Learning words: large group time as a vocabulary development opportunity.

Abstract. This study examines the role of curriculum as a vocabulary learning environment by comparing teacher's Large Group Time instruction in two curricula--Creative Curriculum (Dodge, Colker, & Heroman, 2000) and the Doors to Discovery program (2002). Discourse samples were retrieved from the video data of the Large Group Time activity in five Head Start classrooms. The following discourse features were compared in the two curricula: teachers' and children's amount of meaningful talk, vocabulary diversity, teachers' use of root words and rare words, and vocabulary instruction strategies. The data were analyzed with the CHILDES (Child Language Data Exchange System) and qualitative analysis with analytic induction. The results revealed differences in the children's amount of meaningful talk and in the frequency trends of teachers' use of rare words and vocabulary instructional strategies in favor of the Doors to Discovery program. No significant differences were found in the teacher's amount of meaningful talk, root words, and vocabulary diversity. Limitations of, and suggestions for, using the CHILDES system as a discourse analysis are provided.

**********

Learning words in the early years is of utmost importance to reading achievement because an average or greater vocabulary is essential for adequate reading comprehension from grade 3 onward (Biemiller, 2004). To obtain sufficient vocabulary for grade-level reading comprehension, which is deeply implicated in overall academic success, young children need to acquire about two new word meanings per day, on average, from the age of 1, in order to build up a stock of about 6,000 root words by grade 2 (Biemiller & Slonim, 2001). Root words literally help children to "grow" more words and keep pace with peers in developing age-normal vocabulary that meets the reading demands of late primary--and middle-grade texts. For those children adding less than one word per day (on average), reading comprehension likely will falter and fall below grade level in 3rd grade and beyond (Hart & Risley, 2003). Thus, one of the biggest challenges young children face in realizing their literacy potential is acquiring a sufficient number of word meanings for successful reading comprehension. Creating vocabulary learning opportunities for children in the early years is one of adults' most critical responsibilities.

Young children acquire much of their vocabulary through a process referred to as "fast mapping," whereby words are supplied in the immediate context of meaning (Carey, 1978). Direct explanation of words in narrative contexts, such as book reading, also appears to be helpful (Bus, van IJzendoorn & Pellegrini, 1995; Elley, 1989). Most studies of preschool children's word learning show a strong relationship between the amount of adult-child talk in the home and various aspects of word acquisition. Parent talk, for example, exposes children to meanings, word use, and sentence structures that affect not only their word knowledge, but also their propensity to use language as an expression of experience in literate ways (Hart & Risley, 1995; Heath, 1983; Snow, Burns, & Griffin, 1998). When parent talk is responsive, informative, and playful, children develop rich vocabularies that stimulate accelerated word learning; whereas when it is infrequent and less responsive, children develop smaller vocabularies that limit their capacity to learn new words from everyday experience. It is also the case that those children with richer vocabularies get richer, and those with poorer vocabularies struggle to "catch up," which inevitably gets harder and harder to do as time goes on (Stanovich, 1986).

Empirical evidence about how preschool children learn words in early education and care settings is less forthcoming than research on the role that vocabulary plays in children's success as readers. Much is said about the importance of teacher-child conversation, but research information is scant as to robust relationships between salient features of teacher talk and children's word learning. The fact that teachers converse relatively infrequently with individual children on a daily basis is a consistent finding across studies, as is the observation that much of teacher talk dwells on routine matters rather than the elaboration of children's thinking. An array of descriptive and correlational studies show the differential impact of adults' storybook reading discourse, with some more efficacious than others (Yaden, Rowe, & MacGillivray, 2000). For example, dialogic reading, whereby adults teach children to become storytellers themselves by prompting, encouraging, and adding to their responses during reading, has been shown to produce changes in children's vocabulary and oral language development (Arnold, Lonigan, Whitehurst, & Epstein, 1994). There is also some evidence of differential patterns in teachers' literacy-related discourse outside the storybook reading context that may have a bearing on children's language and literacy learning opportunities in child care environments (Rosemary & Roskos, 2002).

