A small, ethical, deliberate manipulation of one named waiting mechanism — usually control, this course's Week 10 mechanism — tied to a gap your Case Study or a cited source actually left open.
- Waiting Field Notes15% · due Week 4
- Queue Case Study20% · due Week 9
- Waiting Experiment25% · due Week 10
- Waiting Audit40% · due Week 12
Due: 26 April 2027 · Weight: 25%
The brief#
Design and run one small, ethical, deliberate manipulation of a single named waiting mechanism, and report honestly what it does and doesn’t show.
Everything before this assessment has been observation — of your own waits, then of a system you didn’t build. This is the course’s one intervention: you choose something, change it on purpose, and see what happens. It is also the smallest-scale piece of work in the course, and the brief expects you to know that going in — a bounded exercise like this can suggest something worth reporting; it cannot prove very much on its own.
Starting from a gap, not a hunch#
Name your mechanism before you touch any data, and say why it’s the right one to test. There are two legitimate starting points:
- Week 10’s mechanism: control. This week’s teaching is about the sense of control a person has over a wait — whether they can choose when to join, how it proceeds, or whether to leave it. Most students should start here: pick a control-related manipulation (offering a choice over queue position or pacing, versus not) and test it directly.
- A gap your own Queue Case Study left open, or a specific limitation of a named course source. If your Case Study marked a mechanism ambiguous or undeterminable from published information, or if a source this course has cited (Hui & Tse, Pruyn & Smidts, Dietl, Voigt & Kuhn, or any other) explicitly didn’t measure something your own question needs, you may test that mechanism instead — but you must name the exact gap and why it’s worth closing.
A submission that tests an arbitrary mechanism with no stated connection to either path hasn’t met the brief, however well the experiment itself is run.
Running the experiment#
Default to self-observation, a simulation, or a controlled prototype — the same default the Experiment ethics policy sets for this assessment and for the Waiting Audit. A live intervention on a real waiting system is only permitted where every condition in that policy is met: genuinely low-risk, non-disruptive, privacy-preserving, authorised wherever authorisation is needed, and never disrupting an actual clinic, emergency-service, or similarly essential queue.
What you submit#
In order: the protocol, written before you collected any data (mechanism and gap, manipulation, measure, prediction); what you actually observed; the result, including a null, mixed, or contrary-to-prediction outcome reported exactly as it came out; a bounded interpretation of what that result does and doesn’t show; and a limitations section naming the confounds a small, largely uncontrolled exercise can’t rule out.
What this can’t show#
A handful of trials, run by one person, is not a controlled study. It cannot establish strong causal proof, and it cannot be generalised beyond the specific context you tested it in. If your result doesn’t match a cited study’s finding, that’s a limitation of your own small exercise — not evidence that the cited research is wrong.
Afterwards#
The Waiting Audit, due at the end of Week 12, may use this experiment’s result as evidence for one mechanism in the system you audit — but only where it’s actually relevant to that system, only for the one mechanism you tested, and only as strongly as your own design here actually supports.
The spec
The brief poses the problem and leaves room for a response. The spec is the fixed part: what a marker checks. Some lines can be checked mechanically; the rest call for judgement.
- you have named, before collecting any data, the single mechanism you are testing and why — Week 10's control mechanism by default, an ambiguous or cannot-determine finding from your own Queue Case Study, or a specific, statable limitation of a named course source
- you have written a protocol before running the experiment, stating the manipulation, what changes between conditions, the measure, and your prediction
- you have tested the manipulation by self-observation, a simulation, or a controlled prototype; any live intervention on a real system meets every condition in the existing experiment-ethics policy
- you have reported what actually happened, including a null, mixed, or contrary-to-prediction result, without revising the prediction after the fact
- you have stated plainly what your result does and does not establish, and named the confounds and limitations of a small, largely uncontrolled exercise
- you have not claimed strong causal proof, generalised beyond the tested context, or treated a personal null result as evidence against a cited study
- you have tested exactly one named mechanism, not several at once
How it is marked
| Criterion | Weight |
|---|---|
| Precision of the experimental design and stated prediction | 30% |
| Grounding in Week 10's mechanism or an identified gap | 20% |
| Honest reporting of results, including null or mixed outcomes | 25% |
| Bounded interpretation and limitations | 25% |