What Doesn't Belong Here? On Teaching Nonsense Words

  • Assessment adolescents
  • 07 May, 2016
  • 16 Comments

  Obviously you shouldn’t wear an especially short skirt to work, though it might be fine for a night of bar hopping. It would just be out of place. Lil Wayne can do rap, but he’d definitely be out of place at Gospel a Convention, sort of like a love affair with a happy ending in a Taylor Swift lyric.

  So what’s out of place in reading education?

My nominee is the act of teaching kids to read nonsense words. Don’t do it. It don’t belong (it may even be worse than orange and green).

      Why, you might ask, would anyone teach nonsense words? I attribute this all-too-common error to a serious misunderstanding of tests and testing.
Many years ago researchers were interested in determining how well kids could decode. They decided upon lists of words that were graded in difficulty. The more words the students could read accurately, the better we assumed his/her decoding must be. 
But, then they started to think: It’s possible for kids to memorize a bunch of words. In fact, with certain high frequency words we tell kids to memorize them. If I flash the word “of” to a student and he/she reads it correctly, that might not be due to better phonics skills, but just because Johnny had that one drilled into long-term memory.    
That means with word tests we can never be sure of how well kids can decode.            
The solution: nonsense words tests. If we give kids lists of nonse words, that is combinations of letters that fit English spelling patterns, but that aren’t really words, then if students can read them they must have decoding skills, because no one in their right mind would teach these made up letter combinations to children.
Enter tests like DIBELS decoding measure. Tests designed to help determine quickly who needs more help with decoding. These aren’t tests aimed at evaluating programs or teachers; they are diagnostic. 
      These tests work pretty well, too. Studies show a high correlation between performance on nonsense words and real words, and some of the time the nonsense word scores are more closely related with reading achievement than the word test scores!
         But many schools are now using these to make judgments about teachers.
      And, the teachers’ reaction has been to teach nonsense words to the kids. Not just any nonsense words either; the specific nonsense words that show up on DIBELS. That means these teachers are making the test worthless. If kids are memorizing pronunciations for those nonsense words, then the tests no longer can tell how well the kids can decode. 
      We can do better. Please do not use these kinds of tests to make judgments about teachers, it just encourages foolish responses on their parts. And, please do not teach these nonsense words to the kids. It is harmful to kids. It definitely doesn’t belong here.

Comments

See what others have to say about this topic.

EdEd
Apr 06, 2017 06:07 PM

Great post Dr. Shanahan - hope everyone reads this! 5/8/16

Sarah Paul
Apr 06, 2017 06:08 PM

I disagree with this. Teachers are not just using nonsense words to help kids pass tests. what about the one in five kids with dyslexia? Research shows that these kids learn best through systematic phonics instruction. Reading nonsense words provides addition practice for these kids to help them master the sounds and the improve decoding skills. Yes, many many words are not decodable. But many are and this is an important skill. Saying that nonsense words have no place in instruction is a bold blanket statement. I agree we should not be basing our instruction on the tests. However, in my experience, using nonsense words has helped my students become better decoders which has translated to improvement with overall reading. I've had students who are sight readers and memorize words, but have very poor phonics skills. This catches up with them

I'm not speaking about the DIBELS test specifically, but I do know that assessing nonsense words can help teachers identify students at-risk for reading disabilities. I've seen this first hand many times. A student may be able to memorize certain words and seem like they are totally on track but you give them a nonsense word and you realize they can't decode at all. this is a red flag! Teachers need to know this. Early intervention makes a huge difference and assessing nonsense words can be a way to identify who needs that early intervention. With all of this said, I do agree that teachers should not be punished based on any test score. In addition, adding nonsense words to instruction should not be a way to do better on assessments. It's all about the kids and finding ways to help them. My kindergarten and first grade students love the literacy Acticities I provide, including some nonsense word practice. There are ways to make it engaging and they feel such confidence when they become great decoders.

