Should We Be Using Words Correct Per Minute?

  • 02 May, 2020

Teacher question:

When providing fluency instruction, should time, such as the number of words per minute, be an element? Our school has been doing that, but elsewhere I’m hearing that we shouldn’t be doing that.

Shanahan response:

Fluency is a bit of a mash up and not a pure skill.

In fluency, it is important that students read the words correctly. That’s the “accuracy” part of fluency. That obviously depends heavily on decoding skills (and decoding instruction).

Unfortunately, many kids can read words accurately but still aren’t fluent readers. (In fact, that’s what got people teaching fluency in the first place—boys and girls who were on track with their phonics, but still couldn’t read well).

If reading the words right takes a lot of effort it will likely be slow and labored. Dedicating all that attention and effort to decoding the words might result in accurate reading, but it is a sure distraction from reading comprehension. Fluency is about enabling comprehension attention, not distracting from it.  

This ease of decoding is not really about speed. Nevertheless, speed is how we tend to measure it. Faster decoders are probably not needing to exert a lot of decoding effort – at least that’s the theory.

That’s where those words correct per minute (WCPM) measures come in; that’s a combination of accuracy and speed.

The problem here is that we’re using speed, but we’re not interested in speed. We’re only using speed to draw an inference about how easily the reader is decoding.

There are two ways we can increase words correct per minute.

One way is to accomplish high degrees of proficiency in decoding, and second is to hurry. The first of these ways will improve reading comprehension, and the second, not so much.

Many experts (me included), instead of talking about speed or rate these days, have been using the term, “automaticity.” The problem with that approach is that it only changes the word, but not the measure. If speed is the measure of automaticity, then we’re assuming that a change in nomenclature will be sufficient to get teachers and students to focus on ease of processing rather than hurrying.

Recently, I came across some researchers’ speculation that helped me to think about this (Pan, Yan, Laubrock, Shu, & Kliegl, 2013). They viewed fluency as enabling parallel processing—that is, making it possible for readers to multi-task, to do more than one thing at a time.

We’ve long known about what is called the eye-voice span. If readers are reading a text aloud and you suddenly switch off the lights so they can no longer see the text, they’ll keep reading for a few words. The reason that works is parallel processing. Your brain is doing one thing, while your eyes are doing something else. While your eyes are jumping forward and gobbling up information about the next set of letters, your brain is busy turning the previous set of letters into phonemes and meaning.

Good readers have bigger eye-voice span than poor readers.

Those researchers have suggested that a better measurement of fluency might be an eye-voice span measure.

Remember the point of oral reading fluency or text reading fluency is to ensure that the foundational skills are being implemented in a way that enables or facilitates reading comprehension. 

That means text reading fluency includes a third skill – beyond accuracy and automaticity. That’s where proper expression or prosody comes in. 

Readers have to read the words and they have to do that easily, but they also have to organize the words in a way that allows them to be understood. I said fluency is a bit of a mash up, and that means it includes a bit of reading comprehension as well. Not deep thinking or extended analysis or reflective comprehension, but a first-blush sense making.

What is included in this “on the fly” sense making? The reader has to use the punctuation and the meaning in order to put the pauses in the right places and to do things like read heteronyms accurately (it matters if you come up with the right pronunciations of read, live, wind, bow).

There are fancy ways of measuring prosody, but basically it comes down to this: As a listener can you follow the ideas in the text as the student read it aloud? Does it sound like the reader understands it (whether or not he/she does)? Does it sound like language?

I have no real problem with using words correct per minute as a measure of fluency. But encouraging kids to read as fast as they can is not appropriate test preparation. And, great words correct per minute with lousy prosody is not fluent.

The problem, my dear, is not in our measures, but in ourselves.


See what others have to say about this topic.

Sam Bommarito May 02, 2020 05:03 PM

Excellent points. Tim Rasisinki has written extensively about fluency/prosody. His rubric does include a measure of speed. It also includes tables by grade level of what "normal" speed for that level looks like. Speed is only one component of his multi-facited look at prosody. See his book Megabook of Fluency for details. My thought: unless you're trying to get kids ready to be autioneers reading for speed for it's own sake is a decidedly bad idea. Thanks for your insights.

Donna Johnson May 02, 2020 05:25 PM

For my first graders, I tell them that oral reading fluency is for others to be able to listen, understand and enjoy the passage. We have to know the words so we don’t stumble, watch punctuation to get the proper prosody, and use expression so others want to listen. We read aloud for others to enjoy a story and gain understanding. If we only focus on speed it defeats the purpose of understanding the passage being read.