Empirical work on the role of the early literacy curriculum in supporting young children's word learning at preschool toward the goal of school readiness and future reading achievement is very new. Recent research highlights the importance of preschool language experiences (e.g., extended teacher discourse, exposure to rare words) in predicting language and early literacy development and later reading comprehension (Snow, 2002). The increasing pressure for high-quality early language and literacy education for all children increases the demand for new and better use of resources within the early childhood setting--such "hard to get" resources as teacher knowledge, better use of time, well-designed materials, curriculum alignment, and trustworthy assessment information. How to improve the quantity and quality of resources available for high-quality instruction that helps all children achieve reading success is the subject of much debate and also a critical matter for education research (Corcoran & Goertz, 1995). Capacity-building strategies that make efficient, effective use of resources (e.g., early childhood language arts curricula, classroom environments, professional development, etc.) are needed to inform improvement and change at different levels of the education and care system, from local settings to state and federal level policies.

To learn about the role of curriculum as a capacity-builder, we examined the effects of the Doors to Discovery early literacy curriculum (2002) and Creative Curriculum (Dodge, Colker, & Heroman, 2000) on indicators of young children's language and literacy development in preschool (Christie, Roskos, Vukelich, & Han, 2003). Results revealed the Doors curriculum's strengths in promoting Head Start children's vocabulary, as measured by the Peabody Picture Vocabulary Test (Dunn & Dunn, 1997), as well as literacy predictors, as measured by Get Ready to Read! (Whitehurst & Lonigan, 2001), when compared to the Creative Curriculum implemented in similar Head Start classrooms by teachers with similar training and backgrounds. Table I shows the results of two test measures from the previous comparison study between Doors to Discovery curriculum and Creative Curriculum.

It is inferred from these results that the demand for more deliberate integration of learning themes and goals across activity settings (large group, small group, play centers) in the Doors approach, coupled with "scripts" for more explicit instruction on vocabulary, concepts of print, and rhyming in activity settings, may have contributed to Doors' stronger impact on children's learning. But this is not clear, because our research did not go far enough in examining specific variables that might account for the two curricula's differential effects on children's literacy skill or word learning. We did not investigate, for example, the extent to which the curricula may have differentially "pressed" teachers' instructional discourse toward more engaging and powerful language and literacy learning environments.

The purpose in this study, therefore, is to go that next step in determining how a curriculum might help create more powerful conditions for children's word learning at preschool. We take the position that a curriculum provides a support structure for the teacher, albeit a malleable one, that she implements with more or less fidelity to the prescribed plan. Yet the very presence of a curriculum exerts an influence on what the teacher says and does, thus helping to shape the learning environment, and consequently, children's opportunities to learn and use words. The current study attempts to analyze differential features of vocabulary learning opportunities in the two curricula of the earlier study--Doors to Discovery (2002) and Creative Curriculum (Dodge, Colker, & Heroman, 2000), focusing in particular on Large Group Time as a vocabulary learning environment in the five Head Start classrooms of the original study.

As a point of information, Doors to Discovery is a comprehensive pre-kindergarten literacy program that focuses on five areas: oral language, phonological awareness, print awareness, alphabet knowledge, and comprehension. The program's basic design includes a variety of learning contexts--large group, small group, and discovery centers--that offer multiple encounters with language and print in thematic units. Creative Curriculum directs teachers to focus on all aspects of development--social/ emotional, physical, cognitive, and language--and relies heavily on the use of play centers.

This study seeks to compare the teachers' instructional discourse in two curricula, by asking the following set of questions to discourse samples taken from Large Group

Time in each of the Head Start classrooms from the previous study:

1) What are the differences in the amount of teacher and child talk in classrooms using the Doors curriculum and those using Creative Curriculum?

2) What are the differences in the quality of teacher talk between the two curricula on the following indices: (i) vocabulary diversity and (ii) use of root words?

3) How did the quality of teacher talk in the Doors curriculum classrooms change, as indicated by the use of rare words (the trend of content vocabulary of the curricular themes)?

4) How do teachers "teach" new words during Large Group Time in each curriculum (strategies for the vocabulary instruction)?

To answer these questions, the CHILDES (CHild Language Data Exchange System) (MacWhinney, 1995; MacWhinney & Snow, 1990) system, a pioneering effort in the field of early language and literacy research, was used. The CHILDES system offers a full range of discourse analysis such as frequency counts, word searches, and morphosyntactic analysis, with language data, and has been successfully used by linguists for more than 10 years. CHILDES provides detailed guidelines for encoding data within an existing format (CHAT) and a set of software, the CLAN (Computerized Language Analysis) programs, which allow users to do a range of automated analysis of transcribed data. We used the CHILDES system to analyze the necessary data for the first three questions. In addition, qualitative analysis was performed to address the last question, that of the differences in teachers' vocabulary instructional strategies.

