Monday, January 24, 2011

How to handle a large amount of literature

These days, I am less busy in the lab (my setup needs to be changed so I won't be carrying out experiments for about a month), so I've totally dived into my literature review again. Since I am trying to get a good insight on several topics (existing bridges, load distribution, shear in beams, shear in slabs and failure criteria from fracture mechanics) I have gathered a depressingly large amount of papers which I want to work my way through.

Bit by bit I've developed a way to tackle this task.

1. Read by theme or period of time

Initially, I thought that it would help me if I'd be reading several topics parallel. Now, I've found out that it helps me much more to see the connections between different papers when I gather information on the same subject and read everything I can find on the same subject until I've finished that topic. Every now and then I do feel tempting to vary topics or required level of understanding for the considered papers. I tend to alternate papers on experimental work with papers on purely theoretical work.

2. Speed reading

As I wrote before, I've been doing some attempts to speed up my reading. I still feel tempted to savor each and every sentence, but working with a stopwatch has made me more conscious of my time.

3. The egg timer trick

I'm now using the old egg timer trick to track my time. Instead of the egg timer I am using the stopwatch of my phone (a 6-year-old fridge-shaped Nokia), to see how long I can concentrate before I drift off and go walk around the hallway for another cup of tea. I'm trying to stretch my concentration span from 20 minutes to 45 minutes. My attempts to regain focus through meditation seem to have a positive effect on this.

4. Summarize

What was special about this paper? What is the main thought I need to capture? I jot down a few words on the first or last page of the paper for further reference. I also make some notes at the sides of the columns whenever I come across an interesting thought, but I feel it is also important to get the key message and write that down for later in order to avoid having to reread the paper.
 

5. Archive

I use Endnote to organize my references. Most of the papers I read can be found through the online version of Endnote, so I can import all the relevant information to the data entry and attach the paper to it. Searching, as well as looking how much work of a certain author or from a certain period of time I have read, have become much easier this way.

6. Use the relevant data right away

Whatever I need for the report of my literature review goes in the right section of the document - straight away. I also keep some separate files on the influence of different parameters on shear. I don't have any written text in those files yet (the literature review needs to be done first), but I copy the most important graphs and sentences about the studied parameter, together with the reference in those files. I expect that this will speed up my writing process significantly once I start writing about these parameters. In the end, I will only have to mix all the information together, and then discuss it with regard to my own experiments, instead of looking for that information in papers again.

Admittedly, I print out all my papers (reading on the screen hurts my eyes, and I just am a paper and pencil person), so the physical papers themselves also have to be stored. That system is not as good as my Endnote library - I find myself often going through binders, trying to remember under which keyword I stored the paper I need. Any advice on this would be very much appreciated!

7. Let the ideas melt together

While reading all this information, I feel like all ideas are melting together in my head. I plan to start drafting some mindmaps rather soon, to get the links between these ideas and to point out what I will use for my own theoretical work.



   

23 comments:

  1. Eva-

    This is quite helpful; thank you for sharing your strategies.

    This is similar to the methods I have used before, though now I am trying to do all this electronically. Quite a challenge, though I can no longer manage all the huge piles of papers.

    Look forward to hearing more!

    Jeffrey

    ReplyDelete
  2. What software do you use to make this process electronically? And how different is it from the paper-and-pencil approach?

    ReplyDelete
  3. Great article., Eva. Really similar to my approach - I have tried online reading and do it occasdionally but I just prefer having the paper in my hands and handwriting notes, I find it easier to remember things that way - I think because I am quite visual so seeing handwriting helps.

    Re your paper filing - I was doing it by keyword but recently switched to alphabetically by primary author. This is working well because I can use Zotero (which I use instead of Endnote) to find articles by theme and then look for the paper version by author and it means I don't have to rack my brain as to the subject I filed it by!

    Hope that helps...

    ReplyDelete
  4. My approach is rather haphazard - I really need to start being more organised. Thanks for the inspiration!

    ReplyDelete
  5. I'm very interested in how people store and manage all the information they find and generate during PhD research, so thank you for this post!

    Regarding filing your hard-copy papers: you could file them by author name, as Bex suggests, and/or broadly by theme or strand, colour-coding the folders appropriately. Alternatively, you could tailor-make a searchable (and auto-completing) 'keyword' field in EndNote and tag your EndNote records with the same keywords you associate with your printouts?

    ReplyDelete
  6. Thank you for the ideas! I think I'll go for the alphabetic ordering.
    I already use a color code, but all my literature has one code to distinguish it from codes, courses and my own experimental work :)

    ReplyDelete
  7. This post is a great resource Eva. I may write up something that details my own workflow -- very similar to yours in a lot of ways. My own workflow is more tied in to LaTeX/BibTeX.

    Here's another idea regarding paper filing, I can't remember where I heard this before but I think it's a cool idea. Basically, you create your own random indexing structure. It could be something as simple as sequentially numbering them. In Endnote, you document that paper's #, so when you need to look it up, you just go to your file drawer and pull paper 54. Every time you read a new paper, it gets the label n+1 and is added to your archive. This way, you don't physically group papers according to content -- I find it almost impossible to physically organize papers because of content because I often find myself wanting to place papers in multiple buckets.

    ReplyDelete
  8. I'd be really interested to read about your workflow. I never had any more advice here than "you need to make a literature review and publish it as a research report", not even an explanation how our library system works, which I think is quite important information for all graduate students of a university.

    I think I'll use the alphabetic ordering of my papers. Most of the time I have my Endnote library organized alphabetically as well, and I've noticed that I quite like to have the material of the same author together.
    The only drawback for this system would be that I'm sorting by first author. Some work (like the background for the modified compression field theory) is published by (in this case) vecchio and collins, but more recent work is PhD student (first author) + the senior authors.

    Thanks for the good ideas, again!

    ReplyDelete
  9. Hi Eva

    Came across your blog through serendipity while trying to track down some conference proceedings that I needed. I'm currently undertaking an MSc in Civil Engineering, and studying the effective lengths of torsionally braced plate girders.

    Found your tips really helpful, and thought I'd contribute my own system.

    Like you, I use Endnote to attach all the papers that I need, and always access PDF's in this way. After a year of continual use, I'd find it incredibly hard to sort out my references and bibliography manually.

    I also use foxit PDF reader to add bookmarks to all the important points that I need (going through hundreds of codes of practice to find the relevant clause is a nightmare!) Foxit is also useful where figures are on awkward pages, as I can simply cut and paste them around the PDF. And of course, highlighting is absolutely essential.

    Great blog and advice, keep the posts coming. Good luck with the research : )

    Pete

    ReplyDelete
  10. Thanks Pete!

    I hadn't heard of Foxit before, but I'll look it up and try it out.

    Good luck on your studies and research too :)

    ReplyDelete
  11. This comment has been removed by the author.

    ReplyDelete
  12. Good luck with finalizing your literature review!

    ReplyDelete
  13. Really interesting. I use EndNote but I also use NVIVO to manage my literature. It's really for managing qualitative data but I've also found it works brilliantly for managing literature. I read each paper and created a document for each using author name and year. I then coded each paper by themes (the great thing being you can have parent and child nodes for your themes so you can see where they link). You can just code sections of each document to the relevant theme.
    500ish papers later, not only did I have typed notes on each, I also had the titles and subtitles of my lit review which were the codes I had created. Worked really well. It was the lady on the NVIVO training course who suggested it but I have now got quantitative research students using it just for this purpose!

    ReplyDelete
  14. That's a great idea :)
    Thanks for sharing

    ReplyDelete
  15. Do you need to buy NVIVO or can you download a free copy of at least part of the package? I have used Freemind which is a little basic but still allows you to create parent and child nodes in a concept/mind map format. The only issue I have found so that is that I haven't worked out how to create links between the different nodes.

    I think this is a great blog with lots of excellent ideas! Thank you.

    ReplyDelete
  16. Hi Anon
    NVIVO is quite a pricey bit of software :-( but if you work/study anywhere with a social science or psychology or education dept they may well have a multiple license. Check your institutions software library! You can get a free 30 day trial.
    I think the mindmap software can do this but the beauty of NVIVO is being able to double click on the node and bring up a whole heap of relevant text - fantastic. I will do a blog on this I promise and I will post the link here when it's done!
    Mossposs

    ReplyDelete
  17. Hi everyone,

    I use Reference Manager for generating bibliography; however, handling all the PDFs of papers or whatever is very easy to do in Mendeley Desktop, and it is quite useful also for the writing process, as you can search for a specific concept and all the papers containing it will pop up.
    I organize my printed papers by First author's last name, after a filtering by year (before and after 2008 in my case).
    good luck to you all,

    Sara

    ReplyDelete
  18. Hi Sara,

    Thanks for your input. Your approach is similar to what I'm doing now, except for my using EndNote instead of Mendeley...

    ReplyDelete
  19. Hi!

    Nice approach Eva...my main problem is to stick to the "plan"..sometimes I go off the rails.. :p
    I also use the Mendeley Desktop, as Sara. However I generate my bibliography with the same application.. and my papers are organized by main themes/projects (into files) >> first author's last name >> year of publication
    Well I think the aim is the same, we go in the same direction, by parallel tracks! :D

    "First comes thought; then organization of that thought, into ideas and plans; then transformation of those plans into reality. The beginning, as you will observe, is in your imagination.." [Napoleon Hill, american writer]

    ReplyDelete
  20. Hi Otilia,

    Thanks for your input! I also tend to go off on tangents and start reading on remotely related interesting topics, but when I have a clear schedule for myself, then things work out much better.

    Great quote as well :)

    ReplyDelete
  21. Hi Eva,

    Wrt physical filing: Why not file all articles alphabetically? It seems to me the easiest solution, and it actually works. In the upper front corner, you want to write down author and year as to find the paper even more quickly.


    Sry if someone already suggested that, I admittedly didn't work through all the comments.

    "Admittedly, I print out all my papers (reading on the screen hurts my eyes, and I just am a paper and pencil person), so the physical papers themselves also have to be stored. That system is not as good as my Endnote library - I find myself often going through binders, trying to remember under which keyword I stored the paper I need. Any advice on this would be very much appreciated!"

    Best
    V.

    ReplyDelete
    Replies
    1. Thanks - that is exactly what someone else indeed suggested in the previous comments, and I've been keeping my hard copies in alphabetical order ever since then :)

      Delete
    2. :) Oh well, I tried hard not to procrastinate too much with your blog, hence didn't go through all the comments... It's a very enjoyable read!
      V.

      Delete

UA-49678081-1