on student memory. Even with a full page, readers aren’t likely to remember many words or where pauses were needed.

That’s why I prefer more interactive and individual fluency activities such as paired reading with close teacher supervision, for instance.

With partner reading, the two students alternately read or guide. The reader tries to read a sentence or paragraph fluently, and the partner helps. Perhaps the partner helps with a misread word or marks the word for later attention if neither knows it. After a reading, the partner may recommend a rereading and give some feedback on the different criteria.

Kids vary in the quality of this partnering. That’s the reason for the close supervision – with the teacher monitoring different pairs throughout the lesson. By monitoring I mean more than observation or evaluation. This requires real involvement and active guidance.

The teacher both guides the fluency development, while promoting better partnering.

If students read a sentence poorly, have them try again. Often that is all that is needed to make the reading fluent (and partners often fail to ask for it). The teacher’s interjection is a lesson for both reader and guide.

In other cases, the repetition fails to have the impact I’d hoped for. The reread is as bad as the original. That is where the modeling comes in.

I read the sentence in question aloud and ask the student to try to do it like I did. That may take more than one or two models and attempts.

Such modeling informs students of the proper pronunciations of misread words, how to deal with the specific punctuation in that sentence, or how to phrase the sentence beyond those punctuation signals.

That example is at odds with recommendations I’ve seen that tell teachers to model text reading on Monday, and have kids practice that text the rest of the week. There is little likelihood that such a model would impact even the kids’ first reading attempt the next day.  

Another practice I don’t encourage is a heavy emphasis on dramatic reading, trying to make dialogue sound sad, angry, or joyful. Remember the reason for fluency work is to enable comprehension. Reading the words in a text accurately, with automaticity, and proper phrasing whether done aloud (or silently in one’s head), puts the text into a form that promotes comprehension.

Trying to invest the proper emotional tone into a text is something that comes after comprehension, not before. We use our comprehension to determine how a character must have felt so that we can then try to read those words with the appropriate emotional tenor. Guiding students to determine those emotional factors is an important part of reading comprehension instruction. Getting kids to translate that information into oral reading presentations belongs more in a speech or drama class than in one aimed at teaching reading.

What are the big take aways?

1.     The purpose of fluency work is to enable reading comprehension – not to prepare students for oratorical or dramatic presentations.

2.     The idea is to teach kids to translate print to language – both for oral and silent reading.

3.     There are two kinds of oral fluency modeling – general modeling which reveals quality criteria and specific modeling aimed at helping students to successfully read certain sentences.

4.     General modeling should include explanations of the how and why of the various fluency features or criteria.

5.     Specific modeling depends on student memory, so we should keep the modeling brief and this should be followed immediately by student efforts to replicate the model. 

References

Algozzine, B., Marr, M. B., Kavel, R. L., & Dugan, K. K. (2009). Using peer coaches to build oral reading fluency. Journal of Education for Students Placed at Risk, 14(3), 256-270.

https://doi.org/10.1080/10824660903375735

Archer, A. L., & Hughes, C. A. (2011). Explicit instruction: Effective and efficient teaching. New York: Guilford Press.

Chard, D., Vaughn, S., & Tyler, B. (2002). A synthesis of research on effective interventions for building reading fluency with elementary students with learning disabilities. Journal of Learning Disabilities, 35, 386-406.

Donaldson, B. E. (2011). Fluency instruction in contemporary core programs. Unpublished doctoral dissertation, Utah State University.

Harrison, K. S. (2011). Exploring the growth of text-reading fluency in upper-elementary English Language Learners during instruction based on repeated reading. Unpublished doctoral dissertation, University of San Diego.

Josephs, N. L. Using peer-mediated fluency instruction to address the needs of adolescent struggling readers. Unpublished doctoral dissertation, Georgia State University.

McBride, S. N. (2016). A focus on fluency: Reading fluency instruction for Tier 2 and Tier 3 Students at Lancashire Elementary School. Unpublished doctoral dissertation, University of Delaware.

 LISTEN TO MORE: Shanahan On Literacy Podcast