University of Arizona, Emeritus
Last year according to the official DIBELS web site ( http://dibels.uoregon.edu/ ) over 1.7 million children from kindergarten through third grade were processed through the DIBELS computers at the University of Oregon. No doubt the figure is over 2 million for this year as many states go beyond the federal pressure to DIBEL Reading First children and require all K-3 children in the state to be DIBELed
Parents of five year olds are hearing a new answer to the age old question, “What did you do in school today? “I got DIBELed.” Within a few days of entering kindergarten, hundreds of thousands of five year olds are given their first opportunity to taste failure in their ability to say the names of letters in three seconds, say the sound that a picture name begins with in three seconds, and sound out three letter words in three seconds. And if they can’t get enough letters named, initial sounds made, or words sounded in one minute in each DIBELS sub-test then they have failed to bench mark and are thus in need of intensive instruction. From then on they will be DIBELed three times during each year through third grade and sometimes beyond. By mid-year in kindergarten the children also must sound out a page of nonsense syllables. Currently a subtest has been added to DIBELS in kindergarten that tests their ability to “use words orally.” Between benchmark testings, there are monthly progress tests. Those who fail to benchmark are tested every two weeks.
A Battery of one-minute tests
So what is DIBELS that it should be given such importance? DIBELS is nothing more than a set of one minute tests of reading “skills.” It only takes about 10-15 minutes to administer the whole battery to each child each time. It takes less time for many kindergarteners because each sub-test is discontinued if the child fails to respond to a few items in a row. Many five year olds still not adjusted to separation from their mothers are simply overwhelmed by being escorted to an unfamiliar place in the school where a stranger with a top watch rushes them through a series of tasks and stops them before they have had any chance to figure out what is happening. The stranger with the stop watch marks a score sheet while timing each response and cutting off the child at the end of one minute. One child returned to her teacher happily reporting :”You’ll be proud of me! I didn’t say a word to those strangers.” Her scores? All zeroes.
But the benchmark scores on DIBELS becomes reading in the schools where it is used. Learning to read is equated with reaching bench mark scores on each successive sub test. The DIBELS manual says “As stepping stones to literacy development, it means that performance on one of the DIBELS measures is predictive of performance on the next appropriate DIBELS measure(s).” So teachers are led to believe that bench-marking each test is pre-requisite to subsequent tests and that becoming a successful reader requires success on each test.
In an earlier era of behavioral objectives, schools were urged to use a test, teach, test model to determine if objectives were “mastered.” DIBELS goes one step further. It is a test, test, test program. Children are tested with the one-minute tests. Then they are drilled on tasks the tests require which are so specific - naming letters, saying initial sounds, sounding out, etc. that the practice is for test performance. In DIBELS procedures children are not taught to the test. They are taught the test. And with so much at stake it is not surprising that often they are actually rehearsed on the specific items of the test.
So DIBELS takes over the lives of primary children and too often determines their school future. It becomes the curriculum. In a short half-day kindergarten five year olds will get little more than DIBELing and preparation for DIBELing in school. First through third graders get two forty-five minute drills to get them ready to be DIBELed. And those who don’t benchmark will be pulled out for extra drill. Ironically, there’s little time for reading in the DIBELS classroom and even less time for writing.
And all over America children are being retained in Kindergarten or first grade on the basis of their DIBELS scores. Headstart teachers of three and four year olds are being pressured to get their tiny pupils ready for DIBELS.
Some teachers have a bulletin board with nonsense words for the children to practice reading nonsense. And there are stories of children practicing for DIBELS while waiting in line to use the toilet.
The premises of DIBELS
DIBELS is committed to testing the “Big Ideas” as IDEA, its parent, has stated them and as derived from the National Reading Panel report. But it’s focus is on only three of these big ideas: Phonological Awareness, Alphabetic Principle, and Fluency with Connected Text. DIBELS includes a test of Letter Naming although the authors indicate that it does not relate to the Big IDEAS and does not predict success in reading. They also have added to their test of fluency with connected text an optional retelling score to offer a measure of comprehension though they make clear they believe the former is a sufficient test of reading competence and comprehension. A test of Word Use Fluency is also provided apparently to correspond to the Big Idea of Vocabulary. Though it is also optional it is the one sub-test to be used from kindergarten through third grade.
The DIBELS View of Reading and Reading Development
The DIBELS authors believe that competent reading is the ability to read words rapidly and accurately and that comprehension is the result of such rapid accurate reading. They also believe that what happens in one minute tests represents well what happens in reading. Implicit in the choices the authors make of what to test in each sub-test, how the tests are sequenced, and how each component is tested is a view of reading development that involves a single universal sequence of mastering component skills that can be tested in one minute. The whole of reading, they believe is the sum of these reduced parts.
