This experience lead me to reflect on the whole process: what made it so erratic, why would it take longer than expected and why would it be so frustrating? I came up with the following explanations for myself:

1. unspecific guidance: everybody has heard of a “proposal” and you know that it is meant to be written before the actual thesis. But how to even get there? How to start, what to do in the middle? The time from start to proposal can easily cover a year and requires particular activities and techniques – find them out and name them
2. The deliverables are not clearly defined: so you look up some literature, how to structure it? What about these ominous research questions, how do they look like? What is a topic and when do I know that I have found one? At least for me it is very hard to try to reach an unclear goal.
3. Heuristics / strategies / quality gates: How to decide between alternatives? How to judge their feasibility? Again, if you have no clue what to look for, how do you know when you found it?

To find out more about this I went through a number of books and articles and for the most relevant sources I want to share  what I consider most useful about them.

The sources

I have reviewed several books and online articles. Some focussing on academic writing project in general, some focussing on the early stages (up to a research) proposal for computer scientists.

The books

• Booth, W. C.; Colomb, G. G. & Williams, J. M. (2003), The Craft of Research, The University of Chicago Press.
• Davis, G. B. & Parker, C. A. (1997), Writing the doctoral dissertation: a systematic approach, Barrons’ Educational Series, Inc..
• Kruse, O. (1999), Keine Angst vor dem leeren Blatt, Campus Verlag.

and the online articles

• Lauer, H. (1975), ‘Discussion on Ph.D. thesis proposals in computing science’, The Computer Journal 18(3).
• Levine, J. S. (2004),’Writing and Presenting Your Thesis or Dissertation’, eBook, http://www.lulu.com/product/e-book-download/writing-and-presenting-your-thesis-or-dissertation-%28ebook%29/96916. (Also available at: http://www.learnerassociates.net/dissthes/index.htm)

The process of writing a proposal

All sources except Davis and Parker discuss the writing process at least up until the proposal itself. They all more or less assume that there are the following steps

• an orientation phase
• a literature research phase
• a selection or refinement phase
• the write-up of the proposal

Kruse describes the orientation in much detail, he discusses particular questions that any research should ask him/herself when starting a new, big project, how to narrow a subject down, what types of projects there are and much more.  The book looks at all of this from a “writing” point of view, where the act and skill of writing is the vehicle to generate and order thoughts and bring them into a shape that can be used to inform others. Lauer takes the opposite approach and looks at the PhD process as a project and gives appropriate advice on how to manage the different stages.

The literature research phase is presented similarly in all the four sources. The literature review for a proposal is supposed to be comprehensive but not exhaustive (which it needs to be in the thesis itself). Kruse recommends to do a very superficial but broad literature “scan” in the orientation phase (including just reading headlines and abstracts), in order to get a feeling for topics and questions, the amount of potentially relevant material and an idea about who are names to remember in the field.

Deliverables and Gates

Even more important (at least for me) than information about the process were specific and concrete deliverables Ican work towards and “gates” at which I get feedback on how my creations are so far.

One of the most crucial self-assessment seems to be having a concise research question that reflects the current research. Only with a specific question can the available literature and the necessary research work be narrowed down sufficiently to actually get done. Booth, Colomb and Williams offering a short fill-in-the-blanks pattern for this purpose, that basically asks four questions:

• What are you studying? (Topic)
• What do you want to find out? (Question)
• What for? (“Cost of Research”)
• What benefit does it have for the reader? (Practical application)

A beginner’s mistake (count me in) is to stop narrowing down with just a topic. With an actual question, it becomes clear, what your interest in the topic is. This makes it more concise and relevant to you. Unfortunately, you are not researching for yourself, but for a community of researchers. The third part describes what value this research has to them, or conversely, what cost it comes with if it remains unanswered. Finally, especially if you are in applied research, the questions is: what are the implications outside the research community, in every day life.

Once one is a little deeper into any particular question, the “topic analysis” sheet from Davis and Parker might come in handy. They suggest to fill out a questionnaire with six questions, before starting to write the actual proposal. This questionnaire requires you to name a questions, but it also includes a short summary on related work, theories, research approach and possible outcomes. The idea is that you can fill out more than one questionnaire with reasonable effort. If you have more than one topic or several variants on one you can get into them deep enough to know what they are about and so decide which alternative is best, all before you write the actual proposal.

Lauer offers some advice on what should be in the proposal, which goes pretty much along the lines of Davis and Parker. His notion of checking the quality of a proposal leads to idea that this should be a circular or iterative process. He discusses what could go wrong at early versions of a proposal and what is necessary to fix these problems.

