Chapter 5 — Human factors and cyber policy: design alternatives
Lesson map
Lesson 05 published. Map, example, EX021.
flowchart TD
N0["Learning outcomes"]
N1["5.1 Alternatives"]
N2["5.2 Evaluating"]
N3["Lab — Weighted scores and sensitivity for three desi"]
N4["5.3 Selecting"]
N5["Synthesis"]
N0 --> N1 --> N2 --> N3 --> N4 --> N5
N5 --> EX["Worked example"]
EX --> Q["Exercise EX021"]
Logical drawing
flowchart LR
N0["Learning outcomes"]
N1["5.1 Alternatives"]
N2["5.2 Evaluating"]
N3["Lab — Weighted scores and sensitivity for three desi"]
N4["5.3 Selecting"]
N5["Synthesis"]
N0 --> N1 --> N2 --> N3 --> N4 --> N5 --> Q["EX021"]
[Learning outcomes]
|
v
[5.1 Alternatives]
|
v
[5.2 Evaluating]
|
v
[Lab — Weighted scores and sensitivity for three desi]
|
v
[5.3 Selecting]
|
v
[Synthesis]
|
v
[Worked example] --> [Exercise EX021]
Learning outcomes
You will (1) generate alternative programme designs; (2) evaluate them against multiple criteria; (3) select and justify a design.
5.1 Alternatives
Daniel generates three designs for Calloway. Design A refines the current model with better simulations and shorter training. Design B centres on reporting: a prominent button, fast feedback, a no-blame policy and simulations reduced to quarterly under the new governance. Design C centres on engineering: phishing-resistant authentication for all staff, payment verification controls and strong filtering, with minimal training. He describes each in a page, including what staff would experience day to day, so stakeholders can picture the change. Stakeholders read all three.
5.2 Evaluating
Daniel scores the designs with stakeholders on risk reduction, staff burden, cost, trust and regulatory evidence. Weights are agreed before scoring. The lab computes weighted scores and tests whether the ranking changes when each weight is varied by half in either direction. Design C scores highest for risk reduction but lowest for regulatory evidence; Design B balances well. A combination of B and C emerges as the strongest option across most weight settings, which gives the board confidence that the result is not fragile.
Lab — Weighted scores and sensitivity for three designs
# Weighted scores and sensitivity for three designs plus a combination (synthetic 1-5 scores).
crit = ["risk", "burden", "cost", "trust", "evidence"]
w0 = dict(risk=0.30, burden=0.20, cost=0.15, trust=0.20, evidence=0.15)
S = {"A": [2, 2, 4, 2, 5], "B": [4, 4, 4, 5, 3], "C": [5, 4, 2, 3, 2], "B+C": [5, 4, 3, 5, 4]}
def score(w): return {k: sum(w[c] * v for c, v in zip(crit, vals)) / sum(w.values()) for k, vals in S.items()}
print("base:", {k: round(v, 2) for k, v in score(w0).items()})
wins = 0
for c in crit:
for f in (0.5, 1.5):
s = score(dict(w0, **{c: w0[c] * f})); wins += max(s, key=s.get) == "B+C"
print("B+C ranked first in", wins, "of 10 variations")
Run with python3 lab_weights.py. Verified output:
base: {'A': 2.75, 'B': 4.05, 'C': 3.5, 'B+C': 4.35}
B+C ranked first in 10 of 10 variations
5.3 Selecting
Daniel recommends a combined design: the engineering controls of C with the reporting culture of B, retaining brief role-based training to satisfy regulators. He explains the trade-off: higher first-year cost for lasting reductions in risk and burden. He notes risks of the combination, such as project overload, and proposes phasing over eighteen months. The board accepts, asking for a progress review at six months. Selection based on explicit criteria makes the decision easier to defend and to revisit. The phasing plan is agreed.
Worked example
Daniel designs three alternatives, scores them with stakeholder weights, tests sensitivity and recommends a phased combination.
Synthetic case. Designs and scores are synthetic.
Errors and limits
Scores are subjective. Combined designs may overload teams.
Exercises
- EX021 — Design C weakness. Where does Design C score lowest? Rubric: 3 = regulatory evidence;; 2 = evidence;; 1 = risk;; 0 = none.
- EX022 — Weight timing. When were weights agreed? Rubric: 3 = before scoring;; 2 = before;; 1 = after;; 0 = none.
- EX023 — Lab sensitivity. How robust was the combined design in the lab? Rubric: 3 = first in 10 of 10 weight variations;; 2 = robust;; 1 = vague;; 0 = none.
- EX024 — Phasing. Over what period is the combined design phased? Rubric: 3 = eighteen months;; 2 = 18;; 1 = vague;; 0 = none.
- EX025 — Combined risk. What risk does the combination carry? Rubric: 3 = project overload;; 2 = overload;; 1 = vague;; 0 = none.
Synthesis
Alternatives evaluated on agreed weights, with sensitivity tests, support robust selection.
Continue studying — chapter 05 of 12
← Chapter 04 (draft) · Course chapter index · Chapter 06 (draft) →