The number one pitfall of reporting on science
is to take a single work, or single report and treat it as if it is proven as
fact already.  In truth, real science knowledge is that which has stood the
test of time.  That is not really exciting for reporting.  If a reporter wishes
to discuss the latest developments, they must qualify it with the fact that it
is a new study.  Even if it is peer reviewed, that is not even enough yet.   Until
something is shown through consensus built by years if not decades of results to
be either a “law” of nature or a “theory” of science, it should not be reported
as fact.

However, there are times like now, regarding the 2019 novel coronavirus
from Wuhan China, that one cannot wait for consensus and must rely on lesser
levels of review.  Here they are in order of trust ability.

Least Caution, Theories and Laws of nature which
have been backed by decades of results and are agreed upon by consensus. 

A reporter will know these by the fact that they have recognizable
names.  They are things like evolution and General Relativity, the laws of
thermodynamics.  The law that says energy cannot be created or destroyed,
conservation of energy, is a very important one right now.  So many people are
proposing “solutions” to climate change which violate that law or the laws of
thermodynamics.   They can be improved upon, but they are as factual as science
gets.  

Report on such
theories as simply being facts of nature.

Note theories like “string theory”, or other post-modern
theories in physics, do not yet fall into this category.

The basics of germ theory are what one should report on
right now regarding the novel corona virus.  Tell people to avoid unnecessary
public gatherings and mind their handwashing and sanitation.

Peer Reviewed Papers.
These are easily identified by appearing in a journal.  The better known the
journal the more serious you may wish to take it.  If the peer review is “post publication
peer review” … look at the reviews and responses to the reviews for yourself.  
In any case such papers what have been cited, read the citations and make
sure they are not refuting the paper.   To be safe wait until a paper has
been cited, preferably by someone who is not the author.   PhD and possibly MS
theses fall into this category.

Report on these papers with confidence.  Yet also qualify
that further studies may contradict these.  Peer reviewed studies which are coming
out on the coronavirus are worthy of being written about by non-scientist
journalist.

Most Caution, Formally Archived Preprints.

Here things get a little tricky.  Depending on the field
such papers can be almost as good as peer reviewed.  For example, a formally
archived paper by a large scientific collaboration undergoes a lot of review
before it is published.  However, such papers have been shown to be seriously
flawed in the past.  A paper by a single scientist consider the reputation of
the scientist.  What is their current or former academic position?  If they
were someone who was qualified to teach science at college or university, they
are at least not going to “not even wrong”.  It is also wise to see if the
pre-print has been cited by someone.  It is becoming more common to dispense
with formal publishing.  Citation by other researchers then becomes a true mark
of acceptance.

A science reporter should only report on such papers if they
have personal knowledge of the field or if they are being advised by someone
who does.   If you are writing about coronavirus and are not at least a medical
doctor or nurse, then don’t write about non-peer reviewed papers.

Those of us who are scientist, who do write about archived papers
or pre-prints could be providing a sort of public review of the work.  If you
see multiple scientist-journalist praising a paper report on the fact they are
praising it.  If you see that such scientist-journalist are panning a paper or
doubtful, report on the doubts.

Anything that does not fall into one of these categories is
not science that should be reported on as fact yet.  Anything less is a
hypothesis, an educated scientific guess at best.

What of theories like “String Theory”? 

Theories in post modern theoretical physics have at least some math behind them
which is not obviously wrong.  That mathematical framework is often quite substantial
and has been worked on by many.  However, until any such models have made testable
predictions, which have been verified by data, they are not theories in the
classical sense of the word.   Instead they are mathematical frameworks, techniques,
that can be used to construct numerous hypotheses.

Saying “string mathematical framework” just isn’t as catchy
as saying “string theory”.