A parable.
Suppose that someone tells us "I have performed extensive experiments, detailed <here>, and determined that adding baking soda to
vinegar invariably produces, immediately and without any gas production, a blue gel."
We can simply add some baking soda to some vinegar and observe that no such thing happens. At this point we know that the experimental
procedures are flawed.
So, we should start by saying: "No, you are wrong, what happens is this:" with evidence if you want to get fancy.
Then, if you like, you can go through the no-doubt several pages of experimental procedures and point out that in fact what it
being added to the vinegar is actually salt, not baking soda, and that the experiment calls the addition of both blue food coloring
and gelatin.
If you start with the second step, then the intrepid would-be-chemist is likely to start in on why the blue food coloring is irrelevant,
and how salt is basically the same thing as baking soda in this context, or whatever. So, start with "your conclusion is wrong",
before you go into debugging the experimental procedure. And, frankly, the latter step is optional.
That's the strong position, anyways. You can go the other way around if you like, but it leads to trouble and more random quoted bits
and pieces.