Solving Complex Problems and Decision Making

This opening module separates two skills leaders often blur together: problem solving, which asks what is true, and decision making, which asks what to do about it. It sets out two failure modes and their different fixes, shows how to keep the disciplines in sequence in a real meeting, and explains how the programme is organised, from Simon and prospect theory to structured methods and collaborative decision making.

This module builds the shared vocabulary for deciding well. It distinguishes programmed from non programmed decisions, sets out the conditions of certainty, risk, uncertainty and ambiguity, and walks through the classical six step rational decision model, drawing on Daft and Harrison, including the three ordered tests of feasibility, satisfactoriness and affordability. It treats the model as a benchmark and offers four questions to ask before any decision.

Intelligent, well informed people still do not optimise, and this module explains why. It draws on Herbert Simon's administrative model, bounded rationality and satisficing, and on Mintzberg's non sequential model of identification, development and selection. The Kodak digital camera case shows how a run of individually satisfactory decisions can prove fatal, and the module offers practical questions for testing whether good enough is still good enough.

This module explores Kahneman and Tversky's prospect theory and its three building blocks: loss aversion, reference dependence and diminishing sensitivity. It shows how losses loom roughly twice as large as equivalent gains, how gain and loss framing changes choices, and why sunk cost reasoning keeps failing projects alive, using the Concorde case. It offers practical habits and questions for separating a decision that protects a loss from one that weighs a future.

Groups fail in their own distinctive way, and this module explains how. Drawing on Irving Janis's work on groupthink and the Bay of Pigs, it sets out four observable symptoms, including self censorship and the illusion of unanimity, and why capable, cohesive groups are most exposed. It then matches practical countermeasures, such as a devil's advocate, independent sub groups and leaders speaking last, to the failures they address, so dissent is designed in rather than hoped for.

Some problems are hard because people disagree about what the problem is. This module introduces four problem structuring methods for those situations: Checkland's Soft Systems Methodology with rich pictures and CATWOE, Friend's Strategic Choice Approach and its three kinds of uncertainty, Eden's SODA cognitive mapping, and Rosenhead's Robustness Analysis. It explains that soft describes the problem, not the rigour, and offers four questions usable without training.

This module sets two structured methods side by side: Kepner and Tregoe's slow, thorough separation of problem analysis from decision analysis, and John Boyd's fast, iterative OODA loop of observe, orient, decide and act. It explains why assumptions must never quietly become facts, why orientation is the decisive stage, and why a faster cycle can beat a better machine. The core skill is choosing the right speed of thinking for the situation.

Most decisions are shaped before anyone sits down to make them, by who is involved and who owns the outcome. This module uses the Vroom Yetton Jago model to treat involvement as a situational judgement, from deciding alone to delegating, and examines the costs of too little and too much consultation. It then separates involvement from ownership, drawing on decision rights frameworks, with four roles to assign so decisions do not drift.

Once options exist, someone has to work out what each is worth. This module covers cost benefit analysis from Jules Dupuit's original method to social cost benefit analysis, with six steps grouped into three jobs, and the ways an analysis quietly misleads, including forced numbers. It pairs this with market analysis and the 6W customer analysis model, showing how genuine rather than assumed need defines the options worth valuing.

Every technique in this programme depends on the evidence feeding it. This module applies Deming's standard of accurate, valid, precise and reliable data, sets out the main types of data and what each can legitimately support, and covers how primary data is collected, including why a survey is only as good as its design. The 2020 A Level grading algorithm shows how capable people produced a confident, precise and wrong answer.

This module offers a deliberately bounded introduction to design thinking as one problem solving technique among several, as used in Avanti's work with a local authority. It covers starting from the person rather than the specification, combining divergent and convergent thinking in sequence, and IDEO's three tests of desirable, feasible and viable. IDEO's four day shopping cart redesign shows the method in action, and four simple moves bring it into everyday work.

People who approach problems differently often misjudge each other, and this module offers a vocabulary for that. It introduces Michael Kirton's Adaption Innovation continuum, what KAI does and does not measure, and three common misreadings. It explores what adaptors and innovators each bring, how they misjudge one another under pressure, and how to read friction as difference rather than conflict, before turning to designing, matching and sequencing a mixed team.

How you frame a problem decides which solutions you can see. This module covers two simple root cause techniques from the Toyota Production System tradition: the five whys, associated with Taiichi Ohno, which follows one chain of reasoning in depth, and Ishikawa's fishbone diagram, which widens the search across fixed categories first. It shows how to run both well in a meeting, how they fit together, and why the discipline matters more than the technique.

This deliberately narrow module asks whether a decision is genuinely made with the people it affects, or made for them and then announced. It sets out a collaboration spectrum from competing to co-designing, separates coordinating from collaborating, and adapts Habermas's principles of communicative rationality into three tests of comprehensibility, sincerity and legitimacy. Porto Alegre's participatory budgeting shows what binding citizen authority, combined with technical scrutiny, can look like.

This capstone brings no new theory. It reduces the toolkit to three questions, diagnose well, decide well and revisit both honestly, and applies it to three verified cases: the Boeing 737 MAX and MCAS, New Coke, and Birmingham City Council's Oracle programme. Each case is read against decision rights, satisficing, groupthink, market analysis, prospect theory and cost benefit analysis, showing how pressure overrides evidence and how to notice it in time.

Many public decisions involve cost, quality, risk, community impact and deliverability, and forcing them into one number is not always honest. This module introduces Multi Criteria Decision Analysis as a way to make judgement visible: four steps from criteria to choice, simple weighted scoring and Saaty's Analytic Hierarchy Process, and its place in HM Treasury's Green Book. It stresses agreeing weights before scores and shows how the same framework can test or merely decorate a decision.

Under real time pressure, nobody runs the six step model. This module draws on Gary Klein's fireground research and the recognition primed decision model to show how experts decide well under pressure: recognising a pattern, simulating one option once, then committing. Captain Sullenberger's 208 second decision on the Hudson illustrates the pattern, and the module is honest about when recognition lets you down and how to build expertise through after action review.

Put experts in a room and the loudest or most senior voice usually wins. This module explains how the Delphi Method, developed at the RAND Corporation and still used in clinical guideline work, removes that distortion through anonymous rounds, controlled feedback and iteration. It covers the facilitator's responsibilities, when Delphi fits and when it does not, four ways it fails, and why a well documented disagreement is a legitimate answer.

This module turns from the decision to the decision maker. It draws on Barry Schwartz's paradox of choice to show why more options can leave people more anxious and less satisfied, and on decision fatigue to show how judgement degrades across a day of decisions. The hungry judge study and Gloeckner's reanalysis model how to cite challenged evidence honestly, and practical responses work at personal and organisational levels.

Some decisions are not missing information; they are genuine clashes of value. This module shows how to recognise an ethical decision and test it against three classical lenses: utilitarian (Bentham and Mill), deontological (Kant) and virtue based (Aristotle). A council budget dilemma shows how lenses can agree on different foundations or disagree, and NICE's cost per QALY threshold models naming a hard trade off openly rather than hiding it in a number.

This second capstone brings no new theory. It recaps the frameworks from across the programme, adding multi criteria analysis, naturalistic decision making, Delphi, decision fatigue and ethics, and applies them to three current failures: Australia's Robodebt scheme, the Columbia shuttle disaster, and Winter Storm Uri in Texas. Each case shows a warning already in the room that met no route to change course, and closes with diagnose honestly, decide deliberately, own the trade off.