Timothy Rasinski May 02, 2020 05:42 PM

Tim - Agree completely with your blog post. Thank you. We use WCPM in our own reading clinic as a way of assessing automatic word rec. But we also use a rubric for assessing prosody/expression. Both are important. What is troubling is when we have struggling readers show up in our clinic and when we ask them to read a passage orally so that we can observe their reading, they will ask if they should "read it as fast as I can." This indicates that they have been a good dose of reading rate exercises. Speed is a good indicator of automaticity, but it is not automaticity itself. We want students to become fast readers, but we want them to become fast the same way you, I, and other proficient readers became reasonably fast -- we read a lot.

Miriam May 02, 2020 06:21 PM

Well and clearly said. It is a measure. Emphasis on A! It is no good knowing someone weighs 250 lbs unless I also know their height and maybe what kind of shape they are in, possibly their general health and gender etc. etc. It tells me something but not everything. It is funny that teachers always seem to want things clear and simply when most things in education are a complex mess.

Aimee Reali May 02, 2020 08:02 PM

I have recently been working to strengthen my daughter’s iep goals. She can decode but struggles with comprehension. Tim Rasinki’s work helped me understand the relationship of fluency to comprehension (thank you Amplify podcast!). She has always struggled with comprehension but prosody/fluency have never been brought up or used in her iep goals—I suspect many middle school teachers are not thinking about fluency, especially in terms of oral expression. I also think that making sure goals are “measurable” can lead to easily collectible data, like WCPM. I tried to add Tim Rasinki’s fluency rubric as a measure—fingers crossed. Thank you for your insight!

Nancy Nelson May 03, 2020 02:35 AM

ABSOLUTELY NOT!!! I tutor Spanish-speaking middle schoolers. When asked to replicate a "words per minute" test, they read all the easy words and JUMP RIGHT OVER or make an unintelligible sound in place of any word they don't know. Zero comprehension, zero motivation to tackle unfamiliar words. In slower reading for comprehension, they do the same thing. I always ask what they do if they don't know a word. "I call my cousin." "I ask my sister." I now buy a used Spanish dictionary for every student and teach them how to use it. I tell them they have to read all the words to comprehend or school will be unsuccessful. But I teach them which kinds of words do not have to be looked up, when it is safe to guess or not care. EX. "...The scent of the bbq wafts through the trees." The animal was big, hairy, dangerous and [...]." It's still a lot of work so I wonder how often they take my advice.

Charlene Aldrich May 05, 2020 10:13 PM

It pains me that fluency and word recognition have been lauded as the determiners of reading ability in my state. I attended your webinar on Text Complexity and Reading Levels, and I recognize the argument that errors due to speed may not interfere with comprehension. That's a message I want the teachers in my literacy courses (assessments and disciplinary literacy) to focus on; much learning happens from reading in spite of word errors.

Brian Spivey May 06, 2020 01:37 AM

I taught high school, and I still had students struggling with fluency because of the emphasis in the elementary grades on being able to pronounce words over being able to understand them. Because this emphasis happens in the early, formative years while their brains are still developing, it becomes the "normal" way to approach reading. A shift towards comprehension and understanding will, in the long run, benefit students more, because those are the habits they will carry into the secondary grade levels and beyond.

Terry van de Beek May 03, 2020 08:40 AM

In the Netherlands "the words correct per minute (WCPM) measure" is used to discover the children with dyslexia. We now have a larger number children with a diagnosis than could be expected from research. Suppose we don't use this measure anymore, would it decrease the number of dyslexia diagnosis I wonder? Or, is there a relation between dyslexia en pushing children to read as fast as they can?

Lucia Schroeder May 03, 2020 04:22 PM

Prosody comes from understanding the text. I teach a course titled "Teaching Phonics-Basic and Beyond." The basic phonics of this letter "says" this, needs to be taught WITH the caveat of USUALLY and the need to accompany that phonics with sight words that don't follow the "rules" and words used in context. Oral reading WITH the whole class and the teacher leading is so valuable. Starting the day with a poem or song READ together and short discussion of what it means, can go a long with to help students read fluently and understand what they read. [Rasinski, T.V. (2017) Readers who struggle: Why many struggle and a modest proposal for improving their reading. The Reading Teacher, 70(5), 519-525. Dooi:10.1002/trtr.1611.] Yes, the brain is required to multitask and that means giving it time to work before we expect a response from our students.

Hanora Broderick May 03, 2020 04:50 PM

I appreciate your research and dedication to literacy.

A correction: " My dear " at the end of a blog is not appropriate.

Tim Shanahan May 04, 2020 01:53 AM

If you don’t understand literary allusion you shouldn’t criticize it.


Renee Greenly May 04, 2020 02:40 PM

As "literacy experts" we go to great lengths in providing readers with prompts that will enable students to draw meaning from text. Tim's response to this questions is making me think about a very simple idea. Reading is a like listening. Instead of hearing what someone is saying we are reading what someone is saying. Depending on the skills of the reader, a prompt could be, "Would _________ (author, character, etc) say that?" This is another way to nudge students to attend to meaning while decoding. This could make sense for some readers. You could then move into decoding prompts if needed, again depending on the student. Keeping in mind the traits of the reader and the language we use to support them has always been an important mind set to maintain as we work to increase reading proficiency.

