An argument from authority has two forms. The most common is to state that because the arguer has a credential of some kind, then the statement they make must be true. The alternate is to render a statement made by someone without a credential as false.
The truth of a statement does not depend on the credential of the person who makes the statement. For example, if a professor of physics asks their two year old to repeat a statement about gravity, neither the professor of physics nor the two year old are necessarily telling a truth or falsehood due to who they are. Instead the statement itself must be true or false according to current scientific evidence.
In the above example, the professor of physics is far more likely to give an accurate statement about physics than he would about the current state of political support for science. It is tempting to laud his credentials when making a statement about science funding, as if he is an expert on that too. He isn’t.
When an expert witness is called, s/he must still be able to reference the source of their expert knowledge. This isn’t to say “I have a PhD in Physics”, it is to say “this fact is backed up by this evidence found from these experiments performed by these scientists”. An exception to that would be a commonly accepted idea in physics, such at the Theory of Gravity. Even so, s/he should be able to explain the methods of experiments done to test the theory.
Frequently an expert in a school of science is asked on the media to comment on a different school of science. This is poor form. They aren’t qualified to be an expert witness for that form of science. The assumption of accuracy of their statements based on their qualification is illogical. It is like asking a meteorologist which tyres to put on the four wheel drive. After all, they are an expert, right? Wrong. Yet an expert in meteorology may also be a four wheel drive enthusiast and be able to reference where to find the information, why these tyres are good or bad and so forth. Yet you wouldn’t say “John Smith, a meteorologist, recommends these tyres”.
By the same token, in the example above, the two year old’s statement isn’t necessarily wrong because it came from a two year old. The speaker of the information does not equate to validity, the science behind the statement (which is person independent) equates to the validity of the statement. That is, if there is not scientific evidence supporting the statement, then it has no credibility regardless of who makes the statement; and if there is scientific evidence supporting a statement, then it has credibility regardless of who makes the statement.
One hopes that a relevant expert in the field would know the topic better and not leap to faulty conclusions. Unfortunately that isn’t always true. Frequently the media will misunderstand the science involved in an experiment, not understanding the need for good methodology, peer review, the affect of sample size on the validity of results or even the difference between writing about the results and speculating on what it might mean or where to go from here.
A good rule of thumb when reading a mass media write up of something in science (or a infographic) is to check for references to a well known scientific journal. If there isn’t one, take the article/picture with a pinch of salt. If there is a reference, check out the reference and see if what was written reflects the abstract, and then the actual article itself.