The Nature of Reading Problems: How to Assess Reading Difficulties for Students

In this webinar brought to you by CORE, Dr. David Kilpatrick, Professor of Psychology at the State University of New York College at Cortland and author of Essentials of Assessing, Preventing, and Overcoming Reading Difficulties, explores the science of reading and the best approaches for evaluating word-level reading.

In this webinar, you will learn:

  • Several ways to assess potential reading difficulties for students
  • Which tests David Kilpatrick recommends for evaluating word-level reading skills
  • How to correct for students using strategies such as phonic coding when performing word-level reading tests
  • How to bring the science of reading into the classroom
  • And much more!

Below is an overview of this informative and fascinating webinar.

How to Assess Reading Difficulties for Students: Evaluating Word-Level Reading 

When you are ready to assess reading difficulties for students, where should you begin? In this webinar, David Kilpatrick outlines these areas of focus for evaluating word-level reading:

Paragraph reading and fluency

  • “Reading fluency is probably the best thermometer we have to determine if there is a problem,” says David Kilpatrick
  • There are many tests available, but this assessment can be tricky; if a student has an issue with fluency, it is hard to determine which skill is the source of the difficulty

Untimed, context-free word-level reading

  • A very common staple; most test batteries include this (often called simply “word identification”)

Timed, context-free word-level reading

  • “This is the best that we have right now” for assessing the size of the orthographic lexicon, says David Kilpatrick
  • This helps correct for students who are using strategies such as phonic decoding for their context-free word-level reading; these students won’t perform as well when the clock is ticking

How to Assess Reading Difficulties for Students: Skills that Predict Word-Level Reading 

Beyond evaluating word-level reading, David Kilpatrick suggests focusing on the skills that form the basis of word-level reading or predict word-level reading.

Those skills for spotting reading difficulties for students include:

  • Letter sound knowledge
  • Phonemic blending (“I would not ever want to do a reading evaluation without [the CTOPP-2] battery,” says David Kilpatrick)
  • Letter-sound proficiency
  • Phonemic analysis proficiency
  • Rapid automatized naming
  • Phonological working memory

In the webinar, David Kilpatrick recommends specific tests to evaluate each of these word-level reading skills.

How to Assess Reading Difficulties for Students: Why Should We Assess RAN and WM?

 According to David Kilpatrick, there are many reasons we should assess Rapid Automatized Naming (RAN) and Working Memory (WM) for word-level reading:

  • Both RAN and WM moderately correlate with reading (poor RAN and/or WM usually suggests poor word reading)
  • Both RAN and WM predict reading outcomes
  • Both RAN and WM predict response to intervention (this has implications for assigning students to Tier 2 or Tier 3)
  • RAN and WM performance affects the interpretation of the broader reading profile
  • Both RAN and WM provide evidence for Specific Learning Disability (SLD) in reading

 Be sure to watch the full webinar to understand more of the science of reading and how you can apply this learning to assess reading difficulties for students in your classroom.

Video Transcript

Dr. Kilpatrick: The letter-sound knowledge, the phonemic blending and the phonemic analysis, all of them have a ready explanation within what we know about how reading works, particularly when we understand orthographic learning. Rapid naming and phonological working memory are not so obvious. I’ll come back to that as to why we would want to evaluate that even though it’s not so easy where to plug it in. That’s why you don’t see it listed under one of those or the other. You have it as a separate category. All right. With that as background, what should we evaluate? If we want to evaluate a child in terms of their reading, or screen for children, what do we want to do? Well, we want to evaluate their word reading, of course, and at the top level for word reading is going to be paragraph reading, sentence reading, and looking at their fluency, and this is important.

Dr. Kilpatrick: Reading fluency is probably the best thermometer we have to determine if there’s a problem. So if a child is not very fluent at all, there could be a number of different reasons why the child’s not fluent, but it’s definitely an indicator that there could be an issue, and I believe that’s why it got added to the Individuals with Disabilities Education Act in 2004. There are a lot of tests available for this, but here’s the problem. Even though it is a great thermometer, it confounds a whole bunch of different things, and in fact, that’s partly why it is great. In other words, it’s grouping together a bunch of skills. But if a child has problem with fluency, you don’t know what skill is at risk here, or is the source of the difficulty. So here are a few to look at. When a child is reading in a paragraph, in a sentence context, there are a bunch of skills that are being displayed and one of them is what words you already know.