I think with so many followers you should be careful what you are preaching. I respect you and love reading your posts. This one just didn't sit well with me.
Respectfully,
Sarah 5/8/16

Timothy Shanahan
Apr 06, 2017 06:08 PM

I'm a big believer in the teaching of decoding--for dyslexics and everyone else learning an alphabetic language. However, English is the biggest language on earth (in terms of numbers of words). We have no need to make up words to try to come up with instructional examples--and, of course, if there are no examples or few examples of a particular pattern, it isn't worth teaching.

To read an alphabetic language students need to use the letters to arrive at an approximation of a pronunciation of a word that the student is likely to have in his or her language. Teaching kids something else (that is to sound out words, paying no attention to whether they are right or not--because if they are made up words how can you know?) makes no sense. Rather than pretending to teach kids phonics--that is, teaching them to sound out nonsense words, I'd suggest that we teach kids to sound out words so they can get good at reading.

Sarah, even if you are right, it is peculiar that the nonsense word practice these teachers feel that they need to provide focuses all on nonsense words from the tests. What they are doing is making it look like the kids are doing well in decoding, so that they don't have to give the students any extra practice--that's harmful to the kids, even if makes the teacher feel better about herself or allows her to fool the principal into thinking her class is doing well on this skill.

tim 5/8/16

X
Apr 06, 2017 06:09 PM

'm not a big fan of nonsense words. I think they confuse young readers as their vocabulary is being developed. However, I have a question. When it comes to new young readers, can the pronunciation of "nonsense CVC words" really be memorized? In this case, most of those words are good words to apply decoding skills.

5/9/16

Timothy Shanahan
Apr 06, 2017 06:09 PM

X=

They definitely can be memorized. Memory for words should be altered by reading instruction and reading experience, but early on kids have no recourse but to memorize if they want to remember words. Over time, if they become readers, such "memorizing" gets easier and easier as those neural paths are established and the kids learn how the orthography and phonology work.

You can definitely apply decoding skills to such strings of letters, but beyond trying to help kids to do better on the tests, what is the benefit of practicing reading dat, jat, kat, wat, instead of cat, fat, rat, hat? In fact, that kind of practice might even be problematic because while wat isn't a real word, water is and the wat in that is pronounced differently... the same with dat and data.

Testing kids briefly with nonsense words can give a bit more information than testing with real words, but teaching with nonsense words adds no benefit over teaching with real words, and may be misleading to kids in terms of real pronunciations and spellings.

Mrs. McClellan
Apr 06, 2017 06:10 PM

Non-sense words are super difficult for ELL students and I think the post is stating that teaching nonsense words to ensure a student performs well on a assessment is wrong. I enjoyed the perspective because these words are super confusing to my ELL students. 5/9/16

Timothy Shanahan
Apr 06, 2017 06:10 PM

Mrs. McClellan--
You raise an important point about nonsense words and nonsense word tests. It is extremely important when giving those tests to ELL students that you explain to them clearly that these are not real words and that they will not do well if they try to come up with words that they think they know in English. It is important that they use the sounds alone. Of course, that makes such tests somewhat different from what we ask these kids to do when reading words in English, but it still protects against misevaluating their decoding skills because of memorized words.

thanks.

tim 5/9/16

Idlormand
Apr 06, 2017 06:11 PM

Tim, I am a former colleague of you from McGraw-Hill and have been following this thread with a lot of interest. I would go even further as to question the use, or rather misuse of the Dibbles test entirely. When looking at a text, our brains automatically are searching for meaning, or at least they should. Using a test like the Dibbles is a futile exercise to diagnose problems with decoding on the part of a student. It is absolutely counter productive when it comes to ELLs but also to the reader in general. A well trained teacher should be able to identify a lack of decoding skills in a student without having to resort to nonsense "words".

5/13/16

Timothy Shanahan
Apr 06, 2017 06:12 PM

Thanks, Idlormand--
DIBELS and other similar informations are different from reading, but they can be useful (when used appropriately) to identify who is not on track in a specific skill like decoding. They don't tell you more than that, but that is useful.

thanks.

tim 5/13/16

Anonymous
Apr 06, 2017 06:12 PM

Yes, nonsense words are very useful if used as intended and that intent is NOT to master reading nonsense words for the sake of doing well on the assessment. The assessment should reflect how well you're teaching your students to decode and guide your instruction. There are plenty of real words that teachers can use to teach students how to decode and blend. Using nonsense words during instruction is nonsense!! Please continue to get this message out there. Thank you!!!