Summing it up

Hopefully, this little overview of my small library of research aproach sources has given you some hints as to whether or not this is usefull for you. I am sure there is a ton of different literature, but it appears to me that I am approaching “theoretical saturation“, the point at which reading new books on the same topic tell me almost nothing new. Yet there might be stuff that is better written, more compact or more specialized to one or the other discipline. I would love for you to share in the comments.

The Content

The message is hardly something new, but what really makes his book so valuable, in my eyes, is the toolkit he presents to get everyone started. Even those that claim not to be “visual thinkers” or to no be “artistic” enough to draw. Over the course of the book he introduces this framework with good explanation and many illustrated examples.

Problem Categorization

The first step in approaching any problem is to think about which type of problem it is. Dan uses the a simple, common-sense categorization of mostly “W”s and some “H”s: What / Who, How Much, Where, When, How and Why. For each of these problem category there is a appropriate type of visualization: Quantities (the “how much”s) are represented by charts, “where”s are depicted by maps (geographically or conceptually), “How”s are visualized through flow-charts and so on.

The SQUID

The problem categories give a idea about what data to acquire and process and a basic idea of how to visualize them. What is still missing, Dan argues, is a way to image the right aspects of the problem. For this purpose he develops the “SQVI∆” (read “squid”) framework. It takes you through the following process: For each problem at hand, ask yourself: Do I want to focus on:

• S – Simple vs. Elaborate: Do I need to understand the basics or is the full picture with all details important
• Q – Quality vs. Quantity: Is it sufficient to separate different “types” of things or do I need numbers and metrics?
• V – Vision vs. Execution: Must I convince people of the vision, the goal we want to be at, or do they need to understand how to get there?
• I – Individual vs. Comparison: Will the portrayal of our case be helpful or do I need to compare it with other cases?
• ∆ – As is or Change (“Delta”): Should I talk about the status quo or the change between now and then?

Based on your circumstances and the group of people you are interacting with, the answers are different, even for the same problem.

The Matrix

As you can see, the questions about what type your problem is and how to think about it are two more or less independent dimensions: you will have to think about both every time, the one does not follow out of the other. In the book, Dan combines these two and gets a matrix of 30 fields (6 types of problems times 5 Squid questions) , for each of which he suggest visualizations. So if you realize that you have to talk about a “when” problem you will end up with some sort of time line – this determines the “row” of the matrix. Then you look at the SQUID-questions and find out that you want to think about execution comparing the status quo of your company vs. anther company. The matrix suggest you, how to go about visualizing this: You can use two Gantt-charts next to each other, one for your company and one for the other, or maybe you can overlay the two charts, showing the differences.

You can find reference sheets of the matrix, the SQUID and the Problem Categories on the tools page of the book’s website.

Applying the Appproach

Now, this example might appear to you like a complicated solution (running your finger through some table) to find a obvious solution (Gantt-Chart). Any maybe to a point this is true. I compare the use of “The Back of the Napkin” to the blessing of the “Getting Things Done” productivity approach by David Allen. David admits that all he presents is “advanced common sense” – collecting all your stuff, deciding what to do about it, separating between projects and contexts, reviewing your actions regularly, and so on. Nevertheless, his compact, structured description of the process, together with a well written introduction to getting the habit established makes it so much more likely to actually follow it.

The same is true for “The Back of the Napkin”: You will know most of the individual types visualizations, you know that problems can be separated into problems like “where”, “when”, etc. And yet:
The book gives you a structured and comprehensive framework to overlook the different alternatives and guides you through them. It makes a compelling case that anybody can draw the kind of pictures that are necessary for this kind of visual thinking. So even if you knew it all before, it is helpful to convince your team or your client that you should give visualization a try and that everybody can get involved in the process.

On a more practical, personal note, I find it unavoidable to revert to drawings anyway:

• If you are like me, you can work much better, even when creating completely picture free end results (e.g. long texts), when mixing textual and visual information during preparation.
• When you work together with others on a project you will end up drawing pictures anyway, because communicating ideas by saying “just read the memo” will not replcace the times standing in front of a whiteboard with other creating and explaining ideas
• When preparing a PowerPoint / Keynote presentation you will want to focus largely on visuals and will want to think about what to show and sketch out key slides before using the computer.

So, my recommendation is, that if you are a visual thinker, you work in teams or you present your ideas to other people, have a look at “The Back of the Napkin”.

Further Infos

For my German readers: Amazon Germany has the original version available. Apparently, a translated verison is also planned for May 2009.

There is a website accompanoying the book, aptly named www.thebackofthenapkin.com

Dan has a blog, in which he talks about his ideas and about how other make use of visual thinking. Look out for those entries in which he tackles a problem visually. HIs style of step-by-step visualization and textual thought development are really awesome examples of you to develop and present ideas.

There a two videos that I found.
One is Dan himself presenting parts of his ideas at a authors@google talk:

The other one is a short video presenting a summary of the book: