Thursday, 17 July 2014

An oldie but a goodie

Several months ago I came across a reference that I wanted to read:

Price, D. J. de S. (1956) 'The exponential curve of science'. Discovery, 17(1), pp.240–243.

The subject matter of the paper is of little concern for this post but the journey I had in locating it is noteworthy for those reliant on inter library loans or resource discovery services (RDS).

When I refer to RDSs I am meaning new technologies within the library profession that facilitate a single search approach to finding academic materials. Gone are the days of having to look through three or four databases relevant for your subject/topic as the RDS does it all for you in one place. The University of Bedfordshire has an RDS in DISCOVER (lrweb.beds.ac.uk/discover).

Going back to Price’s (1956) paper, the University of Bedfordshire library did not have a copy so I did what all good researchers do and requested it via inter library loan (UoB call it document supply and it is a great service (lrweb.beds.ac.uk/libraryservices/docsupply)).

Unfortunately I got a reply from the British Library stating that they could not find the article even though the information I provided was accurate. The British Library provide an excellent (but automated) service so I started thinking about why this has happened?

I decided to check on the British Library catalogue convinced they should have the paper.

A title search gave me nothing.



But remembering a colleague of mine facetiously rambling “it’s the quality of the metadata” what dawned on me was that the age of the paper (1956) may be too old for there to be metadata about the paper I want. Database providers may not have the resource to provide metadata for old papers, in effect meaning that the only way to know about Price (1956) is by the old ways (checking references lists of papers and browsing through catalogues) but not the new ways (relying on RDS’).

So if I can’t search the article title, I was hoping for some luck searching by publication.



My first thought was that the title of the journal (Discovery) was too generic a word. According to the British Library records there are 35 journal titles beginning with “Discovery”. My solution at this point would be for a visit to the British Library itself and look one by one until I get lucky. This may sound like too much effort for some. The new ways in searching for words and phrases in databases is much of what librarians teach. My solution at this point would be for a visit to the British Library itself and look one by one until I get lucky.

The RDS the British Library has is good but for this particular information request it was the remnants of the British Library catalogue that helped me for an article from 1956. I also looked for the catalogue of a top university library and noticed that the library catalogue had disappeared and had been replaced by an RDS. A researcher may become conditioned (due to the successes of RDS’) that a metadata search will be sufficient (let us call this Plan A) when the Plan B of relying on browsing, looking at references and spending more time for the search process can lead to success.

An accompanying argument can be made that old articles are old and should not be used when compared to more recent articles. Academic librarians work in a system that supports that type of thinking but a more sensible thought could be that it depends on the circumstances. Price’s (1956) paper is perfect for me but it has been buried due to the popularity of new methods getting in the way of old methods.

Ask yourself how reliant you are on Resource Discovery Services dictating your information world and how old journal articles may only be located via the references of papers.








Friday, 30 May 2014

Fun with words


As many other people have done, I went to Google to better understand the meaning of a word (for me it was the word 'discourse').

I went to Google and typed in the following:


I will confess to being none the wiser about what discourse was meant to mean with the Google definitions but what did get me curious was the downward arrow that I could click. Having done so, I was surprised to see the following:


What we are seeing here is a time series chart visualizing the use of the word ‘discourse’. I stopped caring about definitions of the word and started thinking about why this particular word encountered peaks and troughs when it came to its use.

I tried several other examples.



I like this one because it shows how a new word has gradually increased in use.


A very sharp rise here!



This one is my favourite because etymology concerns itself with the origin and development of words. How ironic that over time and space its use has diminished.

I suggest all of you have a play and look for words that have this facility (not all do) in Google. It can be a fun exercise.

My ulterior motive is that it will get you to think about the interpretation and popularity of words and how they can change. New words eventually become old words and get replaced with new words we all start using. Old words can sometimes make a comeback but often accompany a reinterpretation or require an incident. 

With my librarian hat on, perhaps you will also start thinking more about the search terms you put in databases. Librarians often comment about synonyms and having to do OR searches in boolean. The addition I would like to add is to start thinking about the impact of the words you are including/excluding when thinking about search terms for your topic of interest.

Monday, 17 March 2014

Some thoughts on a citation classic


Whenever I try to explain what a citation classic is it this image that comes to mind.













[For any of you that are unfamiliar with Cliff Richard’s Mistletoe and Wine, I suggest a listen on youtube]

Very few British people are unfamiliar with this song. I also dare to say that any ‘Best Christmas Songs Ever’ compilation album will lose credibility if Cliff’s song isn’t there. Like it or not, it’s a classic.

At this point I return to the citation.

When researching a thing, the literature review process will include the possibility of the thing you are researching including citation classic(s). In other words, scholarly materials that are so often cited that you may lose credibility for not having them on your references list. It actually does not matter if you like/agree with the citation classic or not. It is about respecting its constructed position because peers will be expecting it there. Not having a citation classic could be harmful so if an aura of power surrounds the citation classic, bung it in. How you cite it can be up to you.

To demonstrate (please click on images):

Articles with cocitation as Topic according to Web of Science (March 2014)




















What I have done in Web of Science is a topic search (so that means it is somewhere in the title, abstract or keywords) for the word cocitation (which is the thing I am researching) and ranked results according to the number of citations. Henry Small’s paper on cocitation is at the top which makes me think it is a hugely influential paper.

I decided to be a bit of a geek and perform a citation analysis of the 592 papers itself. I noticed the following:











What the list above tells me is that Small’s paper is at the top of this tree as well (40% of the sample citing it is pretty darn strong). There was also overlap between both lists and I became more informed about Authors in this field.

At this point I will confess that by having some knowledge with the subject matter of cocitation, I already knew that Small (1973) was the star at the top of the tree. Most of us looking up a thing will have some knowledge (whether it be tacit or formal) and the citation classic is not a hard thing to work out under these circumstances.

It is however much trickier to work out when entering a domain where your knowledge is somewhat hazy.

This time I am going to go to Google Scholar and write (as phrases) two references plus the author names. I do this because I am interested in this particular references pattern plus my belief that the two references encapsulate the thing I want to research



















The screen grab shows 6 of the 78 results. The third result (Luukkonen, 1997) is on the money for what I wanted to look up. Luukkonen (1997) is not the most cited paper out of my retrieved sample (it is the 11th most cited thing out of the retrieved 78) but the most accurate to my particular interest.

Thus, I have constructed a search to create what I interpret as the current citation classic to my particular interest (that being papers on citation analysis that place emphasis on Latour’s Science in action as well as cite Gilbert’s paper that I consider significant).

In short, ask yourself if there are citation classics in relation to your interest and consider their inclusion. Also consider what techniques (like those suggested here) could help you locate them.