5/16/16

Debbie Hepplewhite
Apr 06, 2017 06:12 PM

Hi Tim. Thank you for your post which is very topical in England.

I have added a comment re your post via the International Foundation for Effective Reading Instruction here:


http://www.iferi.org/iferi_forum/viewtopic.php?f=2&t=589

Please do feel free, of course, to contribute any of your most welcome thoughts re literacy direct to the IFERI message forum.

Warmest regards,

Debbie

7/23/16

Dick Schultz
Apr 06, 2017 06:13 PM

In the way of "mixed metaphors," it seems to me we have "mixed advice" going on here, Tim. US EdLand is currently doing a lot of stupid stuff. Teaching kids pseudo words so that they can pass stupid tests is stupid. Using stupid tests to evaluate teachers escalates and compounds the stupidity.

At the same time, the "meaning/nonsense" distinction of a word doesn't lie in the word; it resides in the conventions that have been attached to the spoken word. All neologisms were "nonsense" before they become "logisms." (I'm not sure "logism" is a word, but if it wasn't before, it is now--for those reading the comment at least. I wouldn't count on the word catching on, but stranger things have happened.)

Texts are a means of coding and recording speech, and reading is a means of re-creating the communication. The communication can range from gibberish to profound, but that applies equally to spoken communication as well as to written communication.

While it's "not good for kids" to teach them nonsense words, it's also "not good" to teach them that all words have meaning. The mal-instruction encourages a child to convert unfamiliar words into a word that is familiar to the child, usually by "skipping" the middle of the word. Children should to distinguish between familiar and unfamiliar words so that they slow down their deciphering. They should also learn to distinguish between a word that has meaning for them and those words that don't. Children have learned how to do these complicated things in spoken communication, so it's easy to build the capability into reading instruction. But it entails more than not teaching nonsense words.

7/23/16

MaryCaren D'Anniballe
Apr 03, 2018 02:24 PM

I'm looking for research against using nonsense words as an early literacy measure for kindergarten and first grade students. My district will be using FastBridge next year to assess our students, and many are in favor of using the nonsense word assessment. As a literacy coach with over 20 years of experience in teaching reading, it doesn't make sense to me to use such a measure. Again, I'm looking for research to support not using this. Thank you

Dawn
Jul 25, 2021 10:10 PM

How do you feel about the assessments that include 'writing' nonsense/ non-words?

Liz Lightner
Aug 31, 2018 02:35 AM

Glad to come across this thread of thought. It came to my attention at a first grade meeting with a teacher new to our team that another school was providing all students with monthly grids of nonsense words like the DIBELS test to practice for homework. In addition, students in intervention had parent helpers quizzing them weekly in 1 minute timed sessions in addition to the progress monitoring taking place. I was dismayed and didn't even know where to begin to explain why this was not only inappropriate for testing validity but sending the wrong message to beginning readers as to the purpose of reading: to make meaning from text.

For the first time since my very large district has been using DIBELS as a screening tool, we will be using the data K-5 to report progress to our school board and community. I fear the use of drilling students on nonsense words may become more wide spread. Educating teachers on this nonsense will hopefully take place!

Anita Kim Venegas
Aug 21, 2020 04:03 PM

Everyone is asking for the research, me included. Our school is about 20 percent with English learners and many students with reading disabilities or difficulties. I think testing them on a phonics assessment that includes nonsense words is WRONG. For those learning to understand the language, I think they are trying to make meaning first, and there is no meaning in nonsense words. I see it over and over again-English learners can read 'real' words, but they cannot read nonsense words. It causes them to not pass the assessment. I've created an online phonics assessment with nonsense words, but I'm VERY tempted to change out the nonsense words in order to use MORE of the digraphs and blends in real words.
Hoping to hear from you.

What Are your thoughts?

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

Comment *
Name*
Email*
Website
Comments

What Doesn't Belong Here? On Teaching Nonsense Words

16 comments

One of the world’s premier literacy educators.

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