Writers' Lab: How to write your Conclusions, Part I: Journal or Conference Paper

If you are the type of linear writer like me, you will typically start your paper from the abstract, then the "Introduction", and on and on - all the way down to the "Summary and Conclusions."

When you start proofreading your entire opus, you might find that you didn't quite keep the same red thread throughout your work that started at the "Introduction" section. You might have trailed off, adding something along the way that seemed relevant when you were writing that section. In the very worst case, you'll end up with a "Summary and Conclusions" section that is completely running in the wild.

Now, what do we not want in a "Summary and Conclusions" section? If you've been reading a number of papers for your literature review, I'm sure you might have come across a few papers where the final sections left you puzzled. Here are the typical ingredients of a poor "Summary and Conclusions" section:

• Contains new ideas: The final sections happily comes to serve you up with an unexpected dessert. No background, no reference to the experiments, no analysis - just an out-of-the-blue statement that leaves you startled.
• Contains nothing beyond the current state-of-the-art: If you are pressed for time and want to quickly get a grasp of a paper, you will typically browse through the abstract, then glance over the figures, and then read the "Summary and Conclusions." If this section is filled up with general statements of things we all know already, then you will start to wonder what original work this paper contains - and toss it to the side.
• Only sums up one part of the paper: If you read a "Summary", you expect a quick recap of all parts of the paper. Some authors (presumably pressed for word count or paper length) come in wham-bam with a list of their conclusions - without the soft bed of a few sentences that repeat the problem, literature review results and methods.
• Is not a self-sufficient unit: Take a summary out of a paper, without knowing the contents of a paper, and the paragraph(s) still need to be a logical unit that requires no further reading of the paper to understand the "Summary and Conclusions."

So now that we got these typical mistakes highlighted, I would like to share with you how I write my "Summary and Conclusions" section. Admittedly, I learned it the hard way, by getting a paper returned and the "Summary and Conclusions" part torn to pieces by the gentle reviewers. To repair the damage done, I used a strategy that I have been implementing ever since.

The approach that I will discuss is mostly suitable for linear writers. If you prefer to nibble and scribble at different subheadings at more random points in time, you might as well find this approach useful - it will help you get a grasp of the entire paper again and focus on the main points to wrap up your writing.

As I said, I typically start by writing my outline, then filling in what I already have from reports or earlier work, and then start to rework the sentences to actually write the paragraphs. The only thing that I never do is the following:

I never write my "Summary and Conclusions" section before a first round of proofreading.

I simply leave it blank - a completely blank section. Instead of finishing up my very first draft -as good as that might feel- I leave it open.
Then, I use the following sequence:

• I sit down to proofread my very first version.
• While proofreading, I take notes of the main points of every subchapter.
• Once I reach the end of the paper, I reread these notes.
• I use these notes to write the "Summary and Conclusions" sections.

Let's quickly look back at the list-of-shame for a "Summary and Conclusions" section, and link that to this strategy:

• No new ideas: it is virtually impossible to take notes of the main points and use these as a guidance and still manage to slip in a new idea.
• No general truths: Noting down your contributions helps you to keep your focus on your own work, and helps you stay clear of the random chatter.
• You cover all the parts: If you take notes of the main ideas of every subchapter, you'll be able to cover all the contents of the paper.
• Self-sufficient unit: Make sure you proofread your "Summary and Conclusions" to see if you wrote it as a stand-alone paragraph. Don't reference to elements of the paper, just keep it sharp and shine a light on the major elements of the paper.

How do you write your conclusions section? Do you agree with my approach?

Silver Linings: Hormones and your health

For those of you who follow me on Twitter, you might have seen that I occasionally share some of the work of Alisa Vitti, founder of FLOliving. Even though mostly aimed at women with PCOS and other hormonal issues, I think her work is valid and interesting for all women.

When she was a student at med school, Alisa was diagnosed with PCOS. Instead of going for a life of medication and more medication, she opted for going out in the wild and finding a way to a better health, by delving deeper into the science of food, genes and hormones.

The result of her studies is what she calls the "WomanCode" - a protocol that suggests which foods, skills, types of exercise and other elements of self-care you should emphasize depending on where in your cycle you are. The result is an organic form of time management, that feels more natural than other linear approaches to time management. Stop wondering why sometimes you feel outgoing and other times you want to stay on the couch with a book - you can fit both elements into a monthly planning.