Carolyn Doolittle May 04, 2020 02:48 PM

How you have described the use of WCPM and its usage in determining fluency progress is correct (that is, automaticity in word recognition and overall pleasing reading rate). But it is just that - a measure of one or two of the building blocks in overall fluency. The problem I am seeing, since returning to the K-12 world after spending 25+ years training preservice teachers in higher education reading instruction, is that school districts are interpreting this measure alone as a means of making some very big decisions for readers. WCPM appears to be the "do all that ends all" for deciding which tiering level a student will be placed, and whether or not he/she is making progress. Yes, it is indicative of word recognition, and yes, it can make a reader sound as if he/she knows what he/she is reading. But I see many students in 5th, 6th, 7th, and 8th grades that have great speed and word recognition - and yet can't tell me much about what they read. All they focus on is getting the words out, and doing it fast. And they have been taught this since their years in the primary grades! Many of my middle schoolers have learned to play the game - make it sound good, and it will keep me out of the reading room! Great reading rate is serving as a mask to significant comprehension problems, particularly for older readers. We have to be very cautious in our interpretations of WCPM data!

Pott Butler May 05, 2020 11:27 AM

This is very interesting. I'm focusing tuition on sentence structure while we can only meet online, including understanding independent and dependent clauses and using the passive voice. I'm mainly doing this for reading comprehension but, of course, it is also leading to stronger writing too. Ideally, this syntax work is based on the topics and projects the child is working on in class. Related to this, and necessary to bear in mind, is the specific language the child will meet in different academic subjects. A really good understanding of syllable types is helpful for developing fluency, so that is part of every lesson. For examples I can hold up in a Zoom or Messenger Kids session, I have been adapting the syllable triangles in Rasinski's Megabook of Fluency mentioned above to match the syllable types the children have learned. I find at any stage that preparing for reading by working separately on the less familiar words or phrases that will appear in the passage to be read, including exploring the meaning, has a marked impact on speed and expression. Given that even the smallest word can subtly change meaning, young readers need lots of opportunity to do this. In this way, the passage to be read is never totally unseen. Finally, going back to Rasinski's book as an example, activities that encourage modelling and experimentation with dynamics while reading do seem to raise a child's own expectations of their phrasing and expression in their reading.

Theresa May 09, 2020 01:11 PM

A pediatrician uses inches/weight on a growth chart to monitor the health of a child (which does not diagnose the problem when a child does not follow a normal trajectory). Similarly, WCPM (given naturally) can offer a measure of student reading progress and indicate (or justify) when further testing/investigation may be necessary. Both are merely progress monitoring tools - nothing more.

Bri West May 12, 2020 04:09 PM

Agreed. I never tell my students to read as quickly as they can during my instruction. So I am sure not to prompt them to do that in a probe. What I do say is to read smoothly, pay attention to punctuation, read in phrases, fix mistakes and move on etc. I have yet to see a “speed reader” fully comprehend a text at hand but then I’ve taught mostly novice readers. While wcpm is an indicator for instruction and student growth (and something that can be nationally normed for an additional piece of evidence) I wish there was more out there to measure comprehension for RtI purposes on a weekly basis . I weight comprehension higher but will always value both. If anything my students for the most part at least have fun with Fluency probes!

Linda Jul 02, 2020 09:01 PM

I worked in elementary schools and DIBELS tested all the elemenrary age readers. Then I went to middle schools and tested junior high readers. Then I went to alternative high schools and listened to many kids read AIMS passages and for whom reading is still hard. Guess what? The older they got, they more they thought good reading was fast word saying. I felt so sad, this is our bad.

L.J. Koning Jul 08, 2020 12:07 PM

Dear colleagues
it is good to read your comments because they are related to a study I am doing in the Netherlands. With us children have the idea that good reading is fast reading. I now follow children who no longer need to read quickly, but read according to adapted standards of about 1 word per second and we see what that does to the children.
Read about is at;

Kelly Parrott Mar 07, 2021 04:34 PM

I completely agree with what DR. Shanahan says: The problem here is that we’re using speed and many kids can read words accurately but still aren’t fluent readers. I have often thought that basing a child's reading ability through WCPM was not a "sole" indicator. I have had many students read with "super speed" yet not be able to retell anything they read. A combination of fluency (WCPM) and prosody are a great resource to attain a more accurate reading level. Yes, we want our students to read faster but they must also read with prosody to better understand the words they are reading.

What Are your thoughts?

Leave me a comment and I would like to have a discussion with you!

Comment *

Should We Be Using Words Correct Per Minute?


One of the world’s premier literacy educators.

He studies reading and writing across all ages and abilities. Feel free to contact him.