Reading Tests that 'Misread' Some Children

November 19, 2007

Screening tests widely used to identify children with reading problems are being misapplied, landing students in the wrong instructional level and delaying treatment for their true difficulties, says new research from the University of Maryland and National-Louis University.

The researchers find that oral reading tests fail to distinguish between children who can't understand words on a page and those who have language problems that make it difficult to prove their reading competence verbally. Children with these so-called "word-finding" difficulties can't manage to say out loud what they read on the page.

The study, published in the November 2007 issue of Reading Psychology, recommends silent reading tests and limited use of oral ones.

The researchers estimate that as many as ten percent of all children may have these speech language problems. Roughly one-in-five children have some kind of learning difficulty, and nearly half of these have the word-finding problem.

"The look on these children's faces captures the problem in the most compelling way," says Diane German, the principal researcher, an expert in disorders of word-finding and a special education professor at National-Louis University in Chicago, Illinois. "They really struggle when they have to read a simple word like 'nest' out loud. Some grimace, others look stuck. Some just blurt out an answer that's almost always wrong. Yet when asked to point to the same word on a page, they almost always get it right. Clearly they've got a problem and need help, but it's not that they lack reading skills."

One child in the study, previously diagnosed with these "word-finding" difficulties, couldn't say "cocoon" as he tried to read a story aloud. When he got to the word, he stumbled and added, "You know, it is that brown thing hanging in the tree."

"Clearly, this child had managed to 'read' the word to himself and comprehend it, or he could never have come up with that kind of description," explains psychologist Rochelle Newman, co-author of the study and a University of Maryland professor of hearing and speech sciences. "He just couldn't retrieve the sound pattern of the word."

While word-finding difficulties are relatively common, German and Newman say it's unclear how many of these children end up mislabeled by schools. All 25 of the students in the experiment diagnosed with word-finding difficulties had been put in remedial reading classes at school - unnecessarily, based on the study results. These are probably not isolated cases, the researchers say.

"This potential for misdiagnosis and under-estimating of children's reading abilities is disconcerting," says German. "Reading teachers and speech and language pathologists need to be more cautious with oral reading screening tests. Often they are surprised when they make this paradigm shift from oral to silent reading assessment. They had truly overlooked how much the children had learned. It's a big wake-up call."

The study recommends the use of silent tasks to determine the actual reading ability of children with speech or language difficulties. For example, to check reading decoding skills, children can be asked to "point to the word" or "find the word." Multiple-choice questions can be used to check silent reading comprehension.

Results and Method

The researchers used an experimental design, testing one group of 15 'typical learning' first-graders and another group of 25 second- and third-graders. The older children had word-finding difficulties and were getting remedial reading instruction. All the subjects were given an experimental reading assessment previously developed and tested by the researchers ("Test of Oral and Silent Reading Recognition"), as well as other language-related tasks.

"We tested the first-graders with the idea that they were rookies at reading," says Newman. "If the second- and third-graders, the ones with the word-finding problems, hadn't learned to read, as their schools believed, then they would perform on the silent tests only about as well as the rookies. But they didn't. The older children surpassed the rookies on these silent tests. They had learned, but just couldn't show it."

The older children scored 98 percent in the silent test compared to 58 percent for the first-graders. The results also verified that children with word-finding difficulties had a much harder time with oral tests than silent ones. Accuracy on oral reading tests for these older children ranged from 57 to 62 percent. But on silent tests using the same words, their reading accuracy was significantly better, ranging from 85 to 88 percent.

Source: University of Maryland