Method

Discourse samples were retrieved from the video data of the Large Group Time activity collected in the original study. We briefly summarize the original study below to frame this current study.

Background

Participants and Setting. The original study was conducted in five Head Start classrooms located in five different schools in a large metropolitan area in the Southwest United States, including two schools with predominantly English-speaking children, two with predominantly Spanish-speaking children, and one school with a mix of English--and Spanish-speaking children. All children were from low-income families, with a maximum income of $17,650 for a family of four. One of the English-speaking children's classrooms and one of the Spanish-speaking children's classrooms were randomly assigned to the experimental treatment (Experimental A and B), and the other English--and Spanish-speaking classrooms were assigned to the control treatment (Control A and B). The classroom with the mix of English- and Spanish-speaking children was assigned to the experimental treatment (Experimental C) in order to investigate how the curriculum functioned in a classroom with a mixed-language environment. The sample consisted of 75 preschoolers, ages 3 to 5 years old; the sample size decreased to 53 children at the end of the four-month study, due to family mobility and high rates of absenteeism in obtaining pre-post test results.

The teachers in the five classrooms had either an A.A. or a B.A. degree in early childhood education, and their years of teaching ranged from three to 15. Prior to the study, all five teachers used the Creative Curriculum (Dodge, Colker, & Heroman, 2000) as a foundation for their academic curriculum. All classrooms had a similar layout, containing the same types of learning centers (e.g., dramatic play, art, blocks, library, math/science, and table toys) and similar materials. All had similar schedules, with a designated "circle time" (Large Group Time), during which the teacher taught the whole class, and a free-choice period, during which children dispersed to the different learning centers.

Procedures. For three and one-half months, the teachers in the three experimental classrooms used three units from the Doors to Discovery language and literacy program: Vroom, Vroom (topic: transportation), Build It Big (topic: construction), and Tabby Tiger's Diner (topic: restaurants and food). Each unit was used for one month, and there was a two-week break between the second and third units for spring break and mandated testing. The curriculum integrates Large Group, Small Group, and Discovery Center activities around stories, songs, and related vocabularies within the same theme unit.

Teachers in the two control classrooms used the Creative Curriculum (Dodge, Colker, & Heroman, 2000) as the basis of their language and literacy curriculum. They began each day with a large group, circle time activity, which usually involved storybook reading, songs, and vocabulary instruction. After Large Group Time, the children dispersed to self-selected learning centers, where they engaged in art, dramatic play, block construction, and a variety of other activities. There was no content alignment between circle time and play center activities.

Prior to the intervention, the three teachers in the experimental classrooms attended a five-hour training session that acquainted them with the goals, structure, and activities of the Doors to Discovery early literacy program. This session was conducted by a professional development specialist employed by the Wright Group/ McGraw-Hill. The teachers were given specific guidance for implementing the first unit, Vroom, Vroom. They were encouraged to use all the Doors curricular components as well as ancillary materials, such as the audio CDs, story character props, activity cards, word cards, and take-me-home books. Prior to the introduction of the second and third units, a research assistant visited each teacher and provided suggestions and guidance as to how to implement the unit.

Each experimental classroom was visited once a week, and each control classroom was visited every two weeks. During these visits, the research assistants videotaped 10 to 15 minutes of ongoing activity and made video journals to supply details for the video clips and field notes to record what was occurring in parts of the classroom that were not being videotaped. The timing and focus of the videotaping and field notes alternated with each visit. On one visit, the research assistant would record what was occurring during the Large Group Time part of the curriculum, and on the next visit, the focus would shift to the Learning Center activities. As a result of this alternating schedule, six Large Group Time and six Learning Center observations were made in each experimental classroom, and three Large Group Time and three Learning Center observations were made in each control classroom. Research assistants also collected copies of each teacher's weekly lesson plans and information on each subject's school attendance.

Discourse Analysis

Sample. Three 10-minute video segments of Large Group Time, one from three different points of time (onset, mid point, end of study), were collected from each of the five classrooms to total a 15-sample pool, consisting of 150 minutes of instruction. Each discourse sample was transcribed verbatim to conform to requirements for Computerized Language Analysis (CLAN), as per the CHILDES system (MacWhinney, 1995).

Analysis. The analytic goal was to examine the amount and verbal qualities of teacher talk in Large Group Time as a vocabulary learning opportunity and to describe the nature of vocabulary instruction in this setting. Information about each sheds light on the role of curriculum in building the instructional capacity of the setting for children's word learning.

