Vocabulary Development in Armenian Children Attending Armenian-English Bilingual Preschools Alice Hovsepian, Ph.D. Candidate Carla J. Johnson, Ph.D. Department of Speech-Language Pathology University of Toronto, 2009
Purpose Investigating Group Differences in Vocabulary Performance In Junior and Senior Kindergarten (JK & SK) bilinguals In Armenian and English languages In Picture Identification and Picture Naming tasks At Time 1 and Time 2, in a six-month interval Predicting Individual Differences in Vocabulary Performance
Rationale Bilingualism, a rule rather than an exception (Bhatia and Ritchie, 2004) Vocabulary related to other aspects of language (Conboy & Thal, 2006) and literacy (e.g., Dickinson et al., 2003; Lindsey et al., 2003) Few studies on young bilingual children (McCardle & Hoff, 2006; Fernald, 2006) No studies on Armenian-English bilinguals
Questions & Predictions Investigating group differences in vocabulary performance 1. Group effect: SK > JK 2. Time effect: Time 2 > Time 1 3. Modality effect: Picture Identification > Picture Naming 4. Language effect: Armenian > English 5. Language by Time interaction: Armenian growth rate < English growth rate Predicting individual differences in Armenian and English vocabulary performance
Participants Bilingual kindergarteners attending balanced bilingual programs Armenian spoken at home at least by one parent
Vocabulary Measure Researcher-developed parallel bilingual vocabulary test Four vocabulary lists, each with 55 items, equated on their mean age of acquisition (AoA) values The lists counterbalanced across four tasks: Armenian Picture Identification (Arm PI) Armenian Picture Naming (Arm PN) English Picture Identification (Eng PI) English Picture Naming (Eng PN)
Participant Characteristics Child Characteristics (measured directly) Nonverbal Cognitive Ability (IQ) Brief IQ Screener, Leiter International Performance Scale-Revised (Roid & Miller, 1995, 1997) Phonological Short-Term Memory (PSTM) The Hebrew Pseudoword PSTM Task (Geva et al., 2000) Family Characteristics (parental questionnaire) Parental Education Armenian & English Exposure/Experience
Child Characteristics JK (N = 20) SK (N = 22) M (SD ) M (SD ) p Range Range Age 51.65 (2.58) 63.15 (3.43) <.01 (month) 47.0 57.0 58.5 69.0 IQ 119.40 (13.68) 102 151 110.91 (13.24) 83 147.048 PSTM 37.59 (13.17) 42.42 (13.31).245 (%) 7.4 63.0 18.5 66.7
Family Characteristics JK (N = 20) SK (N = 22) M (SD ) M (SD ) p Range Range Parental Education 15.71 (2.65) 17.44 (2.90).062 (year) 10.0 20.0 11.0 24.0 Armenian Expo./Expe. 76.85 (22.43) 65.80 (32.06).207 (%) 35.0 100.0 1.3 100.0 English Expo./Expe. 15.69 (18.98) 32.28 (31.29).047 (%) 0.0 65.0 0.0 86.7
Results: Group Differences Mixed Analysis of Covariance Vocabulary performance analyzed as a function of: Group: B/S, JK & SK Time: W/S, Time 1 & Time 2 Modality: W/S, PI & PN Language: W/S, Armenian & English Covariate: English Exposure/Experience
Language*Modality*Time*Group F(1, 39) = 5.765, p =.021, η2 =.13 M ea n % 100 90 80 70 60 50 40 30 20 10 0 Armenian Mean % 100 90 80 70 60 50 40 30 20 10 0 English Time 1 Time 2 Time 1 Time 2 JK PI JK PN SK PI SK PN JK PI JK PN SK PI SK PN
Discussion 1. SK > JK 2. Time 2 > Time 1 3. Picture Identification > Picture Naming Supported Developmental trends similar to monolinguals 4. Armenian > English 5. Armenian growth rate < English growth rate Not supported Additive bilingual environment Language dominance
Individual Differences Stepwise Hierarchical Regression Analyses Time 2 Armenian Vocabulary Time 2 English Vocabulary Time 1 Armenian Vocabulary IQ β =.872, p <.01, R square =.76 PSTM β =.256, p =.002, R square =.06 Parental Education Armenian Exposure/Experience Time 1 English Vocabulary β =.890, p <.01, R square =.79 IQ PSTM Parental Education β =.158, p =.039, R square =.02 English Exposure/Experience
Limitations High parental education and high IQ might indicate a select sample Language exposure measure Home language exposure well represented Community language exposure might be difficult to represent accurately
Contributions Group differences First study on Armenian-English bilingual kindergarteners Cross-language and cross-modality comparisons in vocabulary made possible Individual differences First study to explore several predictors in bilingual vocabulary growth
Contact: alice.hovsepian@utoronto.ca