Lesson
Near-synonym traps: hear the distinction that decides the answer
Goal: Distinguish related terms when a speaker explicitly contrasts them · Track proposals, objections and final recommendations across a discussion · Match paraphrases by meaning and scope rather than shared vocabulary
- At high accuracy levels, distractors often belong to the right topic but describe the wrong property. Affordable concerns cost; feasible concerns whether a plan can be carried out. Balanced groups are not automatically representative of a population. A confidential interview is not necessarily anonymous. Listen for the speaker's local definition and the decision it supports, rather than matching one familiar word to an option.
- Make a brief distinction when two near-synonyms appear: what dimension does each describe, and which one answers the question? The same speaker can accept that a plan is economical while rejecting it as impractical. Another speaker may propose a word and then withdraw it after clarification. Keep the final, qualified position instead of accumulating every positive-sounding term as an answer.
- Use this discussion first as partner-read listening with the script hidden. If no spoken version is available, use it as transcript analysis and do not treat the result as evidence of unaided listening. In review, quote the short phrase that excludes the nearest distractor. Replacing the target word with a plain explanation is often a better check of understanding than memorising a synonym list.
- Recording notes: Mina, female, northern English accent; Joel, male, standard southern British accent; Tutor, female, New Zealand English accent. Read at a natural practice pace of roughly 155–170 words a minute, allowing genuine turn-taking but no extra pauses at answers. Keep the terminology contrasts clear through meaning, without exaggerating individual words. All people and the project are fictional.
- Mina: We have brought the revised plan for our study of the Rellin repair hub. We originally wanted to observe every weekend session for two months. The room itself is free, so we described that as an affordable method. But the hub's insurer requires a supervisor to be present whenever visitors observe repairs, and only one supervisor has agreed to help us. She can attend two sessions, not eight. That leaves us wondering whether the design can actually be carried out.
- Tutor: Exactly. Cost and feasibility are separate questions. Your observation plan may be affordable, but it is not feasible in its current form because you cannot obtain the required supervision. You could observe two sessions as a small supplementary activity. For the main study, interviews look more manageable. However, do not describe people who meet your entry conditions as already selected. They are eligible; selection happens when you choose whom to invite from that group.
- Joel: Our draft invitation list includes equal numbers from three age groups. We thought that made it representative, but the hub's membership is mostly older adults. We also drew every name from current members. Former members and people who have considered joining are absent. So equal numbers would make the groups balanced in size, but the sample would still fail to reflect the wider range of people whose experiences we want to understand.
- Tutor: That is the distinction. You might deliberately balance the groups to compare their answers, provided you explain the choice. Do not then claim that their proportions reproduce the population. The more immediate weakness is recruitment: restricting invitations to current members means you cannot investigate why others left or decided against joining. A more attractive poster would not solve that problem if it were still distributed only inside the hub.
- Mina: We have also promised anonymous interviews. I now think that is too strong, because we will know who is speaking and the recordings will contain recognisable voices. We can promise confidentiality, restrict access to the recordings and remove identifying details from the transcripts. But those measures do not make the interview itself anonymous to us. We need to revise the information sheet so that participants understand what we can actually protect.
- Tutor: Yes, and keep your procedure consistent across interviews. That does not require identical answers or mechanical delivery of every follow-up. It means that comparable issues are explored and that decisions about prompting are made on a common basis. You can clarify an unclear response, but you should not lead one participant towards a complaint while encouraging another to give only positive examples. Consistency concerns the approach, not uniformity in what participants say.
- Joel: In our two pilot interviews, participants gave similar ratings when we repeated one question. We initially called that proof that the question was valid. But it may only show that the rating is repeatable. If the question is really measuring convenience rather than confidence in doing repairs, a stable answer could still measure the wrong thing. We have therefore added a follow-up asking participants what they had in mind when choosing the rating.
- Tutor: Good. Repeatability and whether you are measuring the intended experience need separate checks. Also, your proposed interpretation should remain tentative. That does not mean you have no evidence: several pilot comments point towards uncertainty about tools. It means you have a plausible working account that must be tested against a wider set of responses, including those that do not fit it. An early pattern can guide attention without deciding the final conclusion.
- Mina: One last issue is length. The pilot interviews ran for nearly fifty minutes. We wondered whether to conduct fewer interviews to make the project manageable, but that would narrow the range of experiences. Instead, we could remove repeated questions and bring each interview down to thirty minutes. The core topics would remain; we would stop asking the same thing in slightly different language. That seems a better way to reduce the burden.
- Tutor: I agree. Keep the planned number of interviews and shorten each one by removing repetition. First revise the information sheet, then run two further pilot interviews with the revised questions, and only after that finalise recruitment for the main study. Your plan is much more precise now. Its strength is not that every difficulty has disappeared, but that you distinguish what is affordable, what is workable and what the resulting evidence can legitimately support.
A worked example
BeforeDiscussion: 'The plan may be affordable, but it is not feasible ... because you cannot obtain the required supervision.' Wrong answer to 'What blocks the plan?': cost.
AfterThe blocking issue is the availability of required supervision, which makes the current plan infeasible despite its low cost.
'Affordable' acknowledges a favourable property; 'but it is not feasible' changes the evaluative dimension. The causal clause identifies the actual obstacle.
Use it elsewhere
Use a new spoken discussion containing corrections or qualified agreement. For five answers, explain both why the chosen paraphrase fits and why a near-synonym does not. Keep the explanation tied to the speaker's meaning, not a general dictionary distinction.