News from the Center for the Study of Adult Literacy

It has been awhile since I have posted news from the Center for the Study of Adult Literacy. For those of you unfamiliar with our Center, we are a federally funded national research center charged with focusing on adults who read between the 3.0-7.9 grade levels. We are specifically looking at their reading related strengths and weaknesses, and developing/modifying a hybrid 100-hour reading curriculum (see csal.gsu.edu for more information). 

I thought that many of you would be interested in knowing about our library of web-based texts that we have collected for adult learners. This library includes over 1,500 different materials for learners to read. You can access the library and find out more about it at: http://csal.gsu.edu/content/library  

We would love to get feedback from you about it, so please access the contact form at: http://csal.gsu.edu/content/contact   or email me directly at: dgreenberg@gsu.edu.  We would especially love to hear whether you have other texts that you recommend (they need to be accessible for free on the Web). This resource is only a week old, so we anticipate making changes as time goes on. I hope that you will help us improve our repository!

 

Thanks,

Daphne

 

Daphne Greenberg

Georgia State University

Comments

Thanks for sharing, Daphne!  Looks like you have put together quite a good resource here.  For the Career Pathways COP, I think the section on Jobs and Work under Adult Learners will be especially helpful. 

I noticed that the direct link that is in your message goes to an error message, but if I type http://csal.gsu.edu/content/library into my go box, I am able to access it. Hope that works for everyone who is having trouble!

Donna Brian, SME Career Pathways

Thanks, Daphne! 

This is set up nicely. There's a great range of topics and it's good to have the "easier...medium...harder" categories. I would like to know the criteria you had for including the resources in general, and for putting them into these categories. Can you please share some of that with us?

Julie

Lots of good questions...

Unfortunately, there are not a lot of web-based texts available for our adult learners. So we sent out "feelers" to try to find these texts. I posted a query a long time ago on these community of practice discussion lists asking people for ideas. We also consulted with the CDC and with adult learner organizations. These leads led to further suggestions, until we started to accumulate our various texts. But, we are still looking for more texts (in fact we have about another 300 that we have gathered more recently that we will add to the repository, hopefully in the next few months), so if you or anyone can suggest web-based texts designed for adults who struggle with their reading, please send them to me (dgreenberg@gsu.edu).

You also wanted to know about the easier, medium, harder categories. We used a readability system called Coh-Metrix to help us. As explained on the teacher page of the library (http://csal.gsu.edu/content/are-you-teacher), Coh-Metrix takes into account 108 different text-based factors to estimate the difficulty of a text. Coh-Metrix provides descriptive text information (number of paragraphs, sentence count, word count). It also considers linguistic characteristics (lexical diversity, syntactic complexity) and discourse characteristics, such as text coherence and genre (narrative versus informational text). It is this combination of factors that allows reading materials to be scaled on dozens of aspects of language and discourse. To learn more about  Coh-Metrix go to:http://csal.gsu.edu/content/readability Of course, as with any readability system, familiarity of the topic can make a big difference for the adult learner (making a text harder or easier regardless of the readability system).

I hope I have answered your questions!

Daphne

Thanks, Daphne!

I have never heard of Coh-metrix! I look forward to checking it out. But to give us a sense of comparing it to something we are familiar with, how would you describe your criteria for the easier, medium and harder texts in terms of grade level?

Sorry to press, but I want to be able to compare with things I know! 

Also, I would love to hear your opinion of how Coh-Metrix works as a more complete assessment of reading ease. I have never seen a formula that takes into account the factors you mentioned, which are really important.

Thanks,

Julie

As always, really good questions. Our Center focuses on adults who read between the 3rd and 8th grade levels. Based on the texts we could find, and our Coh-Metrix analyses, the texts were placed into the categories, keeping in mind the grade levels we are focusing on. In terms of your second question, Coh-Metrix is the most complete readability tool that is currently out there. As you mention, it takes into consideration many factors-something that other readability measures do not do. However, readability measures still have a VERY long way to go, especially for our adult learners and therefore should be taken as ONLY one source of information when trying to figure out suitability for a specific adult. Adults come with all kinds of background knowledge, disabilities, and strengths. Therefore, an adult may read at a certain level on a standardized test, but find a text easier or harder to read than would be expected based on their background knowledge about the topic, their motivation/persistence to tackle different types of materials, etc.

