A thesis introduction is not a list of sections; it is a chain of anchored claims. Every move (the hook, the conversation you join, the problem you raise, the question you pose, the contribution you promise) must be tethered to something a reader can verify. Unanchored intros read as opinion; anchored intros earn the reader's next 90% of attention.
Every intro moves the reader through the same five steps. Read this column alongside the Five Anchors sidebar; each stage on the left must tie to at least one anchor on the right.
Plant the topic in Pasteur's quadrant, useful AND theoretically rich. Make the reader curious within the first paragraph.
Two archetypes (Grant & Pollock). The Quote, a startling line, vignette, or industry headline. The Trend, a shift on Main Street or in the Ivory Tower whose causes remain a mystery.
Establish the field through intertextual coherence: synthesised, progressive, or non-coherent. Show the reader the existing terrain before claiming new ground.
Avoid: "no one has studied X" · "to our knowledge, ours is the first…" These are first-mover fictions that rarely survive review.
Pivot from establishing to problematizing. The strongest intros target problems, puzzles, paradoxes, not gaps. Aim for the inadequacy middle ground: a fresh look without burning prior work.
Two routes (Hollenbeck). Consensus shifting, challenging widely held assumptions. Consensus creation, resolving a live disagreement.
Research questions are navigational tools, not fixed coordinates (Agee). Expect them to evolve as understanding deepens. A good RQ is specific enough to be answerable, broad enough to be interesting.
Test: can a peer say what counts as evidence for and against? If not, the RQ is still a topic, not a question.
Preview the theoretical contribution and signpost the structure. The intro should function as a stand-alone mini-summary: a reader who reads only this should know the question, the lens, the method, and the contribution.
Calibrate: avoid overpromising, a 14% reviewer-flagged pitfall. Match the promise to what the rest of the thesis actually delivers.
Synthesised from Sandberg & Alvesson (2010), Grant & Pollock (2011), and Agee (2009).
Reinforces or modestly extends existing theory. Common, but rarely produces interesting contributions.
Names and challenges the assumptions underpinning a literature. Higher upside; harder to do.
Treat the research question as provisional. The cycle below repeats throughout the thesis, not just at the start.
Average rewrites of the intro by AMJ Best Article winners.
Of total writing time spent on the intro alone (2.5× its length share).
Recommended length for the intro (double-spaced). Tight, snappy, scoped.
The intro must answer: Who cares · What do we know · What will we learn.
Tick them off as you go. A draft that fails three or more is not ready for a supervisor, let alone a reviewer.
Two ways to score your Introduction & Background against the five anchors, the funnel stages, the pitfall scan, and the twelve-point audit. The evaluator quotes your draft and returns a prioritised fix list.
Paste into ChatGPT, Claude, Gemini, or any LLM. Then paste your draft. Returns a scorecard out of ten, a funnel-stage diagnosis, a pitfall scan, the twelve-point audit, and the top five fixes.
For Claude Code users. Save SKILL.md into ~/.claude/skills/anchoring-map-evaluator/ and invoke as anchoring-map-evaluator. Same rubric, native integration.