Dealing with PCOS myself, I've been following Alisa's course over the past months, and I've been trying really hard to implement her recommendations into my schedule. Traveling to conferences and moving from one place to the other has been a challenge, but overall I noticed improvements in my energy levels. If you are interested in her work, but don't want to make the time and financial commitment to the course, you can check out her book WomanCode,

If you are interesting in her work, I recommend you to start by checking out her recent GoogleTalk here:

On the road to the defense, part II

Right after the defense, with my husband, paranymphs, full committee and the Beadle

I'm writing the second half of my experiences in the final year of the PhD towards the defense here in hindsight, as I defended successfully on June 14th 2013 and received my PhD degree the very same day.

Previously, I described what happened between finishing my draft dissertation in November 2012 and sending copies out to the committee members in February 2013. In this post, I will continue where I left off in February 2013, and describe those messy final months of working on a doctoral dissertation.

9. Get the comments from the committee members
Most of my committee members preferred in-person meetings. Luckily, most of my committee members are based in Delft or often in Delft for meetings, so that I only had one long train ride to Germany for a meeting. During the meetings, the committee members gave me their input in many different ways. Some preferred my talking them through the dissertation, some interrogated me on topics more related to their field that I implemented into my work, some focused more on the layout and writing style - every reviewer seemed to have his own style of assessing my work.

10. Implement the comments from the committee members
With my piles of notes from these meetings, as well as the pages of comments from the final committee member in London with whom I didn't meet in person, I set off to make some changes to my dissertation. It took me about 6 weeks of full-time work to implement all the comments and respond to all the questions in a way that I felt was satisfactorily. With some committee members I met several times to improve large chunks of my dissertation, or I e-mailed back and forth with them. At first I was surprised as I thought "the work was done", but then I understood how valuable a fresh look at my work can be, not only for my dissertation, but also for my further publications (ie. the papers that I need to write from my dissertation)

11. Iterate if necessary
I felt at times intimidated by the knowledge of some of my committee members, so I felt afraid to ask for their help and input at first, and spent some time studying topics where I felt I was still lagging behind. But then it clicked, and I understood that if they take the time and effort to give some thorough remarks on my work, that they actually show me that they want to help me improving my work. So I went back and forth, learned more things, saw a few more connections in my work and genuinely made a better (yet almost 100 pages bigger) dissertation.

12. Get the approval from the committee members
I remember how my heart jumped up when the final committee member strolled into my office with his "signature". The committee members all need to sign two documents for the Beadle: the first one confirming their availability on the day of the defense and the second one their approval. One of my committee members approved under the condition of making corrections to the dissertation, which then required a written letter from my promotor and additional approval to keep the whole (bureaucratic) process rolling.

13. Revisit with promotor and copromotor
After editing my dissertation, I revisited the work again with my promotor and copromotor, and we needed to tighten my set of propositions (ie. a leaflet with 10 propositions additional to the dissertation that shows your scholarly insight). With a conference in Israel in between, I felt that time was ticking - and that those 4,5 months of preparation that seemed like an eternity where flying by.

14. Send to the printer
With all approval from the full committee, I sent everything to the printer. In the final weeks of working on the edited version of my dissertation, I spent some time in layout-hell, trying to make sure everything fit within the margins (I needed to revamp an entire Annex for that purpose), and making sure all tables are on portrait pages (I used a lot of landscape tables to hold all my data). I sent a draft to the printer before the critical date, and was reminded as well of the fact that all chapters should start at a right page. Lots of tinkering before sending it all off to the printer, as you can imagine. In the meantime, I had a designer at the printer's office developing my cover and invitations - which saved me time for sure, but needed my attention and input as well.

15. Check the print proof
I got my print proof even earlier than the schedule of the printer indicated, and I was very happy to hold the very first physical, book-like copy of my dissertation. I didn't really know how to check the proof: reading from A to Z felt pointless to me, so I ended up going over every page and checking if all looked right - and I did find quite a number of things that did not look as I wanted (yet looked exactly the same as the PDF I delivered to the printer).

16. The final order and delivery
I made more edits, then struggled with the PDF conversion for a few hours. The print-proof arrived right before the start of a conference in Rotterdam, so for the very first time (and equally last time) in my PhD studies, I had to cut back on some sleep to get through the final edits and make it in time to deliver everything to the printer. On the 17th of May, I received a call from the goods delivery point of our faculty that there were "some boxes" for me - and so started my journey to drag my 250 copies of my dissertation into my office...

17. Spread the word
I took the stack of business cards that I had amassed from previous conferences, and started mailing out copies to interested contacts (or at least, contacts that I thought would be interested). I also had two more conferences, 1 in Japan and 1 in Pittsburgh. I took an entire carry-on bag filled with copies of my dissertation, and distributed these during the conferences. By the time of my defense I had about half of all copies distributed.

