The abstract is the only part of your dissertation most people will ever read. In roughly 150–350 words, it needs to convey the problem, methodology, key findings, and significance — clearly enough to stand entirely on its own.
| Element | Typical Length |
|---|---|
| Problem/purpose | 1–2 sentences |
| Methodology | 1–2 sentences — design, sample, key method |
| Key findings | 2–3 sentences — the main results, not every result |
| Implications/significance | 1–2 sentences — why it matters |
That's the entire abstract — typically one paragraph, sometimes split into two. Every sentence has to earn its place; there's no room for background context, citations, or methodological justification.
An abstract is a compression exercise, not a preview — it can only be written accurately once the dissertation itself is finished, because it needs to reflect the actual findings and conclusions, not what you originally expected to find. Drafting the abstract before your results are final is the most common reason abstracts need substantial revision later.
Write five drafts, each shorter than the last. Most over-length abstracts come from trying to summarize everything important rather than the single most important finding per section. Cutting ruthlessly on each pass usually surfaces the version that actually fits the word limit.
Every required element, within your word limit, reflecting your actual findings.
An abstract is shorter and more compressed, typically for academic indexing and search. An executive summary, used in some applied doctoral programs, is longer and more detailed, often including specific recommendations. Check which your program requires.
We recommend waiting until your results and conclusions are stable, since the abstract needs to reflect them accurately — but we can draft a placeholder version early and finalize it once your findings are locked in.
Most institutions require 4–6 keywords listed after the abstract, used for indexing in dissertation databases. We select these to reflect your core topic, methodology, and population.