How we choose what to recommend

Every recommendation on Tasks Made Easier has to pass the same four checks. This page explains them, so you can judge how much weight to give what we say.

1. Evidence from people with the same limitation

We only list a tool if we can point to someone with the same limitation saying it worked (or didn't) for them, or to an occupational therapy supplier that stocks it for exactly that purpose.

What we use today, in order of how much we trust it:

• Occupational therapy supplier catalogs — when a supplier that sells to therapists stocks a specific model as, say, a one-handed nail clipper, that tells us what it was designed for. These appear as a sourced line, not as a quote.
• Customer reviews that mention the limitation — a review from someone who had a stroke and cuts food one-handed is worth more to us than fifty that say "great product". We read those reviews and write up what they add up to, in our own words, under "What reviewers report". We don't reprint the reviews themselves: they belong to the people who wrote them, and Amazon's terms only let us publish what its API gives us. The link under each summary goes to the reviews, so you can check us.

What we are still building. We are recruiting a small panel of people who live with these limitations to try the tools at home; their comments will appear with their initials and permission, marked as ours. No occupational therapist has reviewed these guides yet — the reviewers page says who has, and who hasn't.

Where we do quote someone directly — a tester, or a forum post we read ourselves — the quote is 25 words or fewer and links to where it came from. A tool with no evidence doesn't appear on the site, however popular it is. When the reviews disagree, or there are only one or two of them, the summary says so instead of rounding it up.

2. A fit score, not a star rating

Each tool gets a fit score for each task and limitation, from 0 to 5:

• up to 2 points if therapy suppliers design and sell it for that limitation;
• up to 2 points for evidence from people who mention that limitation;
• 1 point if its measurements match what the limitation needs — for example, a handle at least 32 mm thick for a weak grip.

A tool needs at least 3 points to be recommended for a task. Each step of a task — holding the jar still, getting the lid to turn — shows at most three tools, because comparing eight similar gadgets is exhausting. Anything else that qualifies goes in a plain list further down, so a guide can still name more than three tools in total.

3. "Who it won't suit" is mandatory

Every tool has a plain statement of who should not buy it: too heavy for a weak wrist, needs two hands to set up, won't fit wide jar lids. If we can't write that sentence from the evidence, the tool isn't ready to be listed.

4. Prices are checked daily, content is reviewed on a schedule

Prices, stock and delivery badges come from Amazon and are refreshed at least daily; every price shows the time it was checked. If a price is out of date, we don't show it.

The advice itself changes more slowly, and we keep it separate. When a product is discontinued we replace it and note the change at the bottom of the page, with the date. Every guide is flagged for a re-read three months after it was published or last checked, and the date of that last check is printed at the bottom of the page.

How software is used

Some short product summaries are drafted with software, working only from the quotes and measurements we collected, and an editor checks a sample of every batch. The "try this first" techniques, the "what not to buy" lists and our situation checklists are written and edited by people.

What we don't do

We don't accept payment for placement, we don't publish star ratings we can't verify, and we never promise that a tool will fix a medical problem.