To measure differences in the amount and qualities of instructional discourse in the Doors (experimental) and Creative Curriculum (control) classrooms, the transcripts of each video sample (a total of 15 transcripts) were made for Large Group Time. Then, each transcript was formatted to CHAT transcripts, which enable language transcripts to be analyzed in the CLAN program. Then, the following features were measured using the CLAN program:

* Teachers' ratio of morphemes over utterances, or the amount of meaningful talk

* Children's ratio of morphemes over utterances; also, the amount of meaningful talk in setting exchanges

* Vocabulary diversity of teacher talk, or the number of different words used in exchanges with children

* Frequency of root words in teacher talk

* Ratio of type/token of root words in teacher talk, or the number of different root words used as compared to the total number of root words said

* Frequency of rare words (from the Doors curriculum) of teacher talk within the experimental classrooms.

The rare word analysis consisted of words culled from three Doors thematic explorations. To ensure that the list consisted of rare words, defined as low frequency words, the list was cross-referenced with a list of common words generated in the Home-School Study, based on the Dale-Chall Readability Index (Tabors, 2004). Corresponding or matched words between the rare word list and the common word list were deleted, yielding a total of 275 rare words. Frequency of rare words in teacher discourse was then determined by using CLAN. A list of 769 "easy words" was used to determine the frequency of root words in teachers' discourse in both the Experimental and Control sites using the CLAN program.

Results of these measures were compared, using analysis of variance (ANOVA) (Glass & Hopkins, 1996) to uncover differences in teachers' discourse within two curricula. To examine the trends of rare word frequency used by teachers in the Doors program, each frequency was compared in different time periods by adding time variables into the ANOVA.

In order to answer the question "How do teachers 'teach' new words during Large Group Time in each curriculum?," qualitative methods were used to analyze the teachers' vocabulary instructional strategies during Large Group Time. An analytic induction strategy was applied to the discourse samples, scanning the data to generate categories and strategies of vocabulary instruction (Goetz & LeCompte, 1984). Two basic approaches emerged: "embedded," when teaching words occurred incidentally in the flow of talk; and "direct," when teaching words was more deliberate and set apart from free-flowing conversation. Within these larger frameworks, five strategies of instruction were indicated as shown in Figure 1.

A Two-Step analytic procedure was used to code the 15 sample pools for these instructional categories and strategies. First instances of embedded and direct vocabulary instruction were located and bracketed. Each instance was then coded for evidence of the five strategies.

Results

Data addressing the amount of teacher and child talk (Question One) and the quality of teacher talk (Question Two) between the two curriculums are reported in Table 2.

The Amount of Meaningful Talk Comparison of the ratio of morphemes over utterances in teachers' discourse samples showed no differences (p=.240) in the amount of meaningful teacher talk between teachers using Doors to Discovery and teachers using Creative Curriculum, although the average amount of meaningful talk was higher in the Experimental (Doors) teacher group. Spotlighting this ratio is important, because it shows the proportion of meaningful words to less meaning-bearing sounds and words typically uttered in the speech flow (e.g., connectives, non-word utterances ["hmmm"] or single words ["Yes"]). Based on CLAN, these results indicate that the amount of meaningful teacher talk between the Control and Experimental groups did not differ substantially.

The ANOVA showed statistically significant differences, however, between Control and Experimental groups in children's ratio of morphemes over utterances during the Large Group Time (*p=0.029; p < .05), thus indicating that during Large Group Time, children in the Doors curriculum engaged in more meaningful talk. Comparing this result with that of the Experimental teachers' amount of talk suggests a pattern of teacher-child interactions that may have allowed time for more child talk, thus reducing the opportunity for teacher talk. Teachers talked less, in other words, because children talked more, which may have afforded the children opportunities to say, and practice using, words in the Large Group Time context.

Vocabulary Diversity and Root Words Vocabulary diversity in the teachers' discourse showed no significant differences, although the mean of the Experimental teacher group was higher. That is, these teachers used approximately the same variety of words during Large Group Time. The frequency of root word (p=.716), and the ratio of type/token of root words (p=.178), also showed no significant differences between groups. The large standard deviations in exposure to root words and vocabulary diversity between teachers within each group, however, are notable, and suggest great variability in the instruction provided during Large Group Time in these classrooms. This is to say that children's exposure to diverse words and root words via teacher discourse varied widely across classrooms during the Large Group Time, with some teachers offering far more and others much less exposure to root words.