Daphne

Hi Daphne,

Thanks for posting this resource.  I will check out your content library next.  

The topic of low level readers and content for teaching them to read is very appropriate.  In a discussion thread labeled "Despite a Low IQ, Can Students Still Learn to Read?", we have been talking about this very topic.  

Can you answer this question about your 1,500 different materials?  Do any of them cover Independent living skills and functional, very basic academics skills?

Thanks for the good work you do for our field.

Rochelle Kenyon, SME

 

Hi Rochelle,

Thanks for your question. Unfortunately, we do not have any that cover very basic academic skills. However, we do have a computer skills tab under our resources tab, that may have some digital literacy skills that you may find helpful (http://csal.gsu.edu/content/computer-skills). Independent living skills may be found in various categories, but maybe best found under the advice category (http://csal.gsu.edu/content/advice-test). However, if you are talking about individuals with very low IQ, the types of independent living skills that they need may not be found in our library. We would love to include those kinds of topics if you can suggest any web-based texts for us to include! Please email me at dgreenberg@gsu.edu with any suggested texts.

Daphne

Hello Priyanka,

I completely agree that these materials will be a great springboard.  Corrections classrooms will be able to use the Money, Jobs and Work sections to which you refer, which offer a nice cross-section of basic education, with a great practical application component.  Many thanks to Daphne!

-- Heather 

Thank you for sharing this well organized list of reading materials! If the ESL Literacy Readers are not already included in your collection, you may want to add them. These were developed by our wonderful colleagues at Bowe Valley College in Canada and available at the ESL Literacy Network http://www.esl-literacy.com/.

Cheers,

Susan Finn Miller

Moderator, Assessment CoP

Daphne,

This is great information!  I know that it is mostly for literacy but I can see incorporating learning about pollution (building a context) for later math discovery. As we all know that helping adult learners see the purpose of what they are learning in the real-world is the most effective way to engaged them.  I will use these in both my math and science classroom teachings to help lead into the mathematical concept.

Brooke

Hello,

I realize you are using a new kind of readability index at CSAL; however, I came across an article and a blog post by eminent researchers in the field of literacy that question the entire enterprise of such indexes.  In short, there is no real evidence that such indexes work in the way many think they do: Daniel T. Willingham (http://www.realcleareducation.com/articles/2014/03/26/reading_is_not_formulaic_why_equations_cant_be_920.html); Tim Shanahan (http://www.shanahanonliteracy.com/2011/08/rejecting-instructional-level-theory.html).  What I think an over-reliance on such indexes do is underestimate the capability of adult readers.  For example, your "Hard" section in the library is, I think, far too easy--especially if such selection are being read with tutors/teachers.  I also strongly object to the terms "real-life stories" and "made-up stories."  Adults, even if they are reading at 3-8 grade levels (such levels are not without there own problems/inaccuracies) are not children, and I don't think we've really found a way to teach struggling adult readers as such.  

Thank you,

John Schlueter

Thanks for this reference, John! It was interesting that the Dale-Chall was the only one that had good reliability, and it was the only one that used a word-familiarity metric. That makes so much sense! It also makes me wonder how the Coh-Metrix formula would compare. It is nice to see readability formulas that go beyond sentence length and syllable count.

I also thought it was interesting that they used oral reading fluency as a proxy for reading skill because they say that it relies on "higher-level meaning processing". So, they have evidence that you must have some comprehension in order to have good oral fluency? I had never heard that!

Yes: I think your question regarding the use of oral fluency underscores the difficulty of even measuring comprehension in any formulaic way.  This would, in turn, make the use of formulas that measure the difficulty of comprehending texts, no matter how sophisticated, a dubious enterprise with little chance of meaningful success from the outset. 

While I agree that readability formulas have their drawbacks, they can serve as a starting point to help teachers select texts for students. Other factors that teachers usually consider are student interest and the purpose for reading. Many students can read a higher level text if they are interested in the topic. Also, readability formulas are generally thought of as a predictor of how well the student can "read" a text but not really a predictor of how well a student will comprehend. A student who is a good decoder might be able to read one of the "harder" texts in the library but not necessarily understand what he or she has read. As for the categories of easier, medium, and harder, teachers would likely preview the texts before using them with their students and thus make the final determination as to whether the text is suitable for a particular student or groups of students. In short, readability formulas are a helpful tool but teacher wisdom is what will determine which text is right for the students.

