ACCA SBL 2026: Stop Forcing Lewin Into Every Change Question
The biggest reason candidates fail SBL isn't weak knowledge — it's answering a question the examiner never asked. And the fastest way to do it is bolting Lewin's change model onto a requirement that never mentioned the change process.
The March/June 2025 examiner's report is blunt about it: "Some candidates appear to assume that Lewin must be mentioned whenever a question has anything to do with change – this is not so." If your revision plan is "see the word change, reach for unfreeze-change-refreeze," you are walking into the single most common mark-loser on the paper.
What the examiner actually said
Top of the list of reasons for failure: not answering the task that has been asked. Not wasting time, not weak theory — answering around the question instead of at it. SBL gives you 80 technical marks, and every one of them is for applying knowledge to the scenario. Repeating rote-learned theory scores "few, if any, marks."
Task 1(a) was the trap. The requirement asked candidates to discuss the contextual features of the charity that the CEO needed to consider before her programme of changes. "Contextual features" was the object of the verb — so the answer had to be built around those features. Instead, many candidates structured their answers around the programme of changes itself, and a chunk reached for Lewin's model or PESTEL — which pushed them onto external factors when the requirement was internal. Marks were awarded where features were mentioned, but they were low, because the features weren't the focus.
The other silent killer: copying exhibit information without developing it. Two marks were available throughout the exam for a developed point, but candidates "often had few or no such points," so they had to scrape a pass one single mark at a time.
Wrong answer vs right answer
What lost marks: "The CEO wants to maximise income. Using Lewin's model, HP must first unfreeze current attitudes, then implement the changes, then refreeze the new behaviours to embed them…" — a tidy model dump that never names a single contextual feature of HP.
What scored: "Readiness for change — HP's staff and volunteers have limited recent experience of change, so resistance is likely; the CEO should sequence the least disruptive changes first to build confidence. Power — the trustees, not just the CEO, hold authority, so their visible backing is essential if the income changes prove unpopular with donors." Each point names a feature, then says why it matters for HP's specific situation. That's how you turn a single mark into two.
What to do in the exam
Underline the object of the requirement verb before you write a word. If it says "discuss the contextual features," your headings are contextual features — not the change programme, not a model you happen to remember. The requirement, not your memory, dictates the structure.
Reach for the model the study guide names for that topic — or none at all. Contextual features links to Balogun and Hope Hailey in the SBL study guide. Lewin describes the process of change and was not required here. If a named model doesn't fit the verb, don't force one in.
Develop every point with the "so what for HP" test. State the feature, then explain its consequence for this organisation's change, or support it with detail from the exhibits. A point you can't develop is a point earning half of what it could.
The bottom line
SBL's pass rate sits around 50% — roughly half the room goes home disappointed, and most of them knew the theory. They just answered the question they revised for instead of the one on the screen. Read the verb, answer its object, develop the point. Lewin can stay in the drawer unless the requirement asks him out.