这段话的论述问题不大,但是所得结论(红字部分)值得商榷。
举个例子说吧。
几乎所有自然科学的目的都可以概括成是为了发现某些规律。 而如何设计试验,如何分析数据,这里的基本数学工具是统计学。统计学的根本问题是hypothesis testing. (sorry, I am going to continiue in English, to be a little faster).
The problem of hypothesis testing is to decide whether to "accept" a hypothesis given the data observed or collected. The word "accept" is in quotes, since we do not really accept per se, "accepting" a hypothesis actually means "not rejecting" it.
First of all, what hypotheses to use as the set of candidate hypotheses itself is a question subject to philosophical debate. For example, at time of Issac Newton, the Newton's Law that F=ma was a hypothesis which Newton established and "accepted" via experiments. Why use this hypothesis? why not F=m^{1.00000001}a (namely, F is m to the power of 1.000000001 times a)? If this was to be used as the hypothesis for testing, then with the measurement accuracy at his time, the hypothesis will also be "accepted" with equal and possibly higher confidence (the latter of which would be due to statistical irregularity). But Newton chose the hypothesis F=ma for testing, not the other, or not any other hypotheses -- there are infinitely many such hypotheses that would pass the experimental test.
There is a principle, called Occam's Razor, (
http://en.wikipedia.org/wiki/Occam's_razor ) which is a philosophical dogma that crosses the boundaries of various sciences and which can be regarded to be in the regime of philosophy of science. The principle states, roughly, that one should make the "simplest" assumption (hypothesis). This principle is widely accepted (well, perhaps still subject to controversy at times).
In any area of science, we do not really discover anything that is "the truth", we only claim that our hypothesis can not be rejected with the data we collect and observe. What hypothesis to beggin with is at the very foremost front of any scientific discovery. And different philosophical principle we choose to use would govern the outcome of what we claim as "truth". -- How could we say philosophy has no impact on science?
So far, my example has only covered the choice of hypotheses. In fact, even given the same set of hypotheses, and given the same set of observed data, the choice of "philosophy" used to determine which hypothesis to claim also governs the outcome of the scientific discovery. Some of such philosophies stay in the regime of statistics, and some go beyond and penetrate into the regime of philosophy of science, which I choose to ignore any detailed arguments here, for simplicity .... (by the way, in my opinion, statistics, if you know it well and deeply, in fact sits in between mathematics and philosophy... again, too much to be said there, which, if said, would be a digression ...)
综上所述, and skipping some details, the philosophy of science directly impacts how we should preform scientific experiments, devise our hypotheses, and claim the outcome of a scientific discovery, where literally, I take the "outcome of scientific discovery" to mean "knowledge" as in "哲学并不提供有效的实际知识".