I agree. I use these formulas from a health literacy perspective, to re-write health information, forms, etc so that the average person can easily understand them. Comprehension (and ability to act on the information) is the most important thing in this context, yet there is no way to measure it except to try it out with your audience.

Do you know of any other way to measure or predict comprehension?

The relationship between fluency and comprehension is complex. In order to comprehend a text, one has to read fluently to reduce the cognitive demand of reading. So for example, if you read very, very slowly, by the time you get to the end of the sentence, you will have forgotten what the first few words were, to help you make sense of the sentence, because you are so focused on figuring out each word, you cannot concentrate on meanings and putting the meanings together to make sense of the sentence.

One definition of fluency is speed, accuracy and prosody. To read with prosody, one has to be understanding the words as one is reading.

However, many ESL students who are good at decoding, can read very, very quickly and have no clue what they are reading!

So, it is very complicated!

Daphne

John,

Thank you for your very thoughtful post. You have raised lots of points, and I will try to address each of them.

1. Readability indexes are only ONE source of information. One major problem with them, and especially for our adult learners is that they do not take into account background knowledge. So for example, an adult man who has diabetes and reads at the second grade level, may be able to read a text with words such as diabetes and insulin, as long as the words surrounding the text are easy enough.

2. Coh-Metrix is the closest readability measure we have that takes into account many more factors about words, sentences, meaning, etc than any other readability measure. However, science has yet to uncover the perfect readability tool, and may never be able to-time will tell...

3. I am curious about your objection to "real-life stories" and "made-up stories" We had queried different adult learners/teachers about the use of the word fiction and nonfiction, and it was thought that not all learners (especially reading at the lower levels) would know what those words mean. They were not offended by the terms real life and made-up. But I am curious what you think, and what you suggest we use instead.

4. You are right. We haven't found the most effective way to teach struggling adult readers. That is why research is so critical in our field.

Daphne

Thank you for the replies/perspectives; and it's really great, Daphne, that you're making the development of this project so accessible.

My concern with readability indexes is that they allow us to assume what might be called the "Goldilocks" idea of teaching texts--not too hard, not too easy. However, as Shanahan argues, the evidence for the efficacy of this approach simply doesn't exist.  However, evidence does exist to the contrary: that there is more to gain by teaching more difficult texts, which is also the working idea behind the common core (another evidence-based reading model called Reading Apprenticeship, which I find very exciting, puts this into practice with documented results by independent evaluations).  

In short, my "experience-based" belief is that, when it comes to adults, reading levels do not equate to intellectual levels.  In other words, even if an adult reads at a 5th grade level, she does not have a fifth grader's intellect.  I think we all know this to be true, yet we have not been able to theorize this; consequently, the field of adult literacy seems to fall back on more or less traditional k-12 pedagogy.  My hope is that we can develop principles, evidence-based or otherwise, for teaching adults as such.  

Thanks again,

John

I neglected to address the "real life/made up" categories.  Since the categories of fiction/nonfiction are so widely used, perhaps something like "Fiction (stories that are made-up/invented)" would be a better way to bridge students to knowledge of conventional literary categories.

John

I think that you may have a good idea-we should call the topics:

Fiction (made up stories) Nonfiction (real life stories). This way, we are using the real terms, but providing a definition for those who need it. I will run this idea by my colleagues. Thanks for the suggestion!

Daphne

John,

Just wanted to share that we have changed the real life/made up categories to Non-fiction (real life stories) and Fiction (made up stories) . Thanks for the suggestion!

Daphne

 

I like how you wrote: "making the development of this project so accessible". One thing that I probably should have stressed in my initial announcement is that the repository is not a final product-we are constantly in the process of making changes. We have already made changes based on people's suggestions, and will continue to do so. So thanks everyone!

I agree, reading levels do not equate to levels of intelligence. So, as you say, an adult reading at a 5th grade level does not indicate anything about IQ level.

Daphne

 

Hi,

I have found this entire discussion on readability so interesting.  I used a very basic tool for readability within Microsoft Word.  Possibly, it may be of interest to some of our readers;  To use the tool within a Word document or portion of Word document, go to "Tools," then "Options," then "Spelling & Grammar." then "Grammar," then lastly click on "Show Readability Statistics."  It will give you the readability statistics for everything you have highlighted.

