Chapter 26 — Human factors and cyber policy: architecture alternatives
Lesson map
Lesson 26 published. Map, example, EX076.
flowchart TD
N0["Learning outcomes"]
N1["26.1 Three alternatives"]
N2["26.2 Criteria and scoring"]
N3["Lab — Weighted comparison of three programme designs"]
N4["26.3 An open choice"]
N5["Synthesis"]
N0 --> N1 --> N2 --> N3 --> N4 --> N5
N5 --> EX["Worked example"]
EX --> Q["Exercise EX076"]
Logical drawing
flowchart LR
N0["Learning outcomes"]
N1["26.1 Three alternatives"]
N2["26.2 Criteria and scoring"]
N3["Lab — Weighted comparison of three programme designs"]
N4["26.3 An open choice"]
N5["Synthesis"]
N0 --> N1 --> N2 --> N3 --> N4 --> N5 --> Q["EX076"]
[Learning outcomes]
|
v
[26.1 Three alternatives]
|
v
[26.2 Criteria and scoring]
|
v
[Lab — Weighted comparison of three programme designs]
|
v
[26.3 An open choice]
|
v
[Synthesis]
|
v
[Worked example] --> [Exercise EX076]
Learning outcomes
You will (1) compare alternative architectures for a security awareness programme; (2) weigh alternatives against human and organisational criteria; (3) justify a choice openly.
26.1 Three alternatives
Aiko compares three ways Greyford could organise its programme. The first is compliance-led: an annual online course and quarterly simulations measured by completion and click rates. The second is reporting-led: easy reporting, fast feedback, short monthly refreshers and simulations measured by reporting speed. The third is design-led: invest mainly in technical changes that remove decisions from staff, such as security keys and stronger filtering, with minimal training. Each reflects a different theory of what makes people secure, and each has strengths that the others lack.
26.2 Criteria and scoring
Aiko scores the alternatives against criteria agreed with stakeholders: risk reduction, staff burden, cost, regulatory evidence and staff trust. The lab computes weighted totals. The design-led option scores highest on risk and burden but weakest on regulatory evidence, because the regulator expects some training records. The reporting-led option scores well across most criteria. The compliance-led option scores best only on ease of producing evidence, which matches the critique by Bada, Sasse and Nurse (2019) that such campaigns rarely change behaviour. Stakeholders set the weights before seeing any scores, which prevents the weights being tuned to favour an option.
Lab — Weighted comparison of three programme designs
# Weighted comparison of three awareness programme designs (synthetic 1-5 scores).
w = {"risk": 0.30, "burden": 0.20, "cost": 0.15, "evidence": 0.15, "trust": 0.20}
s = {"compliance": {"risk": 2, "burden": 2, "cost": 4, "evidence": 5, "trust": 2},
"reporting": {"risk": 4, "burden": 4, "cost": 4, "evidence": 3, "trust": 4},
"design": {"risk": 5, "burden": 5, "cost": 2, "evidence": 2, "trust": 4}}
def tot(wt): return {o: sum(wt[c] * v for c, v in sc.items()) / sum(wt.values()) for o, sc in s.items()}
print("base:", {o: round(v, 2) for o, v in tot(w).items()})
print("evidence weight doubled:", {o: round(v, 2) for o, v in tot(dict(w, evidence=0.30)).items()})
Run with python3 lab_options.py. Verified output:
base: {'compliance': 2.75, 'reporting': 3.85, 'design': 3.9}
evidence weight doubled: {'compliance': 3.04, 'reporting': 3.74, 'design': 3.65}
26.3 An open choice
Aiko recommends a hybrid: design-led technical changes combined with the reporting-led programme, keeping short training that also produces the records the regulator needs. She explains the trade-offs openly: the hybrid costs more in the first year, but reduces burden and risk. She notes that the weights reflect stakeholder judgements, and shows that the recommendation holds when the weight on regulatory evidence is doubled: the totals become 3.74 for reporting-led and 3.65 for design-led, both still well above compliance-led at 3.04, so the two components of the hybrid remain the strongest. The board approves the hybrid and asks for evaluation after twelve months.
Worked example
Aiko compares compliance-led, reporting-led and design-led programmes on five weighted criteria and recommends a hybrid with a sensitivity check.
Synthetic case. Scores and weights are synthetic; Bada et al. (2019) is real.
Errors and limits
Scores are judgements. Hybrid designs can lose focus if not managed.
Exercises
- EX076 — Design-led weakness. Where does the design-led option score weakest? Rubric: 3 = regulatory evidence;; 2 = evidence;; 1 = risk;; 0 = none.
- EX077 — Weighted totals. What were the weighted totals in the lab? Rubric: 3 = compliance 2.75, reporting 3.85, design 3.90;; 2 = ordering;; 1 = vague;; 0 = none.
- EX078 — Hybrid robustness. How did Aiko test the hybrid recommendation? Rubric: 3 = doubling the regulatory evidence weight;; 2 = sensitivity;; 1 = vague;; 0 = none.
Synthesis
Comparing programme architectures on agreed criteria leads to an open, tested choice, often a hybrid.
Continue studying — chapter 26 of 50
← Chapter 25 (draft) · Course chapter index · Chapter 27 (draft) →