Scientists don't, when they do science. All they need is calipers, methodology, and math. With just that, they can determine the amount by which we are uncertain about the truth of any given proposition and how uncertain we are about the uncertainty. Of course, that's not God's perspective; so God really doesn't play a role in science, unless one is testing the God hypothesis, itself. That's why science is fundamentally amoral.
Where scientists can go wrong is when they begin to make pronouncements about philosophy and values, as scientists. In that case, they are just like anyone else: The values they bring to the table are only as good as the values they bring to the table. Sometimes good and sometimes not so good--Stephen Pinker being a great example of the later.
When scientists start trying to derive values scientifically (or slide them in thru resort to their AUTHORITY as scientists), we're back to the Sin of Adam. IMHO, there is no such thing as a scientifically derived or supportable value. One can make a statement such as, "Sexually promiscuity in a marriage leads to a 50% higher probability of divorce"--probably roughly a true statement. That sounds like a scientifically supported value statement about the undesirablity of promiscuity. But its value content depends on on the value one places on not getting a divorce, which is not, in this statement, scientifically derived. If you then gather evidence about the effects of divorce, you introduce new, and unscientific, values in an endless recursion.
The scientific method and its accomplishments are magnificent achievements. But we know there are limits to what science and math can know (see LaPlace, Turing and Goedel, e.g.). Simply put, science and math cannot speak to many "scientific" issuse (the orbital trajectory problm r9 spoke about earlier), let alone speak to values or morality. It is only when science and values get confabulated--often intentionally--that things go awry.
Scientists don't, when they do science. All they need is calipers, methodology, and math.
I'd add something else to this list: assumptions.
I'm often reminded of something Einstein said about his development of Special Relativity, to the effect that the first and most important step in his work was to scrutinize the assumptions underlying Newtonian physics; as I recall it, the culprit was a Newtonian assumption that rate of passage of time is invariant.
Once I started looking for this sort of thing, I was surprised how many "scientific" statements are actually statements of what scientists currently assume, and which have attained in their minds the status of physical laws. (Such was the case for Newtonian physics, for example.)