The Use of Rare Words in Doors Classrooms In Question Three, we asked "How did the quality of teacher talk in the Doors curriculum classrooms change, as indicated by the use of rare words (the trend of content vocabulary of the curricular themes)?" Table 3 shows the results of the use of rare words in different time periods: the beginning, middle, and end of the Doors curriculum implementation across the three Experimental classrooms.

The results indicate that teachers' use of rare words increased significantly over time during Doors implementation (p=.004) in all experimental classrooms, thus affording children significantly more exposure to rare words in the third theme than in the first theme. The differences were not significant until the middle of the implementation period. During the last theme, however, the teachers used many more rare words than they did in the previous theme. This result may explain the higher vocabulary diversity of the experimental groups (63.07 vs. 73.81) but the lesser frequency of root words (558.50 vs. 505.90) in Table 2. Teachers in the Doors program may have replaced the frequent use of root words with the increasing use of rare words. Figure 3 shows the rare word frequency trends. Experimental classroom B teacher showed the greatest increases in the use of rare words, from 5 words to 32 rare words, across the three themes. Experimental classroom teachers A and C also increased their use of rare words from only 1 or 2 words to 17 words during the last theme. Such trends in frequency of rare word-use reveals discourse patterns that might indicate the cumulative contributory effect of the curriculum as a resource in the learning environment.

[FIGURE 3 OMITTED]

Vocabulary Instruction in Large Group Time

Question Four asked, "How do teachers 'teach' new words during Large Group Time in each curriculum (strategies for the vocabulary instruction)?" Results of the qualitative analysis of teacher discourse samples are summarized in Table 4. We identified a total of 102 word instruction episodes (instances of providing new words), with the majority occurring during Large Group Time in the Doors classrooms (Doors=71 episodes; Creative Curriculum=31 episodes). Most instances of instruction were of the direct type across both curriculums and consisted of a basic two-step procedure: 1) the teacher says the word (e.g., spatula) and 2) the teacher asks students to say the word, as in "Say "spatula.'"

It is noteworthy, however, that teachers' vocabulary instruction in the Doors classrooms included decidedly more strategies beyond Naming and Saying with 55 percent, including one or more additional features of a word (e.g., identifying attributes), in contrast with the Creative Curriculum classrooms, where only 13 percent of the episodes added more informational features.

For example, in one episode, a Doors teacher said, "This is a menu. Can you say menu? [menu] Menu, right. And what does a menu do? What's it for? No ideas? Well, it tells you what they are serving at the place you are eating. It might say 'hamburgers.' It might say 'steak.' What do you know about that?" This tendency to add more detail about words and phrases occurred more frequently in the Doors episodes than in the Creative Curriculum episodes.

Conclusions

Drawing on the data set from the earlier work, we undertook this study to learn more about curricular influences on teachers' discourse as a critical element of the word learning opportunities and also to explore CLAN as an analytic tool in this research work. The findings were rather complex.

First, the quantitative analysis of teacher talk showed no significant differences between teacher discourse in the two curricula--Doors and Creative Curriculum. However, children in the Doors program displayed significantly more talk than their counterparts in the Creative Curriculum classrooms. This suggests that the two curricula differentially affected the amount of child talk, but not the total amount of teacher talk. When considering the teacher-child interaction patterns during the Large Group Time, it is possible that the Doors curriculum may have allowed more time for child talk, thus reducing the opportunity for teacher talk. Recall that the length of Large Group Time taping was held constant across classrooms at 10 minutes. Hence, in the 10-minute Large Group Time, if children were to talk more, teachers presumably would need to talk less. And talking more may have afforded the children opportunities to say and practice using words more often than their Creative Curriculum counterparts in the Large Group context. It is possible that instruction in which teachers simply talk more may not result in young children learning more vocabulary. This possibility is supported by our earlier finding, which showed that children in the Doors curriculum scored much higher than children in Creative Curriculum on the Peabody Picture Vocabulary Test (Dunn & Dunn, 1997).

