I like to find science news before newspapers and other news media cook it to be more palatable to readers. Why? To begin with, the earlier reports are more likely to be accurate; a newspaper story might become inaccurate when "improved" by the reporter, the editor, and the copy editor.
Also, I don't trust general news media to know what's worth reporting. Is "The White House now has a dog" more important than all the stories which don't get on newspaper front pages? More important than everything which doesn't get reported at all?
And, of course, earlier reports are earlier. Sometimes years or decades earlier.
The downside? Many experimental results aren't confirmed. (If all medical discoveries were real, there'd be no more diseases left; and laboratory rats would live for thousands of years.)
My main source for science news is EurekAlert.org, which publishes scientific press releases. The press releases are simplified enough for newspeople to understand, which means I can usually understand them. Sometimes it's possible to download the complete original article. When that's restricted to newspeople, one can usually download an abstract.
Note: Some scientific journals aren't very scientific. There are journals for researchers who believe human races differ in intelligence, to begin with.
If your level of scientific knowledge is high enough, see arxiv.org for abstracts. These are written for people who understand a particular field, so (for example) a physicist might not fully understand a quantitative biology abstract.
For easily-understandable explanations in paleoanthropology, genetics, and evolution see johnhawks.net.
What are your favorite sources?