Mountains of notes...
March 1, 2012 9:29 AM   Subscribe

Calling all academics and graduate students: Tell me about your workflow for turning a piles of notes and ideas into something that resembles a thesis or research arcticle!

I'm in the process of writing my bachelor's thesis and am a bit stuck on how to move from reading sources to creating an outline and first draft. In particular, I'm having trouble visualizing the transition from reading articles for my introduction (where I have oddles of notes in Mendeley) into a coherent outline, into something readable as a first draft. My supervisor has requested me to start writing with the introduction first, the draft is due in a couple weeks.

My process for writing science reports in the past: (excluding data analyis and collection)

1) Dump article into Mendeley
2) Using Mendeley to read, highlight and take notes
3) Manually copy and paste these notes into a large word document. I then sort my notes into groups based on similarity, flow, etc. Anything useless or repetitive is deleted or rearranged or rewritten
4) After a couple rounds of this, I have something that resembles a rough outline to start writing from.

This has worked ok in the past, but is manually intensive to retag and copy my notes into a word doc. I have difficult writing from scratch - using my notes as the basis of what I write and transforming them is much easier when I sit down to write.

Right now, my biggest challenge is that my thesis is much larger than anything I've written previously, and I would like to use this oppurtunity to learn more about how to write research documents and scientific papers more effectively.

What do you do to go from reading/an idea - outline - first draft of a section? I'm looking for some insights and examples of workflows for creating scientific papers from a jumble of notes, along with some general advice about learning how to write papers. Just ordered this book on writing in a research context. that was recommended as a good resource. Thanks for the help!
posted by snowysoul to Education (14 answers total) 54 users marked this as a favorite
My approach for paper writing was like this: first, I'd do enough basic research and read enough articles to give me a thorough-enough idea about the topic. This becomes easier as you advance in the field because a lot of the basic information becomes second nature to you. Then, I'd think about what I wanted to say. What's the purpose of the paper? Is it just an overview? What's my stance? My focus? Do I want to make a point? If so, what? Just basic decisions with one-word to one-sentence answers.

Only then would I start making an outline, which takes those decisions and transforms them into a skeleton: "Okay, I need an intro. Then I need to review current evidence and schools of thought. That'll be paragraphs x, y, and z. Then I need to start building my case, so sections 2 and 3. Then my conclusions, and a future directions." Something like that. This will change as you write, so don't sweat it. This is the time to step away from your notes for a few minutes. Just use what you've digested from them. You can always edit it later.

Then, I'd start writing. This is the hardest part, bar none, but just start putting words on paper. They could be the worst sentences EVER, but they're going to get revised. You may rethink your entire approach. DOESN'T MATTER. Just write. Words on paper. They'll get fixed later. Sentences will be reappropriated elsewhere. Paragraphs will move. Sections will reorder. You'll definitely find yourself writing things that you don't have citations for. That's why you're writing now and not the night before it's due: you have time to get more citations and research. Put "[citation needed]" there and move on, then come back with the citation when you're revising. Just be bold and unafraid to admit, when necessary, the course you're on needs to be changed. It's not your fault. You just didn't know any better, but now you do, so onward and upward with science!

Once you do that, then revise and write and revise and write, you'll have a paper. Easy. But you just need to do it: just start asking yourself questions. Start making that outline. Start writing sentences. Bad writing is better than no writing, because then it's just a matter of reworking. It'll all come together, I promise.
posted by The Michael The at 9:47 AM on March 1, 2012 [5 favorites]

My workflow is similar to yours, except that my "notes" tend to be more than just notes. As I find things I like, I write out paragraphs -- even pages -- based on these things. This helps me organize my thoughts in a way that just working from a pile of notes can't.

When the time comes to put my "notes" in order, I'm actually organizing entire paragraphs and pages. My research papers tend to be assembled, rather than written in a narrative style. And my outline grows out of that process.

Then I sleep on it or abandon the whole thing for a few days, then I reread everything and massage the writing into a decent narrative. THEN I write the intro (unless I've already written it in the course of note-taking).

I can't imagine beginning with the introduction. You might just as well begin with a conclusion!
posted by coolguymichael at 9:50 AM on March 1, 2012 [4 favorites]

I'm looking for some insights and examples of workflows for creating scientific papers from a jumble of notes, along with some general advice about learning how to write papers.

There are many different possible workflows and this is a big question, so my answer is neither exhaustive nor the one and only way. But for me (and in my field), notes on previous literature are just one component to the writing, and not usually a starting point per se. I need a concrete and specific hypothesis to test, an important question that hypothesis is addressing, an understanding of what previous literature has said about that hypothesis/question, and my own proposal (which is usually a worked out mathematical model along with new data supporting that model). Honestly, I usually start writing up the last part first, especially the data, not the lit review/intro.

