Before you evaluate a source, it helps to know what kind of information source you're looking at. Talk with your advisor about the types of sources you should seek and avoid.
In general books summarize and contexualize original research articles, but aren't considered primary or original research. However, books are sometimes published as collections of original research articles that were published earlier in scholarly journals. Books can help you understand how a research experiment fits within a larger context. Science encyclopedias and dictionaries, written for non-experts, are useful when reading and trying to understand rigorous texts.
You're probably pretty familiar with news articles. They tend to be short pieces written by journalists, not physics experts, and report recent developments in the field. You''ll find news articles in popular publications, trade magazines, and scholarly journals. You won't necessarily cite these in your research, but you might use them to learn more about a topic.
You may also be familiar with editorials, or opinion pieces. These tend to be well-researched opinions with or without citations, and can appear in popular publications, trade magazines, and scholarly journals. Since these aren't original research, you might not cite them in your papers, but you could use them to learn about the nuances of a particular experiment or field of research; they might influence the directions you take as a researcher.
Popular publications like Time magazine and the New York Times publish short articles written by journalists for the general public. You won't cite these in your research but if they parallel specific research articles, they might help you comprehend the scholarly texts.
Trade magazines, sometimes called trade journals, find a middle ground between popular publications and scholarly journals. Articles may be longer than those in popular publications, may include a few citations, and do not have the rigor associated with scholarly articles. Readers are professionals in that trade.
Peer-reviewed articles -- sometimes called research articles, primary research articles, original research articles, and refereed articles -- are written by and for experts and published in scholarly journals. Unlike popular and trade articles which go through an editorial process, peer-reviewed articles go through a rigorous peer-review process. In short, an author submits an article which is refereed by one, two, three, or sometimes four reviewers. These reviewers are experts in their discipline and peers of the author who is also an expert. Typically the review is "blind" meaning that the identities of authors and reviewers are hidden from one another. These reviewers offer recommendations to the author and may approve or reject an article for publication. Some journals take pride in their high rejection rates.
Review articles, published in scholarly journals, review and summarize developments in a subfield. While readers may be experts in a larger field, they may lack expertise in the subfield and use these reviews to stay informed. Review articles often appear similar to original research articles. To distinguish between the two, you'll need to determine if the authors refer to their own work or if they discuss the research of others.
Pre-prints and post-prints refer to scholarly literature before (pre-) peer-review and after (post-) peer-review. Sometimes pre-prints and post-prints are lumped together and considered the unofficial version of a scholarly paper; you can download them from author websites or repositories like arXiv and they'll look like a simple Word document. Although they are unofficial documents that may be riddled with typos, readers value this quick access to new information; the peer review process and publishing both take considerable time.
Conference proceedings are the written versions of research shared at conferences. Experts present their research, and learn from each other at these conferences. Proceedings are sometimes peer-reviewed.
Dissertations and theses are extensive research projects undertaken while working toward a master's or Ph.D. Although thorough reviews are done by an advisory committee, dissertations and theses do not undergo a formal peer-review process. Although dissertations and theses may be excellent sources of information, only a few print copies exist and libraries are unlikely to lend them out. They are best obtained through online repositories or directly from the researchers themselves.
Understanding the flow of information in the sciences may allow you to better evaluate and select sources for your research.