 |
 |
On Aug 15, 2013, at 11:21 AM, Ahmed Abou-Setta, MD wrote: > An interesting paper just came out that rekindles the question on the potential use of Google Scholar in healthcare decision-making. > Great! Another paper to go over VERY carefully at our next Reference Librarians' meeting here. First/quick assumptions: ===> McMaster U. is involved, so I have to take this seriously. ===> They looked at nephrologists searching in Pubmed "vs" nephrologists searching in G-Scholar, so MeSH-based searching is not going to come into play very nicely, except when Pubmed translates the query successfully (which is only a "sometimes" sort of thing). Pubmed will do better with me at the wheel. This is not to imply that I don't love G-Scholar. I do. I used it every day. I just can't explode MeSH terms, use subheadings, limit to publication-type/study design, employ the myriad tricks up my sleeve, download more than one-at-a-time, or properly document my search-strategy & imagine it to be reproducible. ===> The "amount" of articles doesn't necessarily impress me. We know the Google-Scholar is huge. We know that it frequently searches the full-text/PDF itself (which Pubmed won't do) (and which will get good stuff AND lots of noise). We know that the reported number of hits in any sort of Google search is not ever precise. Anyway, I've gotta print it out, get coffee, hunker down... Tom Mead Biomedical Libraries Dartmouth College Hanover, NH 03755