Position Paper: Classify a Relationship
Classify one relationship (yours, a friend's, or a plausible composite) against the model's first five variables and defend the classification
Due: 22 March 2027 · Weight: 25%
The brief#
Classify one relationship against the model’s first five variables, and show your working for each.
Weeks 1 through 5 gave you five variables: the two-party framing, the attraction front-end/back-end split, the reliability rate, the five-line-item dependency balance sheet, and the fungible/specific provision split. This paper asks you to apply all five to one case, in order, and to defend each classification with something a reader could check independently — a computed rate, a listed line item score, a named substitute that either does or doesn’t exist.
The provocation is not “write about a relationship.” It’s “show that a stranger, reading only your evidence, would reach the same classification you did.” A position paper that reports conclusions without the working behind them will not clear the model-application criterion, however correct the conclusions turn out to be.
What you submit#
A single PDF, 1200–1800 words, structured variable by variable (five sections, one per variable, each stating the classification and the evidence for it). De-identify or composite any third party’s private details — the model doesn’t need a true story, it needs a checkable one.
Model application (60%) is marked on whether each of the five variables is applied correctly and specifically to the case. Clarity of reasoning (40%) is marked on whether a reader unfamiliar with the case could follow the argument from evidence to classification without needing the writer present to explain it.
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.
- submitted by the deadline, as a PDF, 1200-1800 words
- names a specific case (with any third-party details de-identified or composited) and classifies it against framing, attraction, reliability, dependency, and specificity
- every classification is backed by a specific, checkable piece of evidence, not by a general impression
How it is marked
| Criterion | Weight |
|---|---|
| Model application | 60% |
| Clarity of reasoning | 40% |