Keto made easy

Why 'Do Your Own Research' Is a Red Flag

Quick answer: When someone makes a health claim and then says "do your own research," it usually means they don't have a source to give you. The person making the claim owns the burden of proof. So what you do is ask for the study, by name or link, then read its abstract yourself. If no study appears, you've learned something too.

On this page

The myth: "do your own research" means "think for yourself"

It sounds empowering. Don't just trust the experts, look into it, make up your own mind. And thinking for yourself is good.

But look at how the phrase actually gets used. Someone posts that a food is toxic, or that a supplement melts fat, or that keto will wreck your thyroid. You ask how they know. They say "do your own research."

That's not an invitation to think. It's a handoff. They made the claim, and now you're supposed to go prove it for them. And if you search for it, you'll find something that agrees, because you can find something that agrees with almost anything.

What holds up: sources are there to be checked

Real research works the other way round. The person with the claim shows their work, and anyone can check it. That's why published studies have methods sections and reference lists: so someone else can see exactly what was done and repeat it.

So the honest version of "do your own research" is "here's my source, check it." Those are very different sentences.

The three-question check

You don't need a science degree for this. You need three questions and about ten minutes.

  1. "What's the source?" A study title, a link, an author and year. Not "studies show," not a video of someone else saying "studies show."
  2. "What kind of study?" Was it in people? How many? Did researchers assign the diet (a trial) or just watch what people ate (an observational study)? A mouse study or a survey of online posts can be interesting and still not tell you what happens to you. See observational vs randomised studies.
  3. "Does the abstract say what they say it says?" The abstract is the short summary at the top of a paper, and it's free to read on PubMed. Read it yourself. Here's how to search PubMed.

If the answer to question one is "do your own research," you can stop there.

A real example of checking

This isn't hypothetical. When the evidence pages behind this site were built in October 2026, the first drafts of the study tables were written by research assistants, and three of them said they'd written the one-line findings from memory rather than from the abstracts. So every study's abstract was pulled from PubMed (74 of them) and every number was checked against it.

Six findings didn't match. One example: a draft described a metabolic-ward study as involving "a 30% calorie cut." That figure isn't in the abstract. What the abstract does report is that 19 adults lost body fat at 53 ± 6 g a day with carbohydrate restriction versus 89 ± 6 g a day with fat restriction (Hall et al., 2015). The page uses the abstract's figures. (More on that study in the metabolic-ward keto studies explained.)

That's the whole method. Not trusting the summary, not distrusting it either, just reading the source.

What the research says

This post is about method, not a single finding, so the "research" here is the checking process itself:

If it happened to a careful process, it happens to forum posts, magazine articles and videos. Including, for what it's worth, KetoDick's own. See red flags in nutrition claims.

KetoDick's take

This is one of his consistent positions. In his 2018 basics video he says he's not a "do your own research" person; he did his research and you can have it, with sources in the comments for you and your doctor to check (video UygKGoP_BtA). He closes the same video calling "do your own research" a red flag: people who correct you should share their sources, and sources are there to be independently evaluated.

In 2017 he said he reads PubMed and NIH publications, cross-checks blog ideas against medical research, and will find you the actual studies if you ask so you can decide for yourself (video xyEZL0HuejI). In 2019 his tip was to go to Google Scholar or add "pubmed" to a search, because a study has to say exactly what it did (video 1JbK1_ClDmY).

He turns it on his own community, too. A keto forum will answer every question with keto, he says, so beware of confirmation bias (video 1m1hPgaRIg4). See keto forums and confirmation bias.

One honest note: his videos cite by kind ("NIH studies," "PubMed") more than by paper. The evidence pages exist partly to put names and links next to those claims, and where his numbers and the papers differ, the posts on this site say both.

Key takeaways