Supplementary materials for the journal article Structural Factor Analysis of Lexical Complexity Constructs and Measures:
Nasseri, M., & McCarthy, P. (2023). Structural Factor Analysis of Lexical Complexity Constructs and Measures: A Quantitative Measure-testing Process on Specialised Academic Texts. Journal of Quantitative Linguistics, 2023. Link (If you need access to this research, please send a "full-text request" via Research Gate and I'll send it to you :)
-
section A: Link access to the code/analysers
-
Section B: Some additional explanation regarding the measure-selection process
-
Section C: L1 backgrounds and ethnicities of EFL and ESL groups
-
Section D: Structural diagrams from the CFA tests along with factor loadings
- Diagram of the structure of lexical constructs and measures with factor loadings for the lexical model A
- Diagram of the structure of lexical constructs and measures with factor loadings for the lexical model A1
- Diagram of the structure of lexical constructs and measures with factor loadings for the lexical model A2
- Diagram of the structure of lexical constructs and measures with factor loadings for the lexical model B
- Diagram of the structure of lexical constructs and measures with factor loadings for the lexical model B2
IN SFA a series of factor analyses are used to verify the structure of constructs and how well the variables represent each construct based on the proposed structures and to further explore the datasets for any additional construct/latent factor that can be revealed. This is usually done based on McArdle's (2011) proposed scheme of first conducting a Confirmatory Factor Analysis (CFA) followed by an Exploratory Factor Analysis (EFA).
Confirmatory Factor Analysis (CFA) is a statistical method used to verify the factor structure of observed variables in order to test whether there exists a relationship between observed variables and their underlying latent constructs (i.e., the overall conceptual understanding of a phenomenon with verifiable/quantifiable variables).
EFA is a statistical method used to explore a large dataset (observed data) and find its underlying structure and the relationship between quantifiable variables. The aim is to find constructs that match/approximate the observed phenomena.
The conceptual representation of lexis (words, collocations, formulaic expressions, etc) that can be measured via objective measures or indices, also called lexical measures.
In this research, lexical complexity measures are quantifiable indices that compute one of the three constructs of lexical density, diversity, and sophistication. Each of these constructs is usually represented by a variety of different measures/indices in the related literature, but usually, some of these measures are more suitable for specific areas than others; e.g., some lexical measures can better distinguish between higher proficiency levels, while others may be more suitable for child language acquisition and development, or the diagnosis of the onset of dementia, etc. Please see the literature review part for more details.