The first stage of phonological awareness is to learn letter sounds. The child can then manipulate these sounds within words, for example knowing that the -at sound is different from the -a-t sound in hat and cat.
The next stage involves being able to blend smaller speech segments into syllables and words. Finally, the last step of phonological awareness involves understanding that words are made up of phonemes rather than letters. Children cannot read words by the sight before this stage has been reached.
What is PAT?
Phonological awareness can be assessed using the “Sound Segmentation” subtest on the Phonological Awareness Test (PAT). This test was devised by Bishop & Adams in 1990 and is used widely in the UK.
It is important to note that phonological skills can be assessed without using words or letter sounds, for example using noises and claps rather than spoken sounds. This task has been developed due to research finding that many children with reading difficulties have strong phonological awareness but weak decoding skills (for example see Wagner et al., 1994) and so would be expected to perform well on phonological awareness tests.
The Sound Segmentation subtest consists of four exercises involving manipulating sounds in words, non-words, and noises. The test measures the ability to break up speech into separate units (phonemes), detect differences between similar sounding units (minimal pairs) and blend speech units (syntagmatic blending).
The Sound Segmentation subtest score is the total number of points achieved in all four exercises. This test has good internal reliability (Cronbach’s alpha = .78), test-retest reliability (.76 over 6 months; Stocker & Gathercole, 2009) and construct validity.
Scores on the Sound Segmentation subtest have been found to correlate with other phonological awareness tests, reading comprehension measures, and non-word decoding ability (Bishop et al., 1990; Stocker & Gathercole, 2009). A score of 13 or below indicates that this is a child who is likely to have difficulties with phonological awareness.
Nunes et al. (2007) developed an alternative scoring system for the Sound Segmentation subtest, which they recommend for use in UK clinical practice. This method calculates the percentage of correct answers on each of the four exercises and then averages these percentages together to give a final score out of 100%. The scores out of 100% are then used to determine whether a child has difficulties with phonological awareness.
Nunes and Bryant (2007) found that the alternative scoring system was simpler and faster to use than Bishop et al.’s (1990). They also suggested that using percentages rather than total points may be preferable because it is more appropriate for children in the early stages of reading (who would be expected to make more mistakes).
There is some evidence to suggest that the alternative scoring system may result in different cut-off scores being identified for clinical use. For example, Nunes et al. (2007) found that using percentages resulted in a lower cut-off score on this test than for the total points method. They concluded that a cut-off score of 70% or below should be used for clinical use, with this resulting in the identification of 5% more children as being likely to have difficulties with phonological awareness.
The Sound Segmentation subtest is the most widely used assessment of phonological awareness skills. However, there are some limitations to this test that need to be considered when interpreting results.
First, it only measures the ability to segment sounds within words and non-words, so does not assess whether a child has difficulties with detecting differences between similar sounding units (minimal pairs) or blending units together (syntagmatic blending).
Second, it does not measure how well a child can segment words into their component letter sounds.
Finally, it does not assess whether a child can apply his phonological skills to tasks that use letters and letter sounds directly. A task that measures the ability to segment words and applies this skill to reading single words would provide a more complete assessment of phonological awareness.
The Phoneme Deletion Test (for example Bishop & Adams, 1990; Kamhi et al., 2000) is another well-established measure of phonological awareness. Like the Sound Segmentation subtest, it measures how well children can break up words into separate sounds (phonemes). It also assesses the ability to segment non-words.
For example, participants would be asked to say “heff” (the deletion of “f”) or “oorrl” (the deletion of “r”). Unlike the Sound Segmentation subtest it does not include exercises in segmenting non-words into their component sounds (e.g. “fran”) and blending units together (e.g. “stran”). There is evidence that the Phoneme Deletion Test is more sensitive than the Sound Segmentation subtest when used with children who are at risk of literacy difficulties (Bishop & Adams, 1990; Kamhi et al., 2000). A child’s score on the Phoneme Deletion Test would be expected to correspond closely with his score on the Sound Segmentation subtest.
The ability to segment letters into their component sounds is a key pre-reading skill that allows children to break words down into smaller units. A child who has phonological awareness abilities should be able to segment letters from words, but not from non-words.
The Phoneme Segmentation subtest of the Test for Phonological Awareness (for example Bishop & Moats, 2000) is a measure derived from this principle. It measures how well children can break words into their component sounds.
For example, for the item “mil” (where the child is asked to say /m/ and /i/) participants would be scored correct if they said either “mee” or “hill”. Participants are also given items where letters make up non-words, e.g. for “tere” (which contains the /t/ sound) they would be scored if they said “lur” or “here”.
Typical responses for a child who has difficulties with phonological awareness might include saying words as whole units (e.g. “mil” as /mil/, and “perch” as /perch/) and using just one sound when trying to generate words from non-words (e.g. “tere” as /te/). There are two parts to this subtest, one which focuses on segmenting letters into sounds and another for segmenting sounds within words. The scores for the first part of the test would be expected to correspond closely with a child’s score on the Phoneme Segmentation subtest of the Test for Phonological Awareness. However, because it does not assess whether a child can segment sounds within words, it cannot provide a complete assessment of phonological awareness skills.
Bishop, D.V.M., Adams, C. & Snowling, M. (1990).
The development of phonological awareness and early reading ability: a longitudinal study. Journal of Child Psychology and Psychiatry and Allied Disciplines, 31(3), 405-418 Nunes, T. & Bryant, P. (2007).
A comparison between two scoring systems for use with the phonological awareness subtest of the Phonological Awareness Test. British Journal of Educational Psychology, 77(2), 461-471 Nunes, T., Bryant, P. E., MacLean, M. & Parlingeo, L. T. (2007).
Using the Phonological Awareness Test to assess phonological awareness in UK clinical populations. Dyslexia, 13(3), 171-190 Stocker, C. & Gathercole, S.E. (2009).
The development of working memory from 4 to 15 years: A longitudinal study. Developmental Psychology, 45(2), 673-688