The second finding raises a question--What other differences seem to result in children in the Doors curriculum acquiring more vocabulary (as indicated by the earlier test results)? The results of Questions Three and Four suggest that teachers in the Doors program displayed significant increases, over time, in the frequency of the use of rare words across the different thematic units. It is possible that this could have a cumulative contributory effect on children's vocabulary learning. Teachers in both curricula initially might have focused on common words. Teachers using the Doors curriculum gradually made changes in their vocabulary instruction by including a greater number of rare words in their Large Group Time instruction, presumably as they became more comfortable with the curriculum. However, this conclusion requires investigation over a longer period of time and the counterbalancing of the use of the themes. In the present study, the three experimental classrooms taught the thematic units in the same order. To ensure that the results are indicative of teachers' vocabulary instruction changing over time, rather than changes due to shifts in the topics under investigation, continued investigation is required.

Third, the qualitative analyses revealed that the teachers using Doors used more sophisticated strategies to introduce new vocabulary words, such as pointing out orthographic features of words, offering definitions, and providing explanations. These "teaching in context" efforts may contribute to children's initial mapping of the meaning of a new word by providing them with some useful information about the word's orthographic and semantic properties, thus rendering it more memorable.

In short, the overall amount of teacher talk in the two curricula was not significantly different. However, the amount of rare words and use of instructional strategies were different in experimental classrooms and between the experimental classrooms (thus suggesting differences in the use of the Doors curriculum by the three teachers, a topic in need of additional investigation). The teachers' use of more rare words and more sophisticated instructional strategies may be partly responsible for our current and earlier findings: 1) that children experiencing the Doors curriculum engaged in more meaningful talk and 2) that they scored significantly higher than their peers in Creative Curriculum classrooms on the Peabody Picture Vocabulary Test (Dunn & Dunn, 1997), as revealed in the earlier study.

Although the curriculum resource clearly seems to exert an influence on the teachers' instructional strategies and kinds of vocabularies that they choose to teach, its influence is not as great on the frequency and types of discourse. Perhaps these kinds of discourse are part of teacher personality and may be hard to change during a short period (3 months). Longitudinal data on teacher discourse are needed to test this hypothesis in different curriculum settings.

Limitations and Future Research

In terms of methodology, a pioneering step was taken in the use of the CHILDES system to analyze teacher discourse within the vocabulary learning environment in the preschool classrooms. In addition, we used trend analysis to understand the teachers' use of rare words and analytic induction as a qualitative analysis. Using these two methods in an integrated way allowed us to reveal the influence of two different curricula on teachers' vocabulary instruction. We encourage our colleagues to join us in the use of the CHILDES system to better interpret classroom discourse.

We have identified several limitations of the work reported here and provided suggestions for future research. First, large standard deviations were observed in the frequency of words, vocabulary diversity, and root word analysis. This denotes great variability within the classrooms using the same curriculum. This variability may have underscored the powerful teacher influences on the curriculum by reducing the chance for a valid comparison between the two curricula, and it likely hindered our ability to show significant differences.

Second, simple comparisons using CLAN that measure differences between the two curricula may not be the best way to study teacher discourse. Teacher discourse possibly is influenced by the activity the teacher select to use during the Large Group Time. That is, a teacher can choose story books, word cards, songs, alphabet charts, or a cloze activity as vocabulary instruction methods. Each activity possibly calls for the use of different teacher discourse. If this is case, comparing teacher discourse during a Large Group Time when a storybook is being read with teacher discourse when a cloze activity is being used may not be an ideal way to study teacher discourse as a word learning environment between curricula. Comparing teacher discourse during Large Group Time when all classrooms are using the same activity (storybook reading, for example) may be a better way to compare teachers' vocabulary instruction in different curricula settings.

Third, we acknowledge that there likely is an inherent weakness in the use of the CHILDES system; this system focuses on root and rare word usage, and does not allow researchers to move beyond the word level analyses to reveal other qualities of teacher discourse, qualities such as engagement in connected conversations with children, that the curriculum might be supporting, if not prompting. In the classic preschool study Under Five in Britain, Bruner described connected discourse as "the forge in which conversational skill is fashioned"--the kinds of skills that prepare children for making themselves heard (1980, p. 62). Such skills develop best in the course of "chats" that transpire while making things, discussing storybooks, telling about what happened, and remembering past experiences. In the Doors curriculum, teachers were guided to organize their Large Group Time around topics of study (e.g., construction, food, environment) that might stimulate more teacher-child "chats," whereby children can hear, say, and use words they know and new words they are learning. The CHILDES system is not sensitive enough to "track" whether or not these kinds of connected exchanges occur and, if so, how often they occur. Thus, a fine mix of quantitative and qualitative methodologies may be necessary to capture and describe teacher discourse and its relationship to the curriculum being used. Those curricula that support teachers' talk in more connected conversations within and across activity settings (e.g., large group, small group, and play) may contribute more generously to children's vocabulary development, because they permit "protracted further experiences" that are necessary for functional word learning, or extended word mapping (Carey, 1978, p. 274).