Rochelle Kenyon, SME

 

Daphne,

Thanks so much for sharing your knowledge of these intricacies! I have a question that is frequently discussed in health literacy plain language work: 

Say you have a text that is written for people who have diabetes, and you want to make it suitable for a 5th grade level reader. Presumably, the target readers are quite familiar with the word "diabetes", even though it is a four-syllable word, and one that in most circles would be considered "medical jargon". When you apply a readability formula to the text, the frequent use of "diabetes" will increase the grade level rating. Some people who are evaluating and re-writing these texts will take out all the incidences of the word "diabetes" and use the resulting grade level measure. They do this because they figure that the readers will know the word well enough. But other plain language reviewers say that the word "diabetes" will still add considerably to the cognitive load of the text, even for readers who are familiar with it. This is because people still will naturally sound it out as they read, and encountering it several times in the text will diminish their overall understanding. What do you think about this?

Thanks!

Julie

Julie,

You wrote that even for readers who are familiar with a word, they "still will naturally sound it out as they read, and encountering it several times in the text will diminish their overall understanding."

I would argue that if someone is sounding out a word, then they do not automatically know the word by sight, it is still a relatively new word. So it might not be a completely unknown word, and if they begin to sound it out, they may be able to remember it, and not have to sound out the whole word, but it is still not part of their automatic sight word reading lexicon.

So I think you raise a very good point. Just because someone, for example, has diabetes, and therefore, we think they have been exposed to the written word enough times, that it should be an automatic sight word for them-it may not be. This has great implications for writers who write specific targeted documents for a specific group. One cannot assume that everyone can read individual words in the same automatic fashion, even if they share the same experience/background.

Daphne

Thanks, Daphne,

I didn't write my question very clearly! :) What I meant was that the argument I often hear assumes that lower level readers will automatically sound out "diabetes" in a text,even though they are familiar with the word, but I was hoping to get your opinion on that.

When new readers learn sight words, even if they are multi-syllabic, can they then read through them fluently in a text?

Thanks!

Julie

I think it all depends on your definition of familiar. If by familiar you mean something like being exposed to the word, then that does not at all mean that the learner "knows" the word-all it means is that there is something familiar about it. So, until the learner knows the word, and can read it w/o thinking-the way you are reading my words right now, there is a good possibility that the learner will need to sound it out or guess.

Once a word becomes automatic (doesn't matter how it was taught), by definition a person reads the word fluently in the text (as long as the word makes logical sense in terms of the depth and breadth of vocabulary that the reader has-for example, if I only know that the word "orange" means a color, and all of a sudden I read a sentence that says: "I love eating an orange for breakfast" The person may read the word orange quickly, but then pause because one does not eat a color).

All words that we know how to read become sight words to us. Think of how you would read the words that I am writing, if they are presented to you as a list of words, out of context-you would read them by sight.

So to answer your questions specifically:

If a person is only "familiar" with the word diabetes, but doesn't "know" the word in order to read it automatically/effortlessly, he/she will need to do some kind of strategy-such as sounding it out or guessing.  If the person knows a sight word automatically, then it does not matter how many syllable it has. Think about how you read the word "illustration" It is multisyllabic, and yet you read it by sight.

Daphne

While all the discussion of lexical levels and measures is interesting, what really struck me when I went to the CSAL site, chose and read an article and worked through the follow-on questions, was how well-organized and easy to understand the site seems to be from a low-literacy adult reader's perspective. My own ABE class is not in session right now (our school closes each summer), but I am meeting with a few literacy students from time to time at our local library, and I plan to have them try out the site and give me feedback (and email me their "scores") so I can see if this would work well in my classroom next fall.

