Chapter 4 Vocabulary, Phonology, and NWR of the Age 2 Cohort
4.2 Vocabulary Development
Children in the age two cohort were still too young at their first testing time to receive any standardized tests. Therefore we measured their vocabulary knowledge, both the receptive vocabulary and the expressive vocabulary, with the MCDI-T checklist (Liu & Tsao, 2010). The parents were required to fill in the MCDI-T checklist in all the three testing sessions. At age 2.5, we included the standardized test PPVT-R (Lu & Liu, 1994) to measure children’s receptive vocabulary, because we believed that children at this age might be old enough to appropriately react to a standardized language receptive task, though a language expressive task might still be too demanding.
It was until when children reached age 3 that we included both standard tests on the receptive vocabulary (PPVT-R) and the expressive vocabulary (REVT, Huang et al., 2011) to assess children’s vocabulary knowledge, in addition to the parent-report measure of early vocabulary (MCDI-T). For the convenience of reference, we would specify the receptive vocabulary score measured with the MCDI-T as MCDI-RECEPTIVE, and the expressive vocabulary score measured with the MCDI-T as MCDI-EXPRESSIVE.
Children’s growth in each task would be examined respectively in the following sections.
53
4.2.1 The expressive vocabulary and receptive vocabulary as measured by MCDI-T. Since both children’s receptive vocbaulary and expressive vocabulary were
inspected with the MCDI-T checklist in this cohort, we first examined the children’s performances in these two vocabulary constructs, and the relationship between these two. Children’s average performances in MCDI-RECEPTIVE (the line with black dots) and MCDI-EXPRESSIVE (the line with hollow dots) across time were graphed in Figure 4.1.In terms of the performances in MCDI-RECEPTIVE, children at age 2 obtained an average score of 471.96 (SD = 99.77), with scores ranging between 232 and 628. At age 2.5, children attained an average score of 638.32 (SD = 60.23), with the range of 485 and 694. At age 3, children had an average score of 686.05 (SD = 10.99). Children’s scores ranged between 651 and 696, suggesting a ceiling effect in using MCDI-T to assess age 3 children’s receptive vocabulary size. In fact, this task has been designed for children below age 3, so it might not effectively reflect the actual variation in children’s receptive vocabulary at age 3.
Correlation analyses were conducted to examine the relationship among children’s receptive vocabulary size at each time points (Table 4.2). MCDI-RECEPTIVE at age 2 was associated with MCDI-RECEPTIVE at age 2.5 (r = .59, p < .01) and MCDI-RECEPTIVE at age 3 (r = .49, p < .05). However, the correlation between MCDI-RECEPTIVE at age 2.5 and MCDI-RECEPTIVE at age 3 only revealed a borderline significance (r = .38, p = .083), probably because children were about to reached the maximum score of the checklist at age 2.5 and age 3. The smaller variation in the children’s performances at these two time points led to the weak correlations.
With regards to the performance in MCDI-EXPRESSIVE, children at age 2 received an average score of 304.87 (SD = 109.89), ranging between 152 and 565. At age 2.5, children attained an average of 567.59 (SD = 85.47), a remarkable increase in size. The score range (409-686) was smaller compared to the score range in age 2. Children
54
seemed to manifest remarkable growth in their expressive vocabulary in this half a year.
At age 3, children’s scores ranged between 574 and 696, with an average score of 665.91 (SD = 28.83), which almost reached the maximum score of this checklist.
Therefore, MCDI-T might not effectively reflect actual variation in children’s expressive vocabulary at age 3.
Then, we examined the correlations among children’s MCDI-EXPRESSIVE score at each time point (Table 4.2). Correlations were found only between MCDI-EXPRESSIVE at age 2 and age 2.5 (r = .59, p < .01). MCDI-EXPRESSIVE at age 3 was not associated with either MCDI-EXPRESSIVE at age 2 (r = .26, p > .05) or MCDI-EXPRESSIVE at age 2.5 (r = .36, p > .05).
