Most teachers circulate during pair work without a clear focus. They drift from pair to pair, listen for a few seconds, move on. They pick up a general impression - "it went OK" or "they seemed engaged" - but no specific, actionable information about what students can and can't do. By the next lesson, the observation has faded entirely.
The problem is not the circulating. The problem is the absence of a listening framework. Without knowing what to listen for, observation produces impressions rather than data. With a clear focus, even two minutes of targeted listening per lesson generates the specific information that drives better teaching decisions.
YapYapGo is a classroom speaking practice tool for ESL and EFL teachers that frees you from logistics during pair activities - pairing, question delivery, timing are all handled automatically - so your full attention is available for observation. Here's what to do with that attention.The five dimensions of speaking to monitor
1. Fluency
What you're listening for: pauses, hesitations, filled pauses (um, er, uh, like), false starts, speech rate. Indicators of good fluency: speech flows without long pauses, ideas connect smoothly, the student fills their speaking time rather than stopping and waiting. Indicators of developing fluency: frequent pauses of more than two seconds, high number of filled pauses, sentences that restart mid-way, speech that stops before the time limit. What to note: which students have high pause rates? Where do pauses cluster - at the beginning of sentences (accessing the topic), mid-sentence (grammatical retrieval), or before specific vocabulary items? What this tells you: fluency blockers are usually vocabulary gaps (can't find the word), grammar uncertainty (don't know the right form), or topic unfamiliarity (don't have enough content). Pauses at the start of most utterances suggest topic or content problems. Pauses mid-sentence suggest grammar. Pauses before specific words suggest vocabulary.2. Vocabulary
What you're listening for: range (do they use varied words or repeat the same limited set?), precision (are words used accurately?), active use of lesson vocabulary. Indicators of good vocabulary use: students use words beyond the most common, choose precise words for precise meanings, incorporate vocabulary from recent lessons naturally. Indicators of developing vocabulary: same words repeated frequently, vague language where precise words are needed ("that thing" instead of a specific noun, "good" where "efficient" or "compelling" would be accurate), avoidance of topics where vocabulary is limited. What to note: three or four specific vocabulary gaps you heard. Phrases students needed that they didn't have. Words used incorrectly. What this tells you: reactive vocabulary teaching after the activity - "I heard three pairs struggle to express X. Here are two phrases that would have helped" - directly addresses observed needs.3. Grammar
What you're listening for: accuracy with the structures the lesson targeted, range of grammatical structures used. Important caveat: during fluency activities, grammar should not be the primary focus of observation. Note patterns rather than individual errors. Address in the debrief, not during the activity. What to note: patterns that appear across multiple pairs (three pairs made the same error with present perfect), not individual errors ("Maria made an error in sentence three"). What this tells you: patterns indicate gaps in acquisition that require systematic teaching, not one-off correction.Tool tip: YapYapGo runs the pair activity so you never need to watch the clock or signal transitions. Your clipboard goes with you as you circulate. A class timer visible to all students means they know when the round ends - you don't have to manage it.
4. Communication effectiveness
What you're listening for: does meaning get across? Do students understand each other? Do they respond to what was actually said rather than delivering prepared responses? Indicators of effective communication: listener asks follow-up questions that show genuine comprehension, speaker adapts when misunderstood, both students are responding to each other rather than talking in parallel. Indicators of ineffective communication: listener asks questions that ignore what was just said (suggesting they didn't follow), speaker continues without noticing their partner's confusion, communication breaks down and neither student repairs it. What to note: moments where communication broke down and what caused it. Moments where students successfully repaired a breakdown. What this tells you: communication effectiveness is a better predictor of real-world English success than grammatical accuracy. Students who cannot repair communication breakdowns are less prepared for real communication than students who can.5. Interaction quality
What you're listening for: turn distribution (equal or dominated by one student?), question generation (who asks questions?), genuine engagement (are they listening and responding, or just waiting for their turn?). Indicators of good interaction quality: turns are roughly balanced, both students generate questions, students respond to what their partner specifically said rather than delivering pre-planned material. Indicators of poor interaction quality: one student speaks 80% of the time, one student answers but never asks, students deliver monologues rather than having a conversation, neither student follows up on interesting statements from the other. What to note: which pairs have dramatically unequal participation. Which students always answer but never initiate. What this tells you: interaction quality problems often reflect pairing issues (a very confident student with a very anxious one), personality dynamics, or a lack of specific question-asking practice. These need structural solutions, not language teaching.The listening rotation
You cannot focus on all five dimensions simultaneously. Rotate your focus across lessons:
- Monday: Fluency (pauses, speech rate)
- Wednesday: Vocabulary (range, gaps, active use)
- Friday: Interaction quality (turn distribution, question generation)
Assess grammar through written work rather than real-time observation during communicative activities.
Using what you've heard
The observation is worthless if it doesn't change anything. A simple structure:
Celebrate: "I heard someone say 'on the other hand' very naturally - that's exactly the kind of connective that makes arguments sound more sophisticated in English." Address a gap: "I noticed several pairs couldn't express the idea of something happening gradually. Here's a phrase that would have helped: 'over time' or 'progressively.'" Name a pattern: "Three pairs made the same error with the present perfect - this tells me we need to revisit it next lesson." Adjust pairings: "I noticed some pairs had very unequal participation today. I'm going to adjust the pairings for next time to mix the groups differently."For more on assessment approaches that go beyond in-class observation, see our posts on formative assessment of speaking and peer assessment for speaking. A random student picker is useful during the debrief when you want to call on specific students. An activity timer labelled with each observation focus reminds you which dimension you're tracking that round.
Sources:
- Luoma, S. (2004). Assessing Speaking. Cambridge University Press. - Dimensions of spoken language and what they indicate about development.
- Skehan, P. (1998). A Cognitive Approach to Language Learning. Oxford University Press. - Fluency, accuracy, and complexity as separable and independently observable dimensions.
- Black, P. & Wiliam, D. (1998). Assessment and Classroom Learning. Assessment in Education. - Observation as formative assessment: the conditions for useful data collection.