O’Connor, M., Geva, E., & Koh, P. W. (2019). Examining reading comprehension profiles of grade 5 monolinguals and English language learners through the lexical quality hypothesis lens. Journal of Learning Disabilities, 52(3), 232–246.
The lexical quality hypothesis posits that reading comprehension is dependent on the quality of the lexical representation of words (Perfetti, 2007; Perfetti & Hart, 2002). The quality of lexical representation is high when the phonological (sounds), orthographic (spelling), and semantic (meaning) information of a word are accessible to an individual allowing for quick and accurate retrieval of word meaning. In contrast, the quality of lexical representation is low when the reader is unable to either read the word accurately or is unclear of the meaning of the word. Lower quality lexical representation in any one or all of the three constituents (i.e., phonological, orthographic, semantic processing) can lead to poor understanding of the content being read.
Past studies have demonstrated support for the lexical quality hypothesis for native speakers of English, Hebrew, and German (e.g., Richter, Isberner, Naumann, & Neeb, 2013; Schiff, Schwartz-Nahshon, & Nagar, 2011). Results from more recent studies have shown that the lexical quality hypothesis framework also applies to second language learners in early elementary grades (e.g., Chung, Koh, Deacon, & Chen, 2017). However, no previous study has tested the lexical quality hypothesis framework for grade 5 English language learners (ELLs) and non-ELLs.
Thus, the focus of O’Connor, Geva, and Koh’s (2019) study was to analyze measures related to the three components of the lexical quality hypothesis to determine whether grade 5 ELLs and non-ELLs have similar reading profiles. In addition to accounting for the three components of the lexical quality hypothesis, O’Connor and colleagues took a novel approach by including a measure of listening comprehension into the model. One reason they did this is because listening comprehension has been observed to be one of the strongest predictors of reading comprehension in upper-elementary and later-grade second language learners (Saiegh-Haddad & Jayusy, 2016). The study addressed two research questions:
Data for the study were taken from a multicohort longitudinal study that collected information on students from grades 1–6 enrolled in a large metropolis in Canada. The authors used a subset of this dataset to conduct their analyses. The study sample included 272 grade 5 students, of which 178 were ELLs. The ELL sample was diverse with home languages including Portuguese (n = 59), a variety of South and East Asian languages (n = 116), and other languages (n = 3).
Individual measures representing each element of the lexical quality hypothesis and listening comprehension were administered to all students. Phonological awareness was measured using the auditory analysis skills task (Rosner & Simon 1971), orthographic processing was measured using the spelling subtest of the Wide Range Achievement Test–Revised (Wilkinson, 1993), and semantic knowledge was assessed using the Biemiller written root word inventory task (Biemiller & Slonim, 2001). Additionally, reading comprehension was assessed using the Gates-MacGinitie Reading Test (MacGinitie & MacGinitie, 1992) while listening comprehension was measured with the listening to paragraphs subtest from the Clinical Evaluation of Language Fundamentals–Third Edition (Semel, Wigg, & Secord, 1995). Finally, the authors also assessed students’ nonverbal ability using the Raven’s Standard Progressive Matrices (Raven, Raven, & Court, 1998) to ensure that both the ELL and non-ELL groups had similar analytical and reasoning skills.
Preliminary analyses were conducted using descriptive statistics to measure the magnitude of difference between ELLs and non-ELLs on all measures. To answer their research questions concerning the reading profiles of ELLs and non-ELLs, the authors conducted latent profile analysis (Bartholomew, 1987). This approach allows for more than one variable to be included in the model to determine subgroups that show similar response patterns across the five key variables: phonological awareness, orthographic processing, semantic knowledge, listening comprehension, and reading comprehension. The authors compared models with up to five subgroups for ELLs and non-ELLs to determine the best fit for each language group. It is important to note that the authors used raw scores rather than standardized scores for all analyses because the norms for all measures were based on non-ELL populations.
Results related to research question 1a: Is there evidence to support the lexical quality hypothesis in ELL and non-ELL grade 5 students?
In accordance with the lexical quality hypothesis, the sample categorized as good comprehenders scored significantly better on measures of phonological awareness, orthographic processing, and semantic knowledge compared to peers categorized as poor comprehenders. This trend was similar in both the ELL and non-ELL grade 5 student sample. Thus, this study provides further support for the lexical quality hypothesis specific to this population of students.