Figure 4.1
Children’s Average Performances in MCDI-
RECEPTIVEand MCDI-
EXPRESSIVEAcross Time
Age
2.0 2.5 3.0
MCDI-T (max = 696)
0 200 400 600
Receptive vocabulary Expressive vocabulary
55
Table 4.2
Correlation Matrix of MCDI-
RECEPTIVEand MCDI-
EXPRESSIVEAcross Time
MCDI-R
The correlations between MCDI-RECEPTIVE and MCDI-EXPRESSIVE were examined (Table 4.2). MCDI-EXPRESSIVE at each time point was highly correlated with the concurrent MCDI-RECEPTIVE (r = .77 - .86, p < .01). Therefore, there was close relationship between children’s concurrent receptive vocabulary and expressive vocabulary. Cross-time correlations were only found between MCDI-RECEPTIVE at age 2 with MCDI-EXPRESSIVE at age 2.5 (r = .71, p < .01), which appeared to suggest a strong association between receptive vocabulary size and the expressive vocabulary size half a year later.
A two-way repeated measures ANOVA was conducted to compare children’s performances of MCDI-EXPRESSIVE and MCDI-RECEPTIVE across time. In addition to the significant effect of time (F(2, 42) = 153.11, p < .001, η2 = .88), the results demonstrated a significant main effect of vocabulary (F(1, 21) = 140.26, p < .001, η2
= .87), and a significant interaction effect of vocabulary and time (F(2, 42) = 46.63, p
< .001, η2 = .689). The interaction effect was caused by the faster growth rate of expressive vocabulary in comparison with the growth rate of receptive vocabulary, as shown in Figure 4.1.
56
Planned analyses were conducted to examine children’s growth in MCDI-RECEPTIVE
size. The results showed that children’s receptive vocabulary size increased from age 2 to age 2.5 (t(21) = 9.47, p < .001), and from age 2.5 to age 3 (t(21) = 3.93, p < .001).
Parallel analyses were also conducted to examine whether children demonstrated significant growth in MCDI-EXPRESSIVE. The results revealed that children’s expressive vocabulary size increased from age 2 to age 2.5 (t(21) = 13.38, p < .001), and from age 2.5 to age 3 (t(21) = 5.77, p < .001).
4.2.2 Receptive vocabulary size as measured by PPVT-R. Children in this age
cohort were tested with the standardized test PPVT-R at age 2.5 (Time 2) and age 3 (Time 3). At age 2.5, the average score they obtained in this task was 18.82 (SD = 6.55, range 10-37). Half a year later, they attained an average score of 29.86 (SD = 8.94, range 13-47). Correlation analysis revealed that children’s PPVT-R at age 2.5 and PPVT-R at age 3 were significantly correlated, r = .62, p = .002. The results of a pairedt test revealed a significant growth effect, t(21) = 7.34, p < .001. Children’s performance
in PPVT-R at age 3 was significantly better than their performance at age 2.5.Since children in this cohort were assessed with two receptive vocabulary tests, i.e.
MCDI-RECEPTIVE and PPVT-R, we examined the correlations between children’s performance in these two tasks. The results showed that PPVT-R at each time point was only associated with its concurrent MCDI-RECEPTIVE score. For example, PPVT-R at age 3 was only significantly associated with MCDI-RECEPTIVE at age 3 (r = .49, p < .05). The correlation between PPVT-R at age 2.5 and MCDI-T at age 2.5 was close to borderline significance (r = .39, p = .073). In addition, MCDI-RECEPTIVE at age 2 was associated with neither PPVT-R at age 3 nor PPVT-R at age 2.5. In other words, children’s MCDI-RECEPTIVEscores could not predict children’s PPVT-R scores.