Fourth, our study showed the influence of curriculum on teachers' use of rare words and instructional strategies on children's vocabulary learning. Considering the accumulated effects of a curriculum on children's experience, our results give us pause and invite us to consider the role of curriculum more critically. The strengths and weaknesses of early literacy curricula need to be examined in terms of their specific influence on the types of teacher discourse. Our study suggests little about the quality of early language and literacy curriculums in terms of design (scope and sequence), research base, teaching protocols, implementation, and self-monitoring tools for teachers so that this resource can be used effectively to the benefit of all children. Putting these stronger curriculums to the test, and conducting a thorough comparison of them, may show the actual curriculum influences on the language learning potential of the setting (including teacher knowledge and skill), and may result in quality curriculum that creates high-quality language environments in every early childhood classroom.

Table 1
The Assessment Results From the Previous Study on Comparison
Between Doors to Discovery Program and Creative Curriculum

                             PPVT Test **

                      Pretest             Post test
Treatment/
classroom           Mean      SD       Mean       SD

Experimental A     43.22     14.87     54.78     16.34
Experimental B     23.69     22.76     30.33     21.23
Experimental C     22.19     15.40     33.06     18.80
Control A          36.44     15.99     34.67     17.07
Control B          22.56     17.23     26.00     15.49

                         Get Ready to Read Test

                      Pretest             Post test
Treatment/
classroom           Mean       SD       Mean       SD

Experimental A     10.56      3.78     11.78      4.47
Experimental B      7.18      4.26      8.75      3.55
Experimental C      7.06      2.93      9.25      3.26
Control A           8.00      4.61      8.56      4.36
Control B           5.44      2.96      5.14      2.91

Note 1. Experimental classrooms implemented Doors to Discovery
program and Control class-rooms used Creative Curriculum.

Note 2. ANCOVA analysis, ** p = .000 (p < 01) for PPVT test,
p = .06 for Get Ready to Read Test.

Figure 1
Strategies of Vocabulary Instruction

Strategy                          Example

Saying the word to children       "tow truck."

Asking children to pronounce      Say "tow truck."
their word

Talking about grapho-phonemic     What sound does "tow truck" start
or semantic properties            with? T-t-t-tow t-t-tr-truck. What
                                  letter is that?

                                  Tow trucks pull cars out of places.
                                  They might take a car to a garage to
                                  get fixed.

Defining the word                 A tow truck is a kind of truck that
                                  hauls other vehicles from one place
                                  to another. This is a "tow truck"
                                  (points to picture).

Explaining the word by adding     Your dad owns a "tow truck," doesn't
details and children's            he, Jared? He uses the tow truck to
experience                        get cars from accidents and takes
                                  them to his garage to get fixed.
                                  Right? I remember you told us about
                                  this once. What color is your dad's
                                  tow truck? How big is it?

Table 2
Differences in Teacher-Child Discourse During Large Group Time
in the Two Curriculums

                     Control group      Experimental group   T score
                       Creative         Doors to Discovery
                       Curriculum

                     Mean        (SD)     Mean        (SD)

Teacher's ratio      4.56        1.44     5.50        1.46      1.23
of morphemes
over utterances

Child's ratio of     1.68         .60     2.62         .79      2.45
morphemes over
utterances

VOCD                63.07       23.63    72.81       23.61       .78

Frequency of       558.50      322.05   505.89      228.23     -.372
root words

Ratio of              .21   3.354E-02      .24   5.240E-02     1.425
Type/token
of root words

Note. ANOVA analysis * p < .05

Table 3
Trend Analysis of the Use of Rare Words Within the
Experimental Classrooms

               Standardized
               Coefficients       T       Sig.
                   Beta

(Constant)                      -.037      .972
Class 1            .000          .000     1.000
Class 2            .293         1.923      .127
Time 1            -.015         -.101      .924
Time 2             .911         5.970      .004 **

Note. ANOVA analysis ** p < .01

Table 4
Percent of Vocabulary Instruction Strategies in Teacher
Discourse Samples

Teaching Strategy                            Control      Experimental

                                             Creative       Doors to
                                            Curriculum     Discovery

Saying the word (Teacher)                      94%            97%
Asking children to pronounce the word          65%            48%
Talking about graphophonemic or semantic        6%            35%
  properties of the word
Defining the word                               6%            13%
Explaining the word                             3%             7%

References

Arnold, D., Lonigan, C., Whitehurst, G., & Epstein, J. (1994). Accelerating language development through picture book reading: Replication and extension to a video tape training format. Journal of Educational Psychology, 86(2), 235-239.

