Syntactic Complexity and Proficiency: A Study of Objective Measures in the TEEP Speaking Test

Syntactic complexity plays a critical role in assessing language proficiency, yet identifying measures that effectively distinguish proficiency levels remains challenging. This study investigates the relationship between objective measures of syntactic complexity and subjective ratings in the Test of English for Educational Purposes speaking assessment. Focusing on three metrics Mean Length of AS-Unit, Mean Length of Clause, and Subordinate Clauses per AS-Unit the research evaluates their ability to correlate with subjective ratings of grammatical range and accuracy and differentiate proficiency levels. Using a quantitative approach, the study analyzed 89 TEEP speaking test transcriptions paired with proficiency scores ranging from 5.0 to 7.5. Statistical analyses revealed weak but significant positive correlations between MLAS and MLC with subjective ratings, suggesting that these measures contribute modestly to evaluative outcomes. However, SCP-AS showed no significant correlation. ANOVA results highlighted that while MLAS distinguished between lower and higher proficiency levels, it struggled to differentiate adjacent levels. MLC and SCP-AS showed limited variation across levels, indicating low sensitivity. The findings suggest refining TEEP rating scales to assess complexity and accuracy separately and incorporating diverse syntactic, lexical, and morphological measures to enhance assessment validity. Future research should expand metrics and task types to deepen understanding of syntactic complexity in language assessment contexts.