I think the accuracy of the lexical scoring algorithm is of secondary importance compared to whether readers can find content that interests them at a level they can understand--and "easier" "medium" "harder" is about the amount of distinction that makes sense to most people in my classes. That's how I label "differentiated" exercises in class and people are usually able to self-select well once they get a sense of what lies under those headings. Also the content buckets are almost exactly the topics that my readers always come up with when we start the year by brainstorming about what topics they want to learn more about. Also, to weigh in on the fiction versus "made-up" labeling debate: most students in my classes get the terms fiction and non-fiction reversed. Using the common-sense labels that CSAL already has and perhaps putting the fiction/non-fiction terms in parentheses after them would be a good way to start bridging to the academic vocabulary--but it would be a mistake, in my opinion, to abandon the current labeling altogether since it is how the readers themselves think of these categories. As you can tell, I'm all for removing the teacher as a necessary intermediary between adult learners and adult learning materials. I want my students to be able to pick their own instructional materials so they can select for maximum interest and "best fit" in terms of level.  My role is to coach them so that they get better at doing that, especially online.

Another interesting free online resource that some of you may already be familiar with is NEWSELA which offers online news articles rewritten to several lexical levels. It has the advantage of offering extremely current articles (which I noticed was not true of the one piece I read from this CSAL site). Some articles include "quizzes," but not all. And sometimes even the easiest lexical level of an article is still too hard for very low level readers in my class, whereas the CSAL article I read from the Easier category was  at a lower level that would probably work for nearly any of my readers. "Joining" the site is a little bit of an obstacle, and I appreciate that CSAL doesn't require that, while still giving learners a couple ways to notify someone of what they've done.

Many thanks to Daphne for sharing this resource with us.  (It reminds me that I really should read more of the LINCS discussions--to find useful gems like this.)

Thanks Wendy! And I agree with you. I think that the major impact of this repository is not the actual readability levels (since that is still an evolving science) but rather that there is one location where all these different types of texts can be found.

And you are also correct-NEWSELA is a great resource for current articles. You hit upon one of the issues we are struggling with right now. Do we exclude articles that are not current? And if yes, how is it that to be determined? We are thinking about putting some kind of disclaimer about current vs. not current on our teacher page. We are thinking that MAYBE we should also put something on our learner page, but we do not want our learner page to become too text heavy...

Daphne

Hi Daphne,

Thinking back to when I was a content manager for a technical support database, I remember we had an "expiration date" in the metadata for each technical note. It was the date by which material needed to be re-reviewed/revised or pulled completely. It entailed additional review and revision cycles for these "potentially expiring content" reviews.

If that's too much to expect from this type of repository, you might consider just not stocking perishables--in other words trying to focus on content that doesn't make reference to things that will date it quickly.

I think the disclaimer approach is perhaps the least useful, since it may only discourage teachers from using the content. And it's really the learners/readers who need the warning, but I agree with keeping the learner page as un-text heavy as possible. I guess at least a disclaimer to teachers might make us think of terming "dated" articles into a teachable moment where we encourage learners to see what's happened since the piece was written by doing a little online research...that's what I do when I hit this type of issue in class.

 

Wendy,

Your note hits upon a very important topic-how do we know whether a website is current and accurate? This is an issue that impacts all of us, of all ages, whether we are new readers, expert readers, or developing readers. I hope that adult literacy teachers consider this a topic worth exploring with learners. As far as our repository is concerned, we have decided that our purpose is to share what we have found on the Internet that is geared towards adults who are not expert readers. We agree with you that having a disclaimer is probably not useful. Unfortunately, we do not have the expertise or money to seek the expertise to determine whether an article is dated. I like your suggestion of teachers using this as  "a teachable moment where we encourage learners to see what's happened since the piece was written by doing a little online research."

Daphne

Yes, this is an important topic--how to know if a website is current, accurate and unbiased. Since so many people use the internet to search for health information (according to Pew, NAAL, etc.), and since health information has a high need for being accurate and current, there have been some good resources developed to address this. Check them out!

Who Can You Trust? Health Information and the Internet: Curriculum Sourcebook

You can also read about the formative research process and the results of the pilot testing here:

Health Information Literacy Outreach: Improving Health Literacy and Access to Reliable Health Information Online in Rural Oxford County Maine

There is also a lesson about evaluating websites in an issue of The Change Agent.  Scroll down to issue #28 called Health: The Big Picture.

The Change Agent is a magazine written for and by adult education students, dedicated to issues of social justice. I would highly recommend it as another source of readings suitable for emerging adult readers. This is developed by the New England Literacy Resource Center (NELRC) at World Education. They can no longer afford to give free access to the back issues, but teachers can subscribe for $20 per year.