57
4.2.3 Expressive vocabulary size as measured by REVT. In addition to
measuring the age two cohort children’s expressive vocabulary with the parent-report checklist, we assessed children’s expressive vocabulary with the standardized test REVT. Considering that this task required children’s expressive performance, and would be too demanding for children below age 3, we included this task when children in this cohort reached age 3. Children gained an average score of 34.57 (SD = 9.77, range = 15-46, percentile range = 52-99%) in this task.We examined how REVT was associated with MCDI-EXPRESSIVE. The results showed that REVT was correlated with MCDI-EXPRESSIVE at age 3 (r = .54, p < .05). Nevertheless, its correlations with MCDI-EXPRESSIVE at age 2 (r = .40, p = .073) and MCDI-EXPRESSIVE at age 2.5 (r = .40, p = .076) approached but fell short of significance.
MCDI-EXPRESSIVE and REVT were modestly associated, because both assess children’s ability in verbally labeling objects. However, REVT provides a more advanced measurement of children’s expressive vocabulary ability by additionally assessing vocabulary categorization, definition, and reasoning. Therefore, it measures not only the breadth, but also the depth of children’s expressive vocabulary.
As stated in the previous section, MCDI-T checklist might not reliably reflect individual variation in vocabulary knowledge in age 3 children. Therefore, we inspected the relationship between expressive vocabulary and receptive vocabulary which were measured with the standardized tests. Correlation analyses showed that REVT at age 3 was significantly correlated with PPVT-R at age 2.5 (r = .47, p < .05), and PPVT-R at age 3 (r = .57, p < .01). The cross-time correlation seemed to suggest that the receptive vocabulary may serve as the foundation for the development of expressive vocabulary, just as we observed in the correlation between MCDI-RECEPTIVE at age 2 and MCDI-EXPRESSIVE at age 2.5. However, since we did not measure REVT at age 2.5, we could not verify whether there was a cross-time correlation between REVT at age 2.5
58
and PPVT-R at age 3.
4.2.4 Summary of vocabulary development from age 2 to age 3. Children in this
age showed dramatic improvements in their vocabulary size, particular during age 2 to age 2.5. As measured by the MCDI-T checklist, during this half a year, children’s receptive vocabulary had an average increase of 28 words per month, and their expressive vocabulary an average increase of 48 words per month. However, the flatter growth curves during age 2.5 to age 3 might not be a genuine slowdown in children’s vocabulary acquisition rates, but the consequence of the limit of the checklist.The use of the MCDI-T checklist for the assessments of vocabulary size in both the receptive and expressive aspects allowed us to make a direct comparison of these two aspects of vocabulary. The results showed that children tended to have larger receptive vocabulary size than expressive vocabulary size, just as found in previous studies (Clark, 1993; Ingram, 1974). Regarding the correlations between these two vocabulary aspects, we found that MCDI-RECEPTIVE at age 2 was associated with MCDI-EXPRESSIVE at age 2.5.
The association between the earlier receptive vocabulary size and the subsequent expressive vocabulary size was also replicated in the significant correlations between PPVT-R at age 2.5 and REVT at age 3. It is likely that receptive vocabulary could serve as the foundation for the development of expressive vocabulary, at least in this age range.
Analyses in the previous sections demonstrated the average trend of children’s vocabulary growth. Close inspection on the idiosyncrasies of children’s performances revealed considerable variation in children’s onset vocabulary knowledge, though all the children experienced growth in the size of vocabulary knowledge during this age range.
It was of interest whether this difference in vocabulary knowledge would be related to performance in nonword repetition. We would investigate this hypothesis in the last
59
section of this chapter.
We assessed the vocabulary knowledge of the children in this cohort with both the parent-report checklist MCDI-T (in all three testing sessions) and the standardized tests PPVT (at age 2.5 and age 3) and REVT (at age 3 only). Though correlations were found between the parent-report checklist and the standardized tests, the magnitude of the correlations suggested a modest-to-weak relationship between them. The differences between the MCDI-T checklist and the other two standardized tests should be noted.