What does DIBELS actually test
With their commitment to testing what a child can do rapidly and accurately in one minute, DIBELS authors reduce each reading “Big Idea” to a task that measures something less than what the name of each implies.
Letter Naming tests the ability to rapidly and accurately identify upper and lower case letters in a single font which is not the one primary children see in print. When asked for a letter name, some children respond with a child’s name - that’s Katie’s letter. Etc. Others confuse sounds with letter names. The tester is strongly restricted in clarifying the task for such children.
Initial Sound test has the child identify the sound that the name of a picture starts with and to do it in three seconds. The tester shows a page of four pictures, names them and then says which one starts with: kuh. The answer is what looks like a bear but is called a cub by the tester. The test is supposed to test phonemic awareness- the ability to separate sounds from words- but it actually involves memory, the ability to abstract a part from a whole and the ability to suspend judgement:
“which picture starts with /duh/?” There is a picture of a frosted donut but maybe it’s a cheese bagel.
Phonemic Segmentation requires the child to sound out a one syllable word the tester has said.
If the tester judges that the child has said the sounds of the word correctly in three seconds a point is awarded for each correct sound. But what if there is a difference between the dialect of the tester and that of the child? Or what if the child has no front teeth, or mumbles, or there is noise outside the test area? How fair and consistent can testers be?
Nonsense Word is supposed to test the “alphabetic principle”. A page of two and three letter “make-believe words” is shown to the child who can either say the whole “word” or say each sound separately. One point for each sound said correctly in one minute. There are many problems with this test. One five year old, the best reader in her class, came back to the room in tears sobbing to her teacher “I couldn’t make any sense of those words.”
All of the DIBELS tests have problems of poor construction- they are badly done. But this test has numerous problems which will make its scores useless. Here’s a few: some of the ‘Nonsense words” are real words in English or Spanish (tam, el, es, wan, mum)
Words that are actually possible English spellings for real English words (in one version of this test all these potential English spellings were found ful, mik, zum, nuf, kun, fod, vep, juj, sug, ov, wam, buk, lef, luk, lof, kom, nol, rez (good word in the Southwest) poz, ol, kav, kic, kis, tek, riz, aj, vej, som, zuz) Children trying to make sense will produce the real word and be counted wrong yet they are in fact using the alphabetic principle
Words that violate English spelling rules. The test includes words ending in j, v, and z none of which occur as final letters in English spelling (final z would be spelled with an s or double z).
The issues of dialect and second language also will influence the scoring.
Oral Reading is the only test that has the child read a real passage. It provides a first person four paragraph essay for the child to read. The score is the number of words read correctly in one minute. Children learn in repeated testing and practice to say as many words as they can quickly identify and not worry about the meaning. Teachers report the passages are difficult for the intended grade.
Retelling was added according the to DIBELS manual when teachers worried that the oral reading score didn’t show comprehension. So after the oral reading the child is asked to retell what was read. The score is the number of words the child used in the retelling. In this test there is no concern for the quality of the retelling. Only the number of words the reader used is counted..
Word use The DIBELS authors have added a word use test which they want used in kindergarten through third grade. The child is asked to “use” a word. The score is the number of words used in using the word.
All of these tests have the word fluency in their names. The authors use the word to mean doing the task that is being tested fast and accurately. Whether naming letters, abstracting initial sounds, breaking words into sounds, saying nonsense words, reading oral passages, retelling the text, or using words orally the child must be fast and accurate. It’s hard to see what how fast a child can name letters has to do with making sense of print.
The Oral Reading Fluency test is the only one that requires the children to read a real connected text. But the score on that test is the number of words in a DIBELS one page passage a pupil can read accurately in one minute. Words read incorrectly are not counted so in subsequent testings and in practice for this test the children learn not to waste time trying to make sense of the passage but rather to skip any words they are unsure of and say the names of those they recognize.
In none of the DIBELS one minute tests is there any measure of the quality of the reading: No score shows comprehension.
A recent study in Alabama (in press) found that, with all this intense DIBELing, 36% of those who fail bench marks in first grade are still failing at the end of the third grade. “Student populations over-represented in the group that failed to benchmark were males, students on free and reduced lunch status, and minority students.” (SEAY, in press)
To summarize: DIBELS is a set of silly little, one minute tests so poorly thought through and constructed that they would be unlikely to pass the review of any school, district or state committee. Education Week has said that there are widespread beliefs among local and state authorities that they could not receive No Child Left Behind funds unless they adopted DIBELS. “In fact, a number of states had intended to use other assessments for screening children and gauging progress in Reading First schools. They changed their plans, they maintain, after federal officials and consultants pressured them to include DIBELS in their grant proposal as a condition for approval.” (Education Week, Sept. 7, 2005.)
No child should suffer what millions are suffering from DIBELS. And no parent or teacher should be party to DIBELing the enthusiasm for school out of children for the sake of the meaningless bench marks that are replacing learning to read in too many American schools.