Biemiller, A. (2004, February). Words and primary grade learners. Presented at the annual Ohio Faculty Seminar, Akron, OH.

Biemiller, A., & Slonim, N. (2001). Estimating root word vocabulary growth in normative and advantaged populations: Evidence for a common sequence of vocabulary acquisition. Journal of Educational Psychology, 93,498-520.

Brunet, J. (1980). Under five in Britain. Ypsilanti, MI: High/Scope.

Bus, A., van IJzendoorn, M., & Pellegrini, A. (1995). Joint book reading makes for success in learning to read: A meta-analysis on intergenerational transmission of literacy. Review of Educational Research, 65, 1-21.

Carey, S. (1978). The child as word learner. In M. Halle, G. Miller, & J. Bresnan (Eds.), Linguistic theory and psychological reality (pp. 264-293). Cambridge, MA: MIT Press.

Christie, J., Roskos, K., Vukelich, C., & Han, M. (2003). The effects of a well-designed literacy program on young children's language and literacy development. In Head Start Bureau (Ed.), The first eight years--Pathways to the future: Implications for research, policy, and practice (pp. 447-448). New York: Mailman School of Public Health, Columbia University.

Corcoran, T., & Goertz, M. (1995). Instructional capacity and high performance schools. Educational Researcher, 24(9), 27-31.

Dodge, D., Colker, L., & Heroman, C. (2000). Creative curriculum for early childhood: Connecting content, teaching and learning. Washington, DC: Teaching Strategies, Inc.

Doors to discovery. (2002). Bothell, WA: Wright Group/McGraw-Hill.

Dunn, L., & Dunn, L. (1997). Peabody Picture Vocabulary Test (3rd ed.). Circle Pines, MN: American Guidance Service.

Elley, W.B. (1989). Vocabulary acquisition from listening to stories. Reading Research Quarterly, 24, 174-187.

Glass, G., & Hopkins, K. (1996). Statistical methods in education and psychology (3rd ed.). Boston: Allyn and Bacon.

Goetz, J., & Lecompte, M. (1984). Ethnography and qualitative design in educational research. New York: Academic.

Hart, B., & Risley, T. (1995). Meaningful differences in the everyday experience of young American children. Baltimore: Paul H. Brookes.

Hart, B., & Risley, T. (2003, Spring). The early catastrophe: The 30 million word gap by age 3. American Educator, 27(1), 4-9.

Heath, S.B. (1983). Ways with words. Cambridge, MA: Cambridge University Press.

MacWhinney, B. (1995). The CHILDES project: Tools for analyzing talk (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

MacWhinney, B., & Snow, C. (1990). The Child Language Exchange System: An update. Journal of Child Language, 17, 457-472.

Rosemary, C., & Roskos, K. (2002). Literacy conversations between adults and children at child care: Descriptive observations and hypotheses. Journal of Research in Childhood Education, 16(2), 212-231.

Snow, C. (2002, May). The central place of language development in early literacy preparation. Paper presented at the annual meeting of the International Reading Association, San Francisco.

Snow, C., Burns, S., & Griffin, P. (1998). Preventing reading difficulties in young children. Washington, DC: National Academy Press.

Stanovich, K. (1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly, 21, 360-407.

Tabors, P. (2004). One child, two languages: A guide for preschool educators. Baltimore: Paul H. Brookes Publishing.

Whitehurst, G., & Lonigan, C. (2001). Get ready to read! An early literacy manual: Screening tools, activities, and resources. Columbus, OH: Pearson Early Learning.

Yaden, D., Rowe, D, & MacGillivray, L. (2000). Emergent literacy: A matter (Polyphony) of perspectives. In M. Kamil, P. Mosenthal, P. D. Pearson, & R. Barr (Eds.), Handbook of reading research (vol. III, pp. 425-454). Mahwah, NJ: Erlbaum.

Myae Han University of Delaware 

Kathleen Roskos John Carroll University 

James Christie Arizona State University 

Sonia Mandzuk John Carroll University

Carol Vukelich University of Delaware