Making Effective Decisions Grade 7 Example Answers (Civil Service)

Grade 7 is where the decision behaviour changes its physics. At every grade below, more evidence is better and thorough analysis is the virtue; the strongest answers gather, weigh, conclude. At G7 the descriptors quietly abandon that comfort: present reasonable conclusions from a wide range of complex and sometimes incomplete evidence, make decisions confidently even when details are unclear. The panel is scoring what you do when the analysis cannot finish — when the evaluation isn't done, the data conflicts, the pilot is half-run, and waiting for certainty carries a price someone else will pay. This is the uncertainty principle, and it produces the G7 answer's signature move: reasoning about the uncertainty itself. Not "I gathered what evidence I could" — but this was knowable, this wasn't, waiting cost this, so I decided on the following basis, watched these signals, and would have changed course if.

What G7 panels score on this behaviour

The incompleteness, characterised. What you knew, what you couldn't know in time, and why not. Candidates who can't describe the shape of their uncertainty haven't reasoned about it — they've just tolerated it.

The cost of waiting, priced. Deferral is a decision with consequences. Naming what delay would have cost — and to whom — is what makes deciding anyway judgement rather than impatience.

A basis for the conclusion. Under incomplete evidence, something still did the deciding: the weight of what was known, an explicit assumption, a reversibility argument, a worst-case comparison. The panel wants that mechanism visible.

Uncertainty managed forward. Leading indicators watched, review points set, the change-course condition stated in advance. Confidence at G7 is instrumented, not asserted.

The right level, held. The descriptors say decisions should happen at the right level — which cuts both ways: not escalated to dodge the discomfort, not hoarded when it was genuinely someone else's call.

Questions this appears as

• "Tell me about a decision you made with incomplete information."
• "Describe deciding between options when the evidence was unclear or conflicting."
• "Tell me about a significant decision you made under time pressure."
• "How have you handled a decision where waiting for more evidence wasn't an option?"

Probes to expect: "What would complete evidence have looked like?", "What did you assume, and what if the assumption was wrong?", "When would you have changed course?"

Fail vs pass at G7: the same funding call

At SEO this would score well — rigorous, structured, defensible. At G7 the panel notices what the thoroughness is hiding: ten weeks bought an analysis, and the answer never asks what the ten weeks cost — two delivery teams in limbo, contracts and staff decisions frozen, partners planning blind. More fundamentally, the framework is presented as having made the decision ("the framework pointed to..."), which is the G7 anti-pattern: methodology deployed as a shield against exercising judgement. Nothing here reasons about what was unknowable, assumes anything explicit, or manages uncertainty forward — because the answer's implicit claim is that enough process made the uncertainty go away. It didn't; it never does. A 4.

The annotation:

1. The incompleteness is characterised, not endured. Three years of data versus fourteen months, an evaluation eight months away — the shape of the uncertainty is mapped precisely. That map is the G7 evidence.
2. Waiting is priced. Notice periods, staff attrition, deferral-as-slow-failure — the cost of the ten weeks Version 1 spent is exactly what Version 2 refuses to spend. Four weeks, chosen and justified.
3. The basis is explicit, assumption included. Known-moderate versus unknown-plausibly-higher, decided on a named 70% conversion assumption. The decision mechanism is inspectable — which is what reasonable conclusions from incomplete evidence means in practice.
4. Uncertainty is engineered, not tolerated. The month-ten gate, pre-agreed thresholds, the £300k bounded exit — confidence with instrumentation. The worst case is designed before the best case is hoped for.
5. The right level, held under pressure. Declining the director's offer to take the decision upward is the decisions at the right level descriptor performed at cost — the sentence "splitting the decision from the reasoning would have served nobody except me" is G7 self-awareness the panel rarely hears.
6. Outcome and judgement, kept separate. "The decision was right when I made it or it wasn't" — the epistemically honest close that tells the panel this candidate's confidence comes from reasoning, not results. It's the highest-trust sentence in the answer.

Calibrating either side of G7

At SEO, Version 1's rigour — sharpened with invited challenge and a carried risk — is genuinely strong; the evidence there is mostly knowable and the virtue is weighing it well. At Deputy Director, the calibration adds strategic scope: bold decisions across political and national pressures. The full ladder: the Making Effective Decisions pillar.

G7-specific mistakes on this behaviour

1. Thoroughness as a shield. Ten weeks of process to avoid four weeks of judgement. Panels at G7 read exhaustive analysis of an unresolvable question as the discomfort dodged.
2. The framework decided. Methodology is how you reason, not who decides. "The scoring pointed to X" hands your marks to a spreadsheet.
3. Uncertainty unpriced. If your answer never states what waiting would have cost, the panel can't distinguish decisiveness from haste.
4. No explicit assumption. Every incomplete-evidence decision rests on one. Naming yours — and what would have falsified it — is the difference between judgement and hope.
5. Confidence without instruments. No gate, no indicators, no change-course condition: bravado. The G7 mark is bounded risk.
6. Vindication by outcome. Citing the result as proof the decision was right invites the obvious probe and fails it. Defend the reasoning; report the outcome separately.

If Making Effective Decisions is on your G7 advert — and it nearly always is — test your answer against the uncertainty principle before the panel does: the AI coach scores it against the real G7 descriptors and flags where the judgement is hiding behind the process. Three sessions free; the G7 Complete Guide runs all nine behaviours through this lens.