Thesis: Information Overload!!
I’m sure all but the most organised of you have experienced something similar, even if you’ve subsequently solved the riddles and moved on!? I was hoping to get some advice that might help me make better use of my time over the coming two months.
Having extended my TMR4 deadline, then had the audacity to take a weeks leave at the end of February, I’m now beginning to feel the thesis crocodile circling closer and closer to the canoe. I’m extremely conscious of upcoming deadlines, but having gone out and collected as much data as I can, I now think that I’ve got too much and I’m struggling to refine it into useful material of the right academic level.
My thesis is on the structure of the Royal Engineers, and how we might consider re-structuring ourselves in order to better deliver effect on Operations. My basic premise is that the current historic structure (Platoons, Companies, Battalions, Regiments etc.) was/is fine for fighting wars, but pretty useless for undertaking construction tasks, not only does it seem to be an ineffective way of managing skilled tradesmen, but it also appears to leave managerial skill gaps within the command structure.
My main issue is that having obtained no less than 20 PXR/POR/PERs from the TICRE (luckily there was plenty of stuff they didn’t have or I’d have ended up with even more), together with the latest Project ANEMOI Lessons Learned document, I’m currently sitting on hundreds, probably thousands of pages of text (of varying quality and command level) that I’m attempting to evaluate in order to find reoccurring themes and patterns. I was keen to avoid any kind of selection bias by simply choosing the reports I liked the best, but in doing so I’m finding it almost impossible to pick out the most useful (and relevant) lessons learned. I could write a whole literature review on just three or four of the reports on their own, so having complied a spreadsheet of lessons learned from all 20, I’m struggling to condense them into useful (academically sound) results. So far, all I seem to have done is hand select obvious anecdotes with little further analysis. It’s like finding passages in the Bible or Koran; search long or hard enough and you’ll find something that backs up your opinion. The rest can just be discarded, even if there are a similar number of anecdotes contradicting the first. It feels like I’ve almost gone the other way and that having too much information is creating its own bias by allowing me to pick and choose what I want rather than what is actually relevant or meaningful.
Did anyone else have similar issues with regards to information overload, and if so what strategies did you employ in order to focus your efforts and obtain useful, valuable data from the sea of literature available?
PS: Please no “I’ve almost finished my thesis you should have started earlier” comments!
Holdfast Training Services Trial Blog 
Kukie
I’m having similar troubles. Having acquired a thesis title with about an equal amount of qualitative and quantitative data I am finding it tricky to hang my analysis on something tangible.
Mark Hill had a good suggestion of looking at ANOVA (Analysis of Variance) to try and pick out trends when there are multiple variables. The Wikipedia page has a decent amount on it, although it helps if you have some numbers to pay with. Other than that I have found it useful to try and turn words into numbers using scoring and weighting systems, then applying some statistical tests – again loads of places where it could go wrong.
Best of luck. If you are like me, you’ll need it.
Kukie,
It’s an interesting ‘trap’ you find yourself in, much like I did during my thesis – What are the key challenges faced by the Corps of Royal Engineers in implementing BIM for Infrastructure projects on operations, can they be overcome and how?
I use the term ‘trap’ as that’s how it felt having to write a purely qualitative thesis, conducting analysis based on the subjective views of others. It’s much easier to compare numbers!
However, as Mark P mentions applying a weighting, based on a priority order, to your literature material is a good way of going about sifting through the mass you have reamsearched.
Using Project ANEMOI as an example, as I too used, is an obvious one because it probably the most current you will find. Not least it offers you the opportunity to interview those who wrote the points made in the PXR (for that particular phase of course). This then gives you an added reason why you have chosen to use that material and offers another media of research other than just another report.
In addition, and I don’t know where you are currently based, but there is an ICE event on in London (8 Mar) titled – ICE and InstRE joint professional lecture: Celebrating 75 Years of the Bailey Bridge: The Past, Present and Future of Bridge Engineering in Military Operations. This may help and again back up your initial research.