18. Prepare for the defense
I didn't really know how to prepare for my defense, and I'll write a post about it specifically in the future, but essentially I didn't have much time for preparing my defense all in all. Certainly, I had to prepare my "lekenpraatje", and I reread my dissertation once more, but preparing for possible questions was a little more vague. I summarized my dissertation in a sentence per page, as suggested by Phillips and Pugh, I went over my meeting notes again, and I tried fishing for information from my committee members (unsuccessfully).

19. Defend
Less than a week after landing back onto European soil after my trip around the world to Japan and the US, I got over my jetlag (mostly) and defended my dissertation. I was nervous at first, but after the first question, I felt my confidence returning to me, and I sort of enjoyed showing my work to the committee members.

20. Celebrate!
After the actual defense and receiving my degree, I organized a reception in the TU Delft Aula building, and in the evening a dinner in a restaurant in Delft (plus some drinks afterwards in a bar). Even though I was exhausted from the stressful day, I was feeling uplifted and enjoyed celebrating my fresh degree very much (well, who wouldn't?).

How different are the preparations at TU Delft and my take on them from your institution? Please let me know in the comments!

Writers' Lab: A drastic revision strategy for improving your paper's story and organization

Today, I once more have the pleasure of hosting Nicole K.S. Barker in the writers' lab. Nicole is a Ph.D. student at Laval University in Québec, Canada, working with Ducks Unlimited Canada and the Boreal Avian Modelling Project. While technically a Forestry student, she classifies her research as "applied ecological modelling". 

Within her Ph.D. she is investigating various methods for quantifying patterns in waterfowl populations over space in time. 
Alongside her research, she continually searches for ways to improve her writing, productivity, and R programming skills, and has found social media to be instrumental in this regard. You can read more about Nicole on her blog or Twitter. 

It seems like revising a piece of writing should be easy. You’ve completed the hard part in actually writing a draft, so it should be simple to fix the grammar and simply be done with it. However, depending on how long it took to write your paper and how much your thinking changed over that duration, you may find some hidden challenges.

If your vision, objectives, or "pitch" differed during the outlining or drafting stages from where it ended up, you’ll have a somewhat disjointed piece of work or section of work. I find that this happens most often for introduction sections. For example, it might not set up your paper’s objectives correctly, or perhaps it provides all of the details but not in an easy-to-follow story.

I recently adopted a new strategy, and now I recommend it to everyone I know. It’s based on chapter 3 of a book by Michael Jay Katz called From Research to Manuscript: A Guide to Scientific Writing. Any time I read a disjointed section of a paper, I remind people about this strategy. I also use it myself for all sections of a manuscript, from methods, to results, to discussion and introduction (keep in mind I’m in the sciences, but this strategy should be relevant for all types of papers, essays, and other writing).

The general procedure is somewhat drastic. It involves start from scratch – yes, a blank page! – once you have already written a draft. It can be intimidating and even discouraging to start from a blank page after spending the time to write a draft, but I guarantee it will lead to a better organized paper.

Steps: Within this new blank document, identify the main idea of each of your paragraphs and write it in sentence form. Ensure that these main sentences tell a story before you add any other detail. Fill in sentences one at a time from your rough draft to the correct paragraph in your new document in a coherent order. Edit each sentence for grammar and wordiness, but keep it in bullet point form. Once you have copied over all sentences and edited for coherence, you can change the sentences to paragraph form instead of point form. Ensure that each sentence flows into the next. Your “main idea” sentences become the topic sentences of each of your paragraphs. Katz’s book leads the reader through a detailed example of what this process looks like, and it’s quite illuminating.

Advantages: Starting from a blank page frees your thinking from the confines of your rough draft. Starting with main message of each paragraph ensures that your work has a flow and tells a story from start to finish. Reworking each sentence individually focuses your energy instead of breezing over individual sentences’ awkwardness or wordiness as can happen when reading a paper as a whole.

Disadvantages: It can be time consuming because you’re starting over. However, keep in mind that much of the work in your rough draft will be kept in this draft – it’s just moved around or reworked.

Let Eva or me know if you’ve tried this method or if you use another strategy for drastic revisions.

Presentation of my PhD Defense, "Lekenpraatje"

On June 14th 2013, I defended my doctoral dissertation at Delft University of Technology.
As I wrote before, the defense itself was the very end of the entire process of writing, editing, re-editing, publishing and finally defending my doctoral dissertation.

One of the elements of the defense at TU Delft itself is the so-called "lekenpraatje" - a presentation for the laymen.

