Skip to main content
A Singapore Government Agency Website How to identify
Official website links end with .gov.sg
Secure websites use HTTPS
Look for a lock () or https:// as an added precaution. Share sensitive information only on official, secure websites.
AST | Elis Home

IRF exchanges in NT and SPE classes

Explore how classroom talk differs between NT and SPE streams. This study found NT lessons had lower-quality instruction, possibly due to reduced teacher expectations.

Last updated 8 September 2026

Hui, C. (2008). A corpus-based analysis of classroom interaction in Singapore secondary schools: The role of IRF exchanges (Master’s thesis). Retrieved from http://libris.nie.edu.sg/

Background

Findings from other countries have shown classroom instruction tends to differ in quality between low and high ability groups.

Aims

This study used corpus analysis to examine variation in the quality of teachers’ management and instructional discourse in Normal (Technical) (NT) and Special (SPE) streams in Singapore secondary schools.

Sample

Eighteen lessons (four English, two Maths, and three Science lessons each from NT and SPE streams) from the Singapore Corpus of Research in Education (SCoRE). A total of 952.4 minutes of classroom data was studied.

Method

The lessons were analysed based on comparisons between the qualities of Initiation-Response-Follow-up (IRF) exchanges in both streams.

Findings

The findings showed IRF exchanges took up a higher percentage of classroom talk in NT lessons (41.1%) than in SPE lessons (28.0%). However, the average word count for each IRF exchanges was higher in SPE lessons (47.3) than in NT lessons (29.3), suggesting that pace of interaction was higher in the latter.

Average student response was longer in SPE lessons (6.7 words) than NT (2.8 words). Both factual and clarification responses were common in both NT (85.3%, 8.34%) and SPE (74.9%, 9.21%). However, SPE lessons contained higher order responses such as elaboration (9.63%) and disagreement (2.09%), whereas NT lessons contained comparatively fewer of them and more of lower order responses such as avoidance (0.68%).

Uptake questions constituted the biggest category of follow-up moves in both NT (23.9%) and SPE (33.3%) streams. However, closed and display questions were significantly more common in NT (98.9%, 87.6%) than in SPE (59.2%, 54.8%) lessons, even though both formed the majority of uptake questions.

In terms of classroom management discourse, regulative talk appeared at significantly higher rates in NT (26.4 times/hour) than in SPE (3.3 times/hour) lessons.

Implications

The author concluded that the lower quality of instructional discourse in NT lessons might be because of the lowered expectations from teachers. In the long run, this could deprive students of productive and higher order learning. The author thought teachers should reflect on their teaching to improve instructional quality.

Original article retrieved from NIE Digital Repository (opens in new tab).