The question of how to review the previous literature, in my own writing, is guided entirely by the hypothesis and my own proposal. Usually it is possible to identify a few core previous answers (or extrapolated) previous answers relevant to the hypothesis, and group the previous literature around these answers. Then I fit my own proposal into one of these, or explain how it is different than any of the existing camps. It sounds like your edict requires you to start with a lit review (though IMO this is often a recipe for later having to rewrite it entirely). So if I had to do this, even without a particular proposal in place, I'd suggest finding a concrete hypothesis to group the review around, in this way. Most pure lit review articles that I've read, if they are any good, also identify a set of core hypotheses/questions even if they are not taking a stand on them.

I agree strongly with the above comment that the best thing is just to start getting words on paper, even if you do have to rewrite it all. In terms of getting students to start writing something, anything, I think we all have our individual strategies, and this may be your particular supervisor's. (Another suggestion: find an article by your supervisor to get a sense of how they deal with this. Or ask them if they can recommend a model. Some of the details and conventions can be quite field-specific.)
posted by advil at 10:04 AM on March 1, 2012 [1 favorite]

It would help to know a bit of what your thesis is about (I never had to write a bachelor's thesis). Is it a proposal, a review of the literature, or what?

Generally I have an outline in mind for a paper before I start researching the topic extensively. If you already know about the topic that you're writing about, then you probably know about: 1.) gaps in the research in that area, 2.) common conflicts over how constructs are defined or measured, and 3.) different theories that are used to explain whatever it is that you're writing about. Each thing I numbered is a potential paper and the process for organizing a paper will vary based on which you're writing about.

Of course, you need an intro. In that intro you need to lay out exactly which of the above questions (or some other question) your trying to address. From there, the process of organization will vary. If you're proposing new research for instance following 1.) to fill a perceived gap in research in your area, then you normally do a section on the lit review, a section on methods (research question, data, constructs, measures, etc) and a section on the contribution that you want to make during the project. Your notes and research should focus on filling out the information that you need for that section. I often start with the methodology section first because I find that once I've fleshed that out and know what I want to measure and how I want to measure it, then it's much easier to go back and find research about those specific things. Otherwise I spend a lot of time looking at irrelevant literature and not enough time actually writing.

If you're doing a 2.) or 3.) I still recommend starting with the 'methodology', but in this case your methodology is going to be your perspective for explaining or reconciling the debate in the field that you've chosen to write about (part c, below). So for instance in my field there's a debate between whether police performance should be measured according to decisions made by police officers about when to use force, or whether performance should be measured according to crime prevention goals. If I'm writing a paper about this topic, then I'd outline the paper like this:

(Intro) A. There's this issue about performance measurement in policing (and give the reader enough information to know what I'm talking about when I say 'performance measurement')
B. There are well-known groups of scholars who take one of two perspectives on this issue (explain what the two groups of scholars, group A and group B, have said about the issue; explain how and why they differ)
C. Group A or B is right, or I think there's an option C
D. Here are the advantages of this approach to understanding the issue
E. This perspective will help the field to move forward in some important way.
(then briefly recap the paper in your conclusion)

Again, Part C is the most important part of the paper, much like the methodology section in the previous example, and it's the part that you need to have thought about before you can start the other section of the paper. When you're reading the literature, you need to be trying to first to identify the question that you want to answer with the paper. Once you know that, you will find it a lot easier to distinguish relevant research from information you don't need.
posted by _cave at 10:14 AM on March 1, 2012 [2 favorites]

Lots of good ideas here already, so I won't belabor the point. Here are 2 of my preferred methods:

Whenever I'm stuck in going from notes to paper, I find a large piece of paper or cardboard. Looking at my computer notes, I pull out some tentative arguments or themes or connections, and I draw a mindmap of sorts (cluster your sources according to arguments/themes/connections). Having it all out in front of me makes it easier to write from the connections I've noticed. Narrative flow emerges later, when you really know what you're trying to say.

Another useful tactic is to imagine the conversation between scholars with varying perspectives on the issue you're studying. (This may be particularly well-suited to my undergrad, which was in political science.) Are you seeing natural allies here, folks who take similar perspectives or use similar assumptions? Anyone really stand out from the pack?

Also, be sure you know how much (if any) original research you are expected to do. At the wonderful liberal arts college I attended, there was nonetheless a good deal of confusion about this point among many thesis-ing seniors. If you're supposed to do original research, ask your professors what they have in mind for this - lab work, interviews, reviewing primary source documents, etc.

Good luck!
posted by brackish.line at 10:30 AM on March 1, 2012

A professor of mine once told me to write a first draft, using all the notes and research you plan to. Then, when finished, look at your concluding paragraph. Copy it, erase the rest of what you have written, and paste it as your introduction. Then start over.

The first-draft conclusion is almost always the best distillation of your thoughts and ideas, and generally makes for the best thesis.

