Test gauges children’s verbal skills UM teacher’s program screens infants, toddlers for language progress

Measuring preschool children’s language development soon could be almost as convenient as glancing at a growth chart to gauge their physical development, a University of Maine professor said this week.

Alan Cobo-Lewis, an associate professor of psychology, has helped develop a computerized test that measures infants’ and toddlers’ language skills in just a few minutes. The program asks parents questions about the words their offspring know and understand.

The tool could be on the market late next year for professionals including pediatricians, speech therapists, and psychologists to screen children ages 8 months to 36 months for language delays, and for determining how much they understand and how well they speak.

The test also could track development in children with language problems.

“Pediatricians typically plot a child’s growth curve to see how they’re growing over time. We want to have a fast tool so pediatricians could plot how well a child’s language is growing over time,” said Cobo-Lewis, who worked on the project with Trefoil Corp., a software development company in Orono.

Other tests to measure children’s abilities are available, but either are not as reliable or take too long to administer, said Cobo-Lewis.

Since his test is shorter, he hopes it will be used more and help identify children with language problems sooner. A lag in language skills could indicate a developmental problem that might affect a child socially and academically.

Cobo-Lewis’ new computerized program is based on the MacArthur Communicative Development Inventories, a paper and pencil test that estimates how a child’s language skills compare to what’s expected at a given age.

While the hour-long CDI requires parents to review 493 words for infants 8 to 16 months, and 797 words for toddlers 15 to 30 months, the computerized version asks just 22 questions and can be completed in five minutes.

Researchers validated the new computerized test by having parents take both the long and short versions and then comparing results. So far, 85 area parents have been tested.

Since the goal is to devise a national assessment test, parents in Massachusetts also are being used as subjects to provide “an ethnically diverse population,” Cobo-Lewis said.

Preliminary results show that the speedier version of the test “does very well,” he said.

The new version is quicker because it focuses on questions that are relevant to the particular child, Cobo-Lewis said. It calculates language level on the basis of each successive word, using a standardized database of test results from about 1,600 children.

For example, parents of a girl around the age of 1 might be asked if their child knows the word “shoes.” If she doesn’t understand the word, or if she understands but doesn’t say it, the program might ask about the word “ball,” which girls usually learn slightly earlier.

If the girl does say “shoes,” the program might ask about the word “brush,” which girls usually learn slightly later.

The program knows that boys tend to lag slightly behind girls in language development, so the first question to parents of a boy the same age might involve the word “ball,” an easier word.

If their son understands “ball,” the program might follow up by asking about the word “shoes.” If he doesn’t understand “ball,” the next question might ask something simpler such as whether the child can reach out and offer a toy he’s holding.

Although there are other tools for assessing language development, they don’t offer the same detailed information as the CDI, according to the professor.

The key to the CDI’s reliability is that it isn’t open-ended, but instead asks about specific words, he said.

Results so far show that girls typically lead boys in language understanding by about a month and in using specific words by about 21/2 months.

But while girls tend to understand the words “doll” and “dress” about three months earlier than boys, boys tend to understand “vroom” and “truck” about half a month earlier than girls, Cobo-Lewis said.

Also, boys tend to understand “vacuum” and “broom” before girls, and engage in sweeping or vacuuming motions earlier.

Now in the final phase of research, Cobo-Lewis is recruiting more parents, including those with children suspected of having language delays.

Some parents liked participating because it helped them reflect on their child’s verbal skills, he said.

Others “tended to be incredulous – how can the test work so quickly?” he said they asked him.

Parents interested in participating in the study may contact research technician Bonnie Blagojevic at 581-3849 or by e-mail, earlylanguageproject@umit.maine.edu. The hour-and-a-half sessions take place at the university. Volunteers will receive $50 compensation. Information about the project is available at www.umaine.edu/language.