Most institutions allow a tolerance of ±10% around a stated limit, and some deduct marks beyond it. Because the penalty is mechanical, checking length is worth doing before every draft, not only at the end.
What usually counts toward the limit
| Element | Commonly counted? |
|---|---|
| Main body text | Yes |
| In-text citations | Usually yes |
| Block quotations | Usually yes |
| Headings and subheadings | Varies |
| Footnotes | Varies — check your handbook |
| Abstract | Often excluded, sometimes capped separately |
| Reference list and bibliography | Normally excluded |
| Appendices, tables, figure captions | Normally excluded |
Because the rules differ, paste only the sections your handbook counts. Copying the body text without the reference list gives you the number your marker will use.
How to check a chapter or a whole thesis
- Copy the body text of one chapter and paste it into the counter.
- Record the word count against your per-chapter budget — for an 80,000-word thesis, introduction 8,000, literature review 16,000, methods 12,000, results 16,000, discussion 20,000, conclusion 8,000.
- Watch average words per sentence: above about 28 signals clause-heavy prose that examiners find slow.
- Repeat per chapter rather than once at the end, so overruns stay local.
Why writers use this instead of a word processor
- It counts exactly the text you paste, so excluded sections cannot inflate the total.
- It reports sentences, paragraphs and average sentence length, which word processors bury.
- It works on any device, including a borrowed or locked-down library machine.
- Nothing is uploaded, which matters for unpublished research: counting happens in your browser and the text is discarded when you close the tab.
Confidentiality is the reason many researchers prefer a client-side tool. There is no account, no server log and no stored copy of your draft.