7. Implications for curriculum

numeracy performance (TEMA) as the outcome variable only the PPVT and PALS accounted for a significant proportion of the variance for younger children. However, for the older children, who were passing more advanced numeracy items on the TEMA, the PPVT was no longer significant, but instead the PALS and the nonwords repetition measure were significant. This suggests that receptive language is more influential for lower level number tasks, but not more advanced tasks. Examples of lower level number tasks on the TEMA were nonverbal items, counting small numbers of objects, and answering questions about cardinality for small numbers. Higher level number tasks involved using number symbols, solving simple addition problems, and comparing numbers to make decisions about size. It appears that as children get older and are able to succeed on more advanced number tasks receptive language accounts for less variance while phonological awareness (PALS) and phonological working memory (non-words repetition task) account for

Of the three measures of phonological working memory only the non-real words repetition task was predictive of older children's numeracy performance. The real words repetition task, non-words repetition task and the Teacher-rated BRIEF-P subscale of Working Memory were not significantly correlated suggesting that they each tapped different characteristics of phonological working memory. This variation is not surprising given that the BRIEF-P rating is based on the teacher's impression across the weeks and months of working with the child while both the real words repetition and the non-words repetition tasks are on-the-spot measurements based on the child's accuracy at that moment. It is also the case that the non-words repetition task is carefully designed to include certain combinations of language sounds rather than actual words and that is not the case for the Brief-P subscale

The non-words repetition task is frequently used to assess language acquisition, and to diagnose and understand the characteristics of language impairment because it measures both phonological working memory and several other phonological components that underlie the learning of words [33]. Success on the non-word repetition task requires children to identify the units of speech (phonemes) that compose words and depends on recognizing the lawful combinations of language sounds. A bi-directional relationship between receptive vocabulary and the nonwords repetition task exists with the skills measured by the non-words repetition task supporting vocabulary growth up to age 5. By age 5 children's vocabularies have reached sufficient size to support the skills of identification of language units, processing and combining language sounds that the non-words repetition task

Our results are consistent with the literature on the relationship between recep-

tive vocabulary and the non-words repetition task. The shift from the PPVT predicting younger children's number skills to the non-word repetition task predicting older children's number skills occurs at about the same time that receptive vocabulary more strongly supports the skills measured by the non-words repetition task. One possible explanation is that the concept of a unit is common to both phonological working memory and number skills. As children are improving in their identification and use of language units their number skills, which are also based on understanding units [34], are also improving. Another possible explanation is that a larger vocabulary contains more advanced words, which supports both phonological working memory and number skills. Although the PPVT does not include many words that are specific to number it is possible that as children's vocabularies grow that more number words are included. Purpura and Logan [35] found that a number specific vocabulary predicts young children's number performance. It is not possible to select from these explanations, or other possibilities

more variance.

Early Childhood Education

of working memory.

measures [33].

144

We agree with Krajewski and Schneider [6] that phonological awareness and early number skills should be taught together early in the preschool years in order to reinforce skill development. An added bonus, as noted by Krajewski and Schneider, is that phonological awareness skills are also necessary to the development of reading so early exposure facilitates the development of two key skill domains. Additionally, it appears from the work of Chu et al. [36] that both numeracy and preliteracy skills, such as recognition of the alphabet in preschool, predict achievement in numeracy and reading by the end of kindergarten. A strong case is emerging for the importance of learning about numeracy in preschool in order to support the development of both numeracy and reading.

With respect to executive functioning, especially phonological working memory, we agree with Clark et al. [11] and Clements et al. [21] that intentional training and practice, presented in a developmentally appropriate way, could scaffold children's development of the executive functioning skills that are specifically geared to learning number skills. Phonological working memory involves remembering the units of language. As children become more proficient with phonological awareness moving to an emphasis on explicit identification of the units of language is appropriate.

## 7.1 Influence of type of childcare setting

Type of childcare setting is a demographic characteristic not often included in research on children's number skills. While there were significant differences between program types for children's scores, there were no differences in the predictors for type of childcare setting. The same patterns held for both center and family childcare children. This particular finding has never before been reported in the literature, to our knowledge and suggests that the same types of training and curriculum are likely to be effective for children in both settings.

#### 7.2 Applied measures

Many studies use laboratory measures for phonological awareness and executive functioning. Our work demonstrates that applied measures of phonological awareness and executive functioning which are composed of activities that could be observed or implemented by teachers serve as predictors of early numeracy performance. Both the PALS and BRIEF-P are measures that can be used by teachers, and our discussion of curriculum implications above indicates how the information gained from these measures could help teachers plan both curriculum and intervention. The measure of phonological working memory, the non-words repetition task, is one that can also be readily understood by teachers and used as the basis for classroom activities.
