Teacher talk in first grade classrooms: The model of language provided by early elementary teachers.

摘要

The purpose of this study was to provide initial documentation of the complexity of first grade teachers' oral and academic language during instruction and interaction with their students. Oral language includes sounds, words, sentence structure, meaning, and the meaning of sentences and words. Academic language, the language that is used in school, is more complex and specific than the oral language used in the home. The findings of multiple studies suggest that children's early oral language may predict later reading comprehension achievement. Children's oral language has also been correlated with the oral language of their primary caregivers. Upon entry to school, teachers become primary caregivers and the oral language model that students experience daily. Yet, teachers' level of oral language during instruction and interaction with their students has not been documented until now. In this mixed methods design study, I asked two questions. My first question was, What is the level of first grade teachers' language use in their classrooms? Randomly selected language samples from four teachers were transcribed and analyzed using Systematic Analysis of Language Transcripts (SALT) software to code teachers' classroom language. The variables used to document teachers' language level were syntax (mean length of utterance, MLU), vocabulary diversity (number of different words/total number of words, type/token ratio), and proportion of academic language words (words identified in kindergarten and first grade district curriculum maps as academic language words and Coxhead's [1998] Academic Word List). Teachers' language was descriptively different for individual teachers. The MLUs for teachers in time samples were 6.91 (SD=.8; T1), 6.04 (SD=66; T2), 6.59 (SD=.99; T3), and 6.62 (SD=1.22; T4). The type/token ratios for teachers were .27 (SD=.04; T1) .34 (SD=.05; T2), .33 (SD=.05; T3), and .33 (SD=.05; T4). The SALT database indicates that the MLU of a first grade student would be 6.9 (SD = 1.25) words. The type/token ratio (word diversity) of a first grader would be .4 (SD = .09). The percentages of K1 academic language words in time samples were 7.70% (T1), 7.29% (T2), 8.75% (T3), and 7.68% (T4). The percent of words from Coxhead's (1998) Academic Word List in language samples for each teacher was below .10%. Language samples were then standardized by taking the first 54 utterances (interquartile range) of each teacher's randomly selected language samples. The combined mean of the MLU in standardized utterance samples of all teachers was 6.47 (SD = 2.00). The type/token ratio in the combined standardized utterance samples of all teachers was .43 (SD = .05). Statistically significant, positive correlations were found between number of different words and MLU for all teachers and all teachers combined in standardized utterance samples. There were also statistically significant positive correlations between teachers' use of words from Coxhead's (1998) Academic Word List and MLU for three teachers and a weak, statistically significant correlation between MLU and Coxhead's (1998) Academic Word List words. My second question used for triangulation asked, What are first grade teachers' perceptions of their role in the development of students' oral and academic language and of the supports and barriers to facilitating students' language development? Interviews with each teacher elicited their perceptions of their oral and academic knowledge, the supports and barriers they experience when supporting their students' language development, and their professional development experiences. Three of the teachers identified the need for more time as a barrier for supporting students' language growth, and the fourth teacher identified the need for her students to have more practice using language. Statements made during their interviews, for the most part, matched with teachers' talk in their classrooms. Suggestions for future research conducted with a larger, more representative sample may allow comparisons of language between teachers and across schools and demographics of the students.