I used this approach throughout grad school and it really works. Though of course I still cut-and-paste the research and quotations, etc., from the first draft. It's the overall idea that works well.
posted by edmondcb at 10:46 AM on March 1, 2012 [3 favorites]

In my field, Ethnomusicology, it's possible to get a lot of conflicting perspectives on a single issue. So, sorting through all of that to come away with something useful can be a challenge, as you get mired in the he-said-she-said of it all.

I adapted an approach that I heard Kurt Vonnegut used to use. I put the varying viewpoints on post-its, and try to group them by sticking them on a wall (pace, my old landlord, who was OK with me doing this!). Sometimes, it helps to physically step back from the argument and look at the broader patterns. Either a way of organizing it sticks out at you, or you realize you're focusing on the wrong aspect or thinking about the problem the wrong way.

For my own writing, I tend to use a mixture of what coolguymichael uses, and doing a straight outline.

Good luck!
posted by LN at 10:49 AM on March 1, 2012

If you are in the experimental sciences, Whitesides' advice on writing a paper can apply to thesis chapters too.....
posted by lalochezia at 10:50 AM on March 1, 2012 [1 favorite]

I think the best way to approach this is to try to tell the story of your research to someone who doesn't know anything about your field. In absence of a real person, imagine being asked to give a presentation to a bunch of smart 14 year olds about your research.

Start with one sentence that sums up everything about your research:

"For a long time, we thought that kids needed praise to bolster their self esteem. Now, we're not so sure. It seems that giving too many compliments is actually bad for them."

Then, you can elaborate. Try to split this sentence into three sentences, the first being the intro, the second developing an idea or hypothesis. The third a conclusion.

1, Since the 1960's, self esteem was thought to be the most important thing we could give our kids.

2, In the 1990's, the first doubts began to appear. Was this really the case? Especially doctor Frankenstein thought we should test this, which he proceeded to do the next ten years.

3, Turns out, self esteem is a double edged sword that can make kids narcisstic.

Now, turn those three sentences into another three. And so forth. Importantly: tell a story. Focus on the stuff that is counterintuitive (because that's the most fun to read) and important.
posted by NekulturnY at 10:56 AM on March 1, 2012 [1 favorite]

I came to mention Whiteside's outline method as well. I find it tedious and irritating, personally, but many of the people in his group swear by it.
posted by bessel functions seem unnecessarily complicated at 11:05 AM on March 1, 2012

My process is pretty much the same as The Michael The's. It sounds like you approach a paper without knowing what you want or need to say, so you start with too many notes or ideas and spend quite a bit of time whittling them down. You might try starting with a barebones outline of the main ideas you want to touch on, find articles that meet your needs and pull from them, and then launch into a "vomit draft" where you get everything out that you want to say. Then go back and edit.
posted by puritycontrol at 1:36 PM on March 1, 2012

If you haven't already, read this.
posted by 5Q7 at 3:05 PM on March 1, 2012

First, I read a lot without taking any notes at all. I just concentrate on getting the big ideas in my head. Taking notes can be distracting for me.

When I'm tired of that, I imagine that I'm explaining my research to someone I just met on the airplane. They don't want anything exhaustive, just the main points. Those main points become the major sections of my outline. I write down the sentences that I would actually say. Sometimes my lips move.

Then I take each section and pretend I'm explaining just that one narrow little topic to someone unfamiliar with it. Again, I write down the sentences I would actually say.

That's usually good enough to get through the difficult part of just getting something down. Then I expand and edit. At that point, I go back to the literature for citations. Sometimes I can't find a citation, and this can lead to an idea for a new research project!

I try not to wed myself to this outline, because I gain a lot of understanding through the process of writing, and this new understanding often leads me to a different flow of presenting the ideas.

I also try not to fall in love with particular sentences. That can lead to a frankenstein document that is crafted around favorite bits, rather than something aimed at communicating broader ideas.

I always write the introduction first, so all that information can inform the way I analyze my data. But then, based on the results, I often go back and re-write the introduction to focus on aspects that fit in better with my results and the story I finally decide to tell.
posted by pizzazz at 9:43 AM on March 2, 2012

Though not specific to scientific papers, a method that I've used successfully beginning by putting notes on index cards. The content of the cards can be an overarching idea, an actual note and/or a citation to some kind of source material (that I'll have on hand later on and is ideally cataloged in some way for easy reference).

When I'm reviewing, I use the floor, a table or a cork board to arrange and visualize the material. With cards, it's easy to see the piece in it's entirety and to move things around if necessary. Once I move onto the writing phase, I transfer my cards (in order) to a card file. It's an easy, organized way to review your notes because they now follow the logic of the paper.
posted by EarnestSchemingway at 1:38 PM on March 2, 2012

« Older Can you help me create funny and encouraging signs...   |   How much does living in Antwerp cost? Newer »
This thread is closed to new comments.