I am planning to do a little (maybe not so little) research concerning the influence of academic and non-academic publications (separated, if possible) on policy. In my case, I care about the indigenous movement and the creation or enhancement of "political opportunities" or the access to resources and how it is influenced by research - via State agencies, for instance.
On a recent conference, I came across a method rarely used in the social sciences that could be of use for this research: bibliometrics. With tools like Google Ngram Viewer (http://books.google.com/ngrams) or jstor Data for Research (http://about.jstor.org/service/data-for-research), you can find out easily and quickly, when a certain term appeared first in publications and how much it was used when.
Nevertheless, those tools have their limits - especially if you're doing a rather specialized research like me.
Is there anyone who has experiences with bibliometrics? Anyone used this method already? Maybe someone knows a better tool than the above mentioned?
Instead of using the Ngram Viewer or tools by other single providers of databases such as JSTOR where the criteria of inclusion are more or less intransparent and possibly biased, corpus linguists prefer corpora of texts that have been sampled according to more elaborate procedures and that are balanced with regard to different criteria (see https://en.wikipedia.org/wiki/Category:Corpora for a list of some of the most important corpora).
Dear Benjamin, thanks for your suggestion - nevertheless, the corporas I tried did not help me at all. Maybe you (or someone) could suggest a certain corpus that works fine for social science purposes?
Dear Philipp, I can recommend a couple of academic sites:
SCOPUS by Elsevier sciencedirect
ISI web of kinowledge by Thompson Reuters
These cites both have built in bibliometrics.
It’s an interesting question, but difficult to answer. With regards to UK- and US science policy, I encountered basically the same problems as you. It is quite rare, that scientific reports were cited, and it is also quite difficult to measure the effects scientific expertise on policy. I recommend that you combine network- and discourse-analysis in order to answer your question.
In order to address this problem, we are trying to use a more indirect approach: First, while crawling through the hearings, speeches and documents of the respective parliaments, certain scientists, their affiliation and their profession were mentioned. My colleagues and I utilized python and the APIs provided by the US and UK-government, downloaded all files. Now – as the second step – we are reconstructing their political stances, additional affiliation (i.e. affiliation to corporation and think tanks), and the arguments given in the political debate by the scientists. In a third step, we’ll check whether or not the arguments used correlate with the argumentation used within the scientific community (in their papers published) and with regards to other affiliations (i.e. corporation and think tanks).
- Badges
- Science method
- Similar topics
- Methodology