Results related to research question 1b: How well would ELL and non-ELL latent profile groups perform on measures of phonological, orthographic, and language comprehension skills?
Preliminary analysis of the data showed that non-ELLs in general outperformed their ELL peers on measures of semantic knowledge (d = .30) and reading comprehension (d = .25). The groups did not differ significantly on all other measures.
Latent profile analysis for ELLs and non-ELLs showed a two sub-group solution as the best fit. Authors identified the two subgroups as good comprehenders and poor comprehenders. A large majority of the ELL (80.9%) and non-ELL (70.2%) students in the sample were identified as good comprehenders. Good comprehenders had average to above-average scores on all three constituents of the lexical quality hypothesis and the reading comprehension measure. In contrast, 19.1% of the ELL sample and 29.8% of the non-ELL sample were identified as poor comprehenders who performed below-average on all reading measures. Notably, poor and good comprehenders did not differ significantly on the measure of listening comprehension.
Results related to research question 2. Are there differences in the emerging profiles of ELLs and non-ELLs?
There were no significant differences in the emerging profiles of ELLs and non-ELLs. Results based on this particular study’s sample showed two similar latent profiles for both subgroups. Good comprehenders in both subgroups performed above average on all reading measures. On the other hand, latent profiles of ELLs and non-ELLs identified as poor comprehenders generally showed deficits in areas of decoding, reading comprehension, and language-related skills. In summary, the study reports that grade 5 ELLs and non-ELLs who struggle with reading-related tasks have similar reading profiles and have deficits in multiple reading components.
One limitation of this study is that authors used only one measure to represent each reading construct. Past studies have shown that individual measures may tap into slightly different reading constructs and, therefore, recommended using latent variables to study associations (e.g., Cutting & Scarborough, 2006).
August, D., & Shanahan, T. (Eds.). (2006). Developing literacy in second-language learners: Report of the National Literacy Panel on Language—Minority children and youth. Mahwah, NJ: Erlbaum.
Bartholomew, D. J. (1987). Latent variable models and factor analysis. London, UK: Charles Griffin.
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(3), 498–520.
Chung, C. S., Koh, P. W., Deacon, H., & Chen, X. (2017). Learning to read in English and French: Emergent readers in French immersion. Topics in Language Disorders, 37(2), 136–153.
Cutting, L. E., & Scarborough, H. S. (2006). Prediction of reading comprehension: Relative contributions of word recognition, language proficiency, and other cognitive skills can depend on how comprehension is measured. Scientific Studies of Reading, 10(3), 277–299.
MacGinitie, W. H., & MacGinitie, R. K. (1992). Gates-MacGinitie reading tests (2nd Canadian ed.). Toronto, ON, Canada: Nelson.
Perfetti, C. (2007). Reading ability: Lexical quality to comprehension. Scientific Studies of Reading, 11(4), 357–383.
Perfetti, C. A., & Hart, L. (2002). The lexical quality hypothesis. Precursors of Functional Literacy, 11, 67–86.
Raven, J., Raven, J. C., & Court, J. H. (1998). Manual for Raven’s progressive matrices and vocabulary scales, section 1: General overview. San Antonio, TX: Harcourt Assessment.
Richter, T., Isberner, M. B., Naumann, J., & Neeb, Y. (2013). Lexical quality and reading comprehension in primary school children. Scientific Studies of Reading, 17(6), 415–434.
Rosner, J., & Simon, D. P. (1971). The auditory analysis test: An initial report. Journal of Learning Disabilities, 4, 383–392.
Saiegh-Haddad, E., & Jayusy, A. (2016). Metalinguistic awareness in reading Hebrew L2. Acquisition and Development of Hebrew: From Infancy to Adolescence, 19, 353–386.
Schiff, R., Schwartz-Nahshon, S., & Nagar, R. (2011). Effect of phonological and morphological awareness on reading comprehension in Hebrew-speaking adolescents with reading disabilities. Annals of Dyslexia, 61(1), 44–63.
Semel, E., Wigg, E. H., & Secord, W. A. (1995). The clinical evaluation of language fundamentals (3rd ed.). San Antonio, TX: Psychological Corporation.
Wilkinson, G. S. (1993). Wide Range Achievement Test–Revised (WRAT 3-R, 3rd ed.). Wilmington, DE: Wide Range.