Dr. Kilpatrick: So chances are, when the child’s reading, there are already some words they already know, but also there’s the skill of figuring out a word on the spot. So they’re reading along, they come across the word, how good are they at figuring out that word using phonic decoding? A third skill that’s involved in that is what’s called set for variability. And more and more research has been coming out on that in the last three, four years. What’s set for variability? The name itself won’t jump out at you as obvious to what it could mean. But set for variability, I’ll give you my definition. It’s basically the ability to properly identify in this pronounced word. So a child is reading along and, well, let me back up. Let me tell you how they’ve studied it.

Dr. Kilpatrick: They would say to children, first out of context, then in context, they’d say, I’m going to say a word, what word do you think I’m saying? I’m not going to say it the right way. And they will pronounce a word that’s irregular, but they’ll pronounce it in a phonically regular way. So they’ll say wash. And of course they’re trying to say the word wash. And as a group, about 25, 30% of children will get those correct. And then when they put it in a very, and not a totally neutral context, but not a context that’ll give it away, they will say the dog needed a wash, and the kids will say, wash. Well now putting it even in that level of context, it jumps up to about 60 or 70% accuracy. Here’s the issue, as I’ll come back to in a moment, children with higher vocabulary do better at set for variability than children with lower vocabulary.

Dr. Kilpatrick: And that really shouldn’t be surprising. But the idea is that if you’re good at set for variability, you’re more likely to properly identify a brand new word that either is phonically irregular or we could consider it phonically regular, but maybe it’s a multi-syllabic word where you don’t know where to place the stress. So the child says markette for market. And then the last is the use of contextual guessing. If you aren’t doing well with the other ones or maybe contextual guessing is going to help when you have made it through the second and third one, and you’re kind of there, like markette for market. Once again, that has a lot to do with language functioning and background knowledge, in this case, for the fourth one. So children that are stronger in that area can create the illusion that they’re better readers than they really are, which is interesting.

Dr. Kilpatrick: And so we have to watch out for that. We have kids who are compensators that are pretty weak at the first two, but the third and the fourth allow them to create the impression they’re better readers than they really are. With all that said and with all the caveats there, we need to realize that fluency is still one of our best overall thermometers for determining if there’s a problem with reading. Then, the next thing we can do maybe to eliminate some of the confounds, would be untimed, context-free word reading. This is very common staple and research studies. The most common up until just recently when there was a revision, the most common way we tested word reading was with the Woodcock Reading Mastery Test word identification subtest. I got to guess that in the hundreds and hundreds of studies I’ve read, 50% of them use that particular subtest, and the other 50% involved other commercially available tasks or experimenter design tests.

Dr. Kilpatrick: So you read words off a list, they get harder and harder and harder. That’s context-free word reading. Often called just word identification, and most test batteries have this. The problem is this also confounds three of the four. A child taking such a test already knows some of the words, in all likelihood, and then they might be able to sound some of them out, and if you didn’t know the word beforehand and you sounded it out on the spot, you get the same score as if you already knew it. It can’t distinguish between those. And then set for variability. As I said earlier, even without a context, children can do decently well with set for variability. Then if we want to narrow it down even better, this is the best that we have right now. Timed context-free word reading. Because if you have to use some sort of strategy like phonically decoding or inferring from set for variability, the clock is ticking and you’re not going to do as well.

Dr. Kilpatrick: So here are some tests, three that are available. Fourth, that’s going to be available shortly. The Test of Word Reading Efficiency, Second Edition (TOWRE-2), you have a list, 45 seconds. Start with easy words. They get harder as time goes on. The Test of Silent Word Reading Fluency (TOSWRF-2). Some people might not like it. They may not think of it as authentic. I think it’s a great test. It’s been used in research studies where you have a whole bunch of words that run together and you have to put a slash mark between the words, and if you have a large orthographic lexicon or sight vocabulary, those words jump out at you. You know right where to put the lines. If you don’t, the words aren’t jumping out at you and it takes you longer. The new Kaufman (KTEA-3) has both timed and untimed word identification, and then the upcoming WIAT-IV, the Wechsler Individual Achievement Test, the fourth edition in 2020 will also have a timed and untimed word reading.

Dr. Kilpatrick: Now, what those are doing is they’re giving us a much better estimate of the size of the sight vocabulary, so kids that already know a whole lot of words because of the timed nature of them, they’re going to get quite a few of those words right. Now, the kids that don’t have so many words that are already familiar with them, they’re going to be trying to figure it out as they go along, the clock is ticking, they get a lower score. So this is a very important set of assessments that we can use that, in this case, narrow it down to looking at the size of the sight vocabulary.