For my "Lekenpraatje" I decided to touch upon a few elements of my dissertation:

• the background: why was this research necessary, and what has been (or has not been) done previously
• an overview of the experiments: because all my friends and family know is that I spent crawling around on a dusty laboratory floor for a little less than 3 years, resulting in my ever-white-dusted appearance at random events, or my needing to wash my labjeans at high temperatures to get the cement dust out.
• beams versus slabs: or something that I have been trying to repeat over and over again: slabs in one-way shear behave differently than beams!
• Modified Bond Model: the theory that I worked on, and I tried to explain it as graphically as I could.
• Application to practice: What have we won by investing in experiments? We found that we can decrease the loads for calculation with almost 20% as a result of the recommendations that we formulated from our experiments. That does NOT mean we've "won" 20%, as the shear provisions in the current code result in lower shear capacities. Overall, we did see a modest improvement based on our recommendations and the Eurocodes as compared to the previously used national code.

All in all, I had about 20 minutes' worth of time allotted for the presentation - and it was a little on the short side. 30 minutes would have been more comfortable to stop and expand on a few topics along the way, but I think (and hope) it gave my friends and family an idea of what I worked on for the past few years.

If you are interested in the slides of this presentation, you can find them here:

Lekenpraatje from Eva Lantsoght

Quick Scan Method - Presentation at the International Bridge Conference

I recently presented in a technical session of the International Bridge Conference in Pittsburgh - a great venue for practitioners and researchers that work on practical applications for the bridge engineering field.

The proceedings will be published over the summer, and the abstract of my paper is the following:

To assess a large number of reinforced concrete slab bridges for shear, recommendations based on experimental research are implemented in a spreadsheet-based “Quick Scan” method. In particular, transverse load redistribution is studied. The resulting method is used to study cases of existing slab bridges in The Netherlands, indicating the benefits of the proposed method as compared to previous “Quick Scan” approaches.

You can see my slides here:

Quick Scan Method from Eva Lantsoght

Writer's Lab: Constructing effective outlines using assertive language

Today, I have the pleasure of hosting Nicole K.S. Barker in the writers' lab. Nicole is a Ph.D. student at Laval University in Québec, Canada, working with Ducks Unlimited Canada and the Boreal Avian Modelling Project. While technically a Forestry student, she classifies her research as "applied ecological modelling". 

Within her Ph.D. she is investigating various methods for quantifying patterns in waterfowl populations over space in time. 
Alongside her research, she continually searches for ways to improve her writing, productivity, and R programming skills, and has found social media to be instrumental in this regard. You can read more about Nicole on her blog or Twitter. 

Most writers start with an outline. And if they don’t, they should.

Outlines help you think through your topic. They help you identify the thread of your story - the message you’ll carry through from start to finish. Nothing is more frustrating than writing a full introduction to your paper only to discover that it doesn’t mesh well with your results or discussion once you get to those sections. As you fill in the finer details of your outline, you can revise the order of sentences (or even paragraphs) more fluidly than if you were writing in paragraph format.

I’ve been using outlines to guide my writing since the beginning of my academic studies, and I've used basically the same format the whole time. However, I recently revised my outline format following a tip from one of my writing group members. I’ve already found it useful, so I wanted to share it.

Étienne suggests that instead of using an outline to simply describe the content of each paragraph, we should use assertive statements. That is, summarize the main message or conclusion of each paragraph in one sentence or phrase. In this way, the story of the paper is apparent within the outline before we’ve even started writing the full content.

These outlines are harder to write, because they require you to really think through what you are trying to say. (You might find that this style works best for a second outline once you’ve got the very rough ideas down.) However, the assertive outline makes it easier to identify problems with flow between paragraphs, or areas where your story becomes inconsistent. Because the ideas are still in outline format, it’s easy to rearrange or adjust as necessary.

Give it a try and let Eva or me know how it went for you.

As an example, the introduction and discussion from a hypothetical paper about visualizing music with paint as a means to understand hearing impairment.

Old

Introduction:

• Broad opening statement – hearing impairment
• Previous research 1 – cochlear implants
• Previous research 2 – visualizing sound
• Our study and objectives – effect of music on paint

Discussion

• Summary of results
• Relation to previous research – other attempts to visualize sound
• Relation to previous research – other attempts to understand cochlear implants
• Conclusions and implications for hearing impairment

New

Introduction:

• XXX number of people experiencing hearing impairment in one form or another
• Cochlear implants have substantially improved hearing in XXX people since 19XX
• In an effort to understand effects of cochlear implants, researchers have attempted to visualize sound using a variety of methods.
• We applied a novel technique to visually represent sound using paint and multiple styles of music.

Discussion

• We found that different musical styles showed distinct patterns in the corresponding paint splatters.
• Our research corroborates previous results that visual representations of sound display meaningful differences.
• The differences between paint splatters from Mozart and those from Metallica help explain why those with cochlear implants prefer music with a strong beat.
• Our results provide a novel means to make sound relatable for those with hearing impairments.