Decide whether a paper is worth reading in full
For a literature review, most papers fail on two facts: too few participants, or the wrong design. Kriterio finds the numbers the paper prints and says Read, Skim or Skip. It never invents a statistic.
How it works
- Add Kriterio to Chrome. In its settings, turn on the Researcher preset (marked Beta).
- Set your bar: the smallest sample you accept (default 200) and the design you require (default: randomized).
- Open a paper and click the Kriterio icon, or press Alt+Shift+D.
- Get Read, Skim or Skip with the reason. If the sample size isn't printed, it asks you instead of working one out.
What it checks
- In code, every time: the sample size and the study design, using only numbers and words the paper prints.
- By the AI, quoting the page: whether the population studied fits your question (by default, adults in general practice).
- Where: the preset turns on by itself on arXiv, PubMed, PubMed Central, bioRxiv, medRxiv, Nature, ScienceDirect, PLOS and Frontiers.
Questions
Does it summarize the paper?
No. It checks the paper against your bar and tells you whether to read it. The reading is still yours.
Is the Researcher preset ready?
It's marked Beta: it works today, and we're still tuning how it reads papers. A number it can't find is a question, never a guess.
What does it cost?
Free covers all presets with your own limits and 25 checks a week. Pro ($12/month) adds your own rules in plain words, Compare for your open tabs, deep check and 100 checks a day.
Kriterio is not affiliated with the sites named on this page.
More ways to use Kriterio
- Skip low-budget Upwork jobs in one click
- Stop reading freelance briefs that don't pay enough
- Catch the catch in any Amazon listing
- Know what you're really getting on eBay
- Your rental checklist, checked on every listing
- Screen a profile in seconds, against your own job description
- Read the pricing page before it reads your wallet
Deciding as Researchern ≥ 200
Read
Randomized trial, 1,240 adults
A randomized trial with 1,240 participants, in adults in general practice.
Sample sizeat least 2001,240
Study designincludes “random”randomized
Right call? YesNo