How to Tell a Good Non-Fiction Book from a Bad One: A Checklist

A bad non-fiction book rarely announces itself. It arrives with a handsome cover, a blurb promising to change how you see the world, and praise from a writer you admire. Then, a hundred pages in, you notice the same anecdote has been told twice, the central claim rests on one unverifiable story, and you could not repeat a word of it to a friend. None of that is unavoidable. The warning signs are almost always inside the book, and most of them can be found in ten minutes — standing in a shop, or sitting in a library chair — before you commit your money or your evening.

Before you read a word

Start with the scaffolding, the parts most readers skip.

• The contents page. Specific chapter titles ("The Salt Tax, 1930") suggest an author who deals in particulars. Vague ones ("Crisis", "Change") often hide padding.
• The index. A thin index in a fat history or biography is a quiet alarm. Detailed indexes usually mean a careful author and a publisher who took the book seriously.
• The notes. Open a chapter at random and turn to the back. Do the notes cite documents, interviews, datasets and page numbers — or do they mostly repeat other popular books on the subject?
• The bibliography. Ask whether it looks like a working list or a display. Consistent, dated, well-published sources are a good sign. Twenty undated websites are not.

You are not looking for perfection. Trade books compress research, and every writer leans on secondary sources somewhere. You are checking whether the author has shown their working at all.

A source you cannot check is not evidence. It is atmosphere.

The sourcing test

Numbers need a home

A statistic that arrives without a source — in a business book, a diet book, an economics primer — should be treated as decoration until proven otherwise. You do not need to verify every figure. You should be able to see, from the notes, where the big ones came from and how old they are.

Anecdotes have a job

Stories are fine; they are how arguments become memorable. The test is whether an anecdote is doing the work of evidence. One vivid interview can illustrate a trend; it cannot establish one. If the author generalises from a single conversation — "everyone I spoke to said…" — the claim is carrying more weight than the material can bear.

If a book offers health, money or legal advice, treat it as a starting point rather than a verdict, and check anything you intend to act on with a qualified professional.

Bias is not a dealbreaker — hidden bias is

Every writer has a point of view. A book with a clear argument, stated early and tested throughout, is more trustworthy than one that pretends to be neutral while steering you steadily towards a conclusion. Watch for:

• Loaded adjectives doing the work of evidence ("the disastrous policy", "the obvious solution").
• Adverbs that bully: surely, clearly, obviously, undeniably.
• Opponents described in caricature, or not described at all.
• Research invoked as "studies show" with no name, date or place to find it.

Check the strongest opposing case

Find the passage where the author deals with people who disagree. Does it quote their best argument and answer it properly, or the weakest version, knocked down for effect? A writer who engages seriously with critics usually ends up more convincing, not less. Occasional bias is human. Constant bias means you are holding not a piece of research but an argument — worth reading only if an argument is what you came for.

Is the prose doing any work?

Quality of writing is not decoration; it usually tracks quality of thinking. Read a page aloud. If you lose the thread twice, the problem is probably not you.

• Does the sentence structure vary, or does every sentence sit in the same shape?
• Is the jargon shorthand or performance? Real specialists explain; pretenders impress.
• Are people named, dated and quoted specifically, or does the prose float above the evidence in abstractions?

The best non-fiction makes difficult things clear without flattening them. If it reads like a set of slides with sentences attached, expect the same from the substance.

The ten-point checklist

Run through this before you buy. A single mark against a book is rarely fatal; a pattern of them is.

1. Does the author say where the information came from? Sources, interviews, archives, datasets — anything visible.
2. Are key claims checkable? Notes that point to real things, not just to other bestsellers.
3. Are the numbers sourced and dated?
4. Does a story ever stand in for proof?
5. Is the language specific? Names, dates, places, figures.
6. Are strong counterarguments handled fairly?
7. Are the hedges honest? "Perhaps", "probably", "the evidence suggests" where the evidence is thin — not everywhere.
8. Does the book keep the promise of its title? Read the last pages of the introduction, then the first pages of the conclusion.
9. Could you summarise the argument in two sentences? If not, the book may not have one.
10. Would you read one more chapter for pleasure? Interest is not a luxury; it is fuel.

A sixty-minute audition

If you want a decision rather than a hunch, give the book an hour. Read five pages from the middle, somewhere the argument should be at full stretch. Check the notes for those pages. Read the final three pages of the introduction and ask whether the author has promised something they can actually deliver. Then try to argue back: can you state, out loud, what the book is claiming and what would count as evidence against it? If you can do both, you are dealing with a writer worth your time.

Very few books pass all ten points, and they do not have to. The checklist is not there to make you a harsher reader, only a quicker one — so the hours you give to a book go to the writers who did the work.

Photo: lil_foot_ / Pixabay