Spelling is a speeded test of the ability to spell regular and irregular words (see Section 1.1.2 for explanation of speeded tests). Each form comprises 30 regular words and 20 irregular words that are presented in order of difficulty (which was established by previous trials with large numbers of students). In each item the target word is spoken by the computer both in isolation and in the context of a sentence. An illustration associated with the sentence appears on the screen. The task is to type in the target word as quickly as possible. Note that students can begin typing in the target word as soon as they hear it and do not have to wait for the contextual sentence to be spoken.
The time allowed for each item is a function of the length of the word (3–14 letters), with a minimum of 9 seconds and a maximum of 31 seconds. An audible prompt is given 3 seconds before the allowed time is up. However, the time allowed does not increase as a function of the difficulty of the words. There are two reasons for this. The first reason is that if the time allowed did increase as a function of the difficulty of the words, this would reduce the effectiveness of the test to identify poor spellers. The second reason is that, at this age, spelling is mostly automatised – that is, through practice the student has learned how to spell the word without much conscious effort – and hence the time taken to produce a spelling is largely due to how many letters have to be written or typed. Hence the test is efficient in revealing the lack of automaticity in poor spellers (see Section 1.1.2 for a discussion of the importance of automaticity in skilled literacy).
The spelling test is adaptive insofar as the entry point is determined by performance on a number of ‘probe’ items of increasing difficulty given at the start of the test. When a student fails a probe item (or all the probes have been successfully answered) the program jumps to the appropriate place in the test. This preserves the sensitivity of the test for assessing poor spellers whilst avoiding boredom of more able students, who would otherwise find it very tedious and demotivating to have to spell lots of very easy words. Items that are jumped in this way are credited as having been passed correctly by the student. The test is automatically discontinued when four out of the last six items attempted have been answered incorrectly. Since all the items are in order of difficulty the score obtained still accurately reflects the student’s spelling ability even though they have not necessarily attempted all the items.
Results are provided for overall words (see Section 3.1.1), and for regular words and irregular words (see Section 3.1.4). A breakdown of all responses in this test is also provided for assessors who wish to use this information diagnostically. In secondary school students, spelling is often a more accurate measure of dyslexia than reading ability, and standard scores below 85 may be another pointer to that diagnosis (for an example, see the case study presented in Section 4.1).