Why the four mechanisms your own log could report on directly turn into open, sometimes unanswerable questions once you're standing outside someone else's wait — and the two separate questions, kept apart, that let you say "cannot determine" or "undetermined" honestly instead of guessing.
A café chalkboard reads: “Orders usually ready in 10 minutes.” You can read it, note it down, and move on — that much is right there in front of you. What you can’t do is ask the next customer what they actually expected walking in, or whether that sign is what set it. Nothing about this week’s mechanism has changed since Week 3. What’s changed is your position: you’re no longer the person in the queue, you’re the person looking at one from outside.
This week’s claim, plainly#
The four mechanisms your own log could report on directly become open, sometimes unanswerable questions once you stand outside someone else’s wait; learning to say “undetermined” honestly is the skill this week teaches.
Weeks 1 to 4 built time, attention, expectation, and information entirely out of your own log — you were both the observer and the person whose experience was being recorded, so every field had a direct answer available just by asking yourself. Weeks 5 to 9 taught five different mechanisms — interface, environment, queue structure, priority, value — that are externally observable by design: a queue’s posted rule or a room’s layout doesn’t require access to anyone’s mind. This week is where those two halves of the course meet a genuine problem: the Queue Case Study asked you to classify the second five from outside a system, but it never asked you to say how strongly your evidence actually supported any claim about effect. And the first four have never been asked of an external system at all. Both gaps close here, on the one system you already know.
Two axes, kept apart#
- Mechanism status — present / absent / ambiguous / cannot determine. This is the question your Queue Case Study already asked, for a different five mechanisms: what does the accessible evidence indicate is actually in the system?
- Evidence tier — directly evidenced / plausible interpretation / undetermined. This is a different question, asked for the first time this week: however I’ve classified the mechanism, how strongly does what I actually have support the specific claim I want to make about it?
These are not two names for the same judgement, and they never combine into one score. A mechanism’s status can be settled with confidence from a single observable feature. A claim about that mechanism’s effect on anyone’s experience usually can’t be, because that would require exactly the kind of access — asking the person directly, running a controlled test — that looking at a system from outside doesn’t give you.
Reading absence and uncertainty correctly#
The evidence tier attaches to the specific claim you’re making, not to whether the mechanism is present. Two priority findings, both from a published policy, show how differently that can land:
Example A. “The published policy states no priority criterion applies here.” Mechanism status: absent. Evidence tier: directly evidenced — the policy says so, in writing, and you can point to exactly where.
Example B. “No priority mechanism could be identified from what’s published.” Mechanism status: cannot determine. Evidence tier: undetermined — not because nothing was found, but because the available evidence isn’t sufficient to establish whether an unstated mechanism exists at all. A silence in a published policy is not the same kind of evidence as a policy that explicitly rules something out.
Both are legitimate, complete findings. Neither is a weaker answer than “present,” and neither should be written up as if it were a guess.
The five you already have#
Your Queue Case Study gave interface, environment, queue structure, priority, and value each a status finding — present, absent, ambiguous, or cannot determine — on your own system. It did not give any of them an evidence tier, because that question hadn’t been taught yet. This week doesn’t ask you to redo that analysis. It asks you to go back to each of those five findings and add the second axis: for whatever claim you made about what that mechanism does, how strongly does your original evidence actually support it? A status finding you’re confident in can still carry a claim about effect that only earns “plausible interpretation” — that combination is common and correct, not a sign your Case Study was weak.
The four you’re building from scratch#
Time, attention, expectation, and information have never been classified on an external system at all — only on your own log. Doing it now means applying both axes, not just the one your Case Study used:
- Time. A posted schedule, or a duration you time yourself, gives you a directly observable clock duration. Whether anyone standing in that system judged it as longer or shorter than that number is not observable — you have no channel to their estimate.
- Attention. A screen, seating, a view, silence, or nothing at all is a directly observable feature of the setting. Whether it actually occupied anyone’s attention, or moved anyone’s judged duration the way Week 2’s sources describe, stays at most a plausible interpretation.
- Information. A display, an announcement, or their absence is directly observable. Their effect on anyone’s evaluation of the wait is a separate question this week doesn’t try to settle — it’s the same mechanism Week 4 named, just no longer measured on yourself.
- Expectation — this week’s fully worked case, below.
Expectation, worked in full#
Back to the chalkboard: “Orders usually ready in 10 minutes.”
- Expectation-setting feature: present. The sign exists, says a number, and you can quote it exactly. Not in question.
- A specific customer’s actual expectation: cannot determine. The sign tells you what the system invites people to expect. It tells you nothing about what any particular person walking up to the counter actually believed — that would need asking them, which this kind of observation doesn’t permit.
- Claim: “the sign shaped that customer’s expectation”: plausible interpretation, unless direct evidence exists. It’s a reasonable inference — posted estimates are a standard way of setting a reference point, and Week 3’s Pruyn and Smidts material supports that kind of reference point mattering in general — but nothing here confirms it for any actual customer at this café. The sign being there is never, on its own, evidence that anyone’s internal expectation was actually set by it.
What Weeks 1 to 4’s sources do and don’t give you here#
Block and Zakay’s prospective/retrospective finding, Block, Hancock and Zakay’s cognitive-load finding, Pruyn and Smidts’ acceptable-waiting-time finding, and Hui and Tse’s information finding were all built the same way: someone asked the person doing the waiting, directly, under a controlled protocol (Block & Zakay, 1997; Block, Hancock & Zakay, 2010; Pruyn & Smidts, 1998; Hui & Tse, 1996). That’s exactly the channel you don’t have when you’re reading a system from outside without interviewing anyone in it. These four sources aren’t being reapplied to your café or your Case Study system as if their findings transferred automatically — they’re evidence for why “undetermined” is the honest answer to a claim about effect, not evidence about your particular system at all.
Why this matters#
Every mechanism from Week 1 onward has, until now, been measured either on yourself or on a system whose structure doesn’t require anyone’s inner state to read. This week is where those two kinds of evidence have to sit side by side on the same system, honestly tiered. Control is not part of this week’s table at all — it’s supplied separately by your Week 10 experiment, tiered according to how strong that specific design actually was, and carried into the Audit on its own terms rather than folded into the nine mechanisms here. By the end of this week you hold nine mechanisms, each with a status and a tier; Week 12 adds the tenth and asks for the whole system at once, under the same thesis this course opened with — waiting is an experience, not merely a duration.
What you’re doing this week#
The session has you build one working table: your five Case Study findings, re-tiered: your four remaining mechanisms, classified on both axes for the first time, on the same system. That table is not a rehearsal for the Waiting Audit — it’s the first nine rows of it.
Teaching team
- Marisol Quaye — convenor