Master of Wine common mistakes
Master of Wine · Understand the traps and how to avoid them.
Framing the research question as a broad topic area ("Bordeaux's response to climate change") rather than a specific, answerable question with a defined scope, time frame, and evidence base.
Candidates often start from a subject they find interesting and assume the dissertation will narrow itself as they read. Without a sharp question, every subsequent decision - sources, structure, chapter length - has no anchor, and the work drifts into a survey rather than an argument.
What to do instead
Before writing anything, draft the question in one sentence with three constraints: a defined population or region, a defined period or vintage range, and a defined outcome you will evaluate. Test it by asking what evidence would prove the answer wrong. If you cannot name that evidence, the question is still a topic. Get the question signed off in writing by your supervisor before starting research.
Writing a literature review that summarises sources one after another ("Smith says X. Jones says Y. Lee says Z.") without organising them around the debate or the gap the dissertation addresses.
It feels productive to demonstrate that you have read widely, and summarising is the easiest way to show effort. But an annotated bibliography is not a review; examiners look for synthesis and a clear statement of what is unresolved.
What to do instead
Structure the review by themes or competing positions, not by author. For each theme, state what is settled, what is contested, and what remains unexamined - then show how your question sits in that gap. End the review with an explicit paragraph naming the gap and how the dissertation will address it. Delete any source that does not serve one of those themes.
Presenting tasting notes or winery interviews as objective data without stating how samples were selected, how many tasters were involved, whether panels were blind, and how judgements were calibrated.
Sensory and interview material is the most accessible evidence for many candidates, so it gets treated as self-evidently valid. Candidates often do not realise that unstructured tasting impressions are anecdotal unless sampling and protocol are specified.
What to do instead
Write a short methods section before collecting data that specifies selection criteria, sample size, blind or sighted conditions, the tasting sheet or interview guide used, and how you will handle disagreement between tasters. Where the design is exploratory, say so and label the findings as indicative rather than generalisable. Report the limitations explicitly rather than leaving them for the examiner to find.
Overstating causal claims from correlational data - for example, attributing a rise in regional prices to a single critic score or a change in appellation rules when other factors were moving at the same time.
Candidates want a striking conclusion, and a clean causal story reads better than a hedged one. Confounding variables such as vintage quality, currency movements, and global demand are easy to overlook when the data set is small.
What to do instead
For every causal claim, list at least two alternative explanations and state in the text why the data cannot rule them out. Use language that matches the design: 'associated with' for correlation, 'consistent with' for suggestive evidence. If the design cannot support causation, say what design would, and treat that as a recommendation for further work.
Treating the dissertation as a long essay with no explicit methodology chapter, so the reader cannot tell how evidence was gathered, analysed, or judged.
Many candidates come from a humanities or trade background where method is implicit. They assume the reader will infer the approach from the narrative, and they spend their word count on content instead of justifying process.
What to do instead
Include a dedicated methods chapter of roughly 10–15 per cent of the total length. Cover research design, data sources, collection procedure, analytical approach, and ethical considerations in that order. Justify each choice against the research question rather than describing it neutrally. If the approach is qualitative, explain how themes were derived and how you guarded against cherry-picking.
Relying on secondary and trade-press sources for factual claims that require primary data or official records, and citing them without checking the original.
Trade magazines, blogs, and marketing material are quick to find and often contain the exact statistic the candidate needs. Repeated citation of the same secondary summary creates a false sense of corroboration, and errors propagate.
What to do instead
Trace every key statistic to its primary source - official appellation records, customs data, producer accounts, or peer-reviewed studies - and cite that source directly. Where only secondary sources exist, say so and flag the reliability. Keep a source log recording where each figure came from and the date accessed.
Producing a dissertation whose structure does not mirror the research question, with long descriptive background chapters and a rushed analysis and conclusion.
Background is comfortable to write and easy to research, so it expands to fill the available time. Analysis is harder and gets deferred until the deadline, leaving conclusions thin and unsupported.
What to do instead
Build a chapter plan that allocates a fixed word budget to each chapter before drafting, with the analysis and discussion taking the largest share after the literature review. Write the conclusion early as a skeleton and revise it as evidence arrives. Cut background that the examiner already knows; keep only what the argument needs.
Failing to engage critically with counter-evidence and rival interpretations, presenting only findings that support the thesis.
Candidates become attached to their hypothesis and read contradictory data as a threat rather than a contribution. Selection bias creeps in gradually and is rarely deliberate.
What to do instead
Deliberately search for disconfirming evidence and give it a section of its own. For each main finding, state the strongest objection and respond to it. If the objection cannot be answered, revise the claim rather than omitting the objection. Ask a peer to read the draft and identify what has been left out.
Mishandling citation and attribution - inconsistent referencing style, missing page numbers for direct quotations, and close paraphrase of a source without acknowledgement.
Note-taking that mixes the candidate's words with the source's words, combined with a citation style applied only at the end, produces accidental plagiarism. Candidates often do not realise that unattributed close paraphrase counts the same as copying.
What to do instead
Adopt one referencing style at the outset and apply it as you write, not afterwards. In notes, mark quotations clearly and record page numbers immediately. Paraphrase by closing the source and writing the idea from memory, then checking against the original. Run the final draft through plagiarism-detection software before submission and resolve every flagged passage.
Submitting a draft that has not been proofread for structure, argument flow, and consistency of terminology, treating the dissertation as a formatting exercise in the final week.
Candidates underestimate the time needed for revision and assume supervisors will catch structural problems. Late-stage formatting consumes the remaining time, and inconsistencies in units, terminology, and tense survive to submission.
What to do instead
Reserve at least three weeks for revision and plan two full read-throughs: one for argument and structure, one for language and consistency. Produce a style sheet fixing terms, units, and abbreviations, and check the document against it. Give the penultimate draft to a reader unfamiliar with the topic and ask them to summarise your argument; where their summary diverges from your intention, revise.