A new Castles and Coltheart reading test available on-line

Authors: Anne Castles and Max Coltheart, Macquarie Centre for Cognitive Science, Macquarie University, Australia.

Two key processes that children need to acquire when learning to read are sounding-out ability and whole word recognition ability.

Sounding-out ability involves converting printed letters into their corresponding sounds, and is best assessed by measuring a child’s accuracy in reading aloud nonwords (nonsense words), such as “gop”, since these cannot be read by any other means.

Whole word recognition ability involves accessing stored knowledge about familiar written words, and is best assessed by measuring accuracy in reading aloud irregular words, such as “yacht”, since these cannot be read correctly via sounding-out rules.

In 1993, Castles and Coltheart developed a test designed to directly assess how well children can use the sounding-out and the whole word recognition procedures.

The test consisted of 30 irregular words and 30 nonwords for assessing each of the key reading processes.

Thirty regular words, like “cat” were also included: These kinds of words can be read accurately by either sounding-out or whole word recognition and so provide a measure of the combined functioning of the two processes.

There were, however, two major limitations of this original version of the Castles and Coltheart test.

Firstly, the test did not have a stopping rule. As a result, even a child who could successfully read aloud only a few of the simplest items on the test had to be presented with all 90 items. This was both time-consuming for the tester and potentially stressful for a child who could only read a few items.

Secondly, for older age groups, the test was too easy, particularly for the regular words and the nonwords.

We have now developed a modified version of the test to address these limitations. The new test contains an expanded set of items, with 40 each of regular words, irregular words and nonwords, rather than the original 30 items of each type. The new items extend the upper-end of the difficulty range of the test, making it more suitable for older children.

The test also incorporates a stopping-rule, which  makes administration of the test less time-consuming, and removes the stress on children who can only read a few items.
The test has been normed on over 1000 Australian children.

The test is free of charge and is available to teachers, professionals and researchers at our new Macquarie On-line Test Interface ­(MOTIf: http://www.motif.org.au).

The test can be administered on-line to one or more children and the results stored in an individual and secure test page. Alternatively, hard copies of the test materials can be downloaded as pdfs from MOTIf in order to administer the tests off-line.

We hope that this modified test and the new facility for administering it will assist professionals in assessing key components of reading in children.