Hope that helps.
Fran
PS. I have finished my thesis – about a year ago!
Kukie, I feel your pain. I think what helped me is decide on the structure to the thesis and methodology. What are your aims? What hoops do you nee to jump through to get to those aims. Break it down and then break it down some more.
As a fellow dyslexic I know how it can feel overwhelming and how difficult it is to get clarity. It might help to talk it through with your mentor but I’m happy to FaceTime. (dougnelson33@hotmail.com). I have found speaking to John M really useful. It sounds like Richard may be your mentor so maybe speak to him about methodology.
Regarding information you could try to classify the projects by type or complexity. Quantify their success and problems faced. Were they high or low risk. Set you self a list of questions that you need answered and then look at them.
Your project is really a management topic how do civvy street organise and why what are the project team structures, functional, matrix etc. In order to change you need to see where we are failing and how it would be better to change. You need to consider DLODs (Tepidoil) as you are generating a capability.
Anyway happy to chat when it’s convenient but, you are not alone.
Kukie,
Great idea for a thesis, sounds genuinely interesting.
I was sort of at the other end of the spectrum for a while and found I couldn’t get hold of the information I wanted – I ended up finding a few key people who knew the topic (thanks Mark Palmer for putting me in touch with one of them) and spoke to them – arrange an interview or by email if you can.
I know you have a lot of information – but what it will help you do, is eliminate some of the information that may not be relavent. Appart from that – my advice is just keep putting information down and then reread it about 10 times – editing it all the time.
Kukie, I don’t want to step on your mentors toes, so please take this advice as “for what it’s worth” rather than formal guidance. All of the advice above is excellent and will no doubt be useful at some point during the writing of the thesis. The issue that I sense is at the heart of your post though is the identification of themes that you can then apply the suggested techniques to. Mark’s suggestion of ANOVA is an excellent technique for achieving this where you are working with statistical data. If however, the base information is textual you may be better applying Pareto analysis. This technique is often used in the diagnosis of engineering problems to determine the best course of remedial action from amongst a number of competing alternatives.
Effectively what you do is identify issues that can be grouped together under one title and then see which of the titles has the most number of occurrences. If I am analysing why an escalator is failing this would involve recording the different reasons for non operation over a period of time, then counting and ordering the reasons for failure by frequency of occurrence. The most numerous reasons are then determined to be the most significant, so that is where your investigation initially focuses. There is a lot of good stuff on the net explaining the technique, but it isn’t rocket science, so the Wikipedia entry is probably all that you need for this particular problem.
Jim
Limiting scope through introduction/aim and use of thematic analysis is all I’ll add, although I might just email you directly!!
For those of you exploring qualitative data then you really have two main avenues to explore:
Mixed methods – the classic weighted approach where both a qualitative pass is undertaken, followed by a quantitave pass. Once each of the main “themes” has been given some kind of score, I’m sure you’ve all done something like this during your attachments. The crucial bit is to make the score realistic for each element and to justify that weighting through [previous/current] research. That’s really where ANOVA comes into it (although its easier to apply it to quantitative data as it highlights effect size in controlled experiments rather well (but don’t get me started on the design of experiments!!)
Kukie. One way of qualitatively searching through the thousands of pages is to employ a filtering method which looks for keywords, the use of word maps to search for popular phrases often brings up interesting phrases that can help you focus on the important elements or key themes. A simplistic method of doing this may be to use “wordle” (http://www.wordle.net) which produces lovely “word clouds” – damn I hate qualitative analysis!
If nothing else word clouds highlight common themes; given the documents you look at all originate from the same cultural background there shouldn’t be too much noise in the results and then you could consider focussing in on these themes or phrases, as Chris says the key is to eliminate the irrelevant but to do so would require some justification – the size of the word cloud/map/bubble can be used to justify that.
Regardless of method used the ability to filter out irrelevant information is probably one of the biggest challenges, good luck!!