Thought of the day

December 4, 2013

Advertisements

38 Responses to “Thought of the day”

  1. Odyssey Says:

    Yep. Lacks synthesis of data into story and requires less intellectual input.

    Like

  2. Jonathan Says:

    IIRC, Finnish PhDs are a compilation of five peer reviewed papers.

    Like

  3. bsci Says:

    A thesis that is read by 4 people before being buried in a university library probably never to be seen again is inferior to research that is indexed in PubMed and easily downloadable.

    While I’m sympathetic to not setting the bar unnecessarily high for a PhD, a system where the norm for a PhDs ends with no change in communal knowledge is extraordinarily wasteful. I also suspect that, even for PhDs who don’t enter the academic track, having a peer reviewed publication or two under their belt doesn’t hurt career prospects.

    Like

  4. AcademicLurker Says:

    I don’t mind the stapled papers so much (most dissertations are “chapters” that are basically just manuscripts). It’s the “thin introduction” part that’s not good.

    I found that writing an extensive overview-and-synthesis-of-the-literature introduction for my diss was a very useful intellectual enterprise.

    Like

  5. odyssey Says:

    A thesis that is read by 4 people before being buried in a university library probably never to be seen again is inferior to research that is indexed in PubMed and easily downloadable.

    If that were the sole purpose of the dissertation, then it would indeed be inferior. But that’s not it’s purpose.

    Like

  6. dr24hours Says:

    Judging inferiority or superiority (and always finding ourselves on the superior side, yes?) seems like a hallmark of insecurity.

    I presume you’ll argue you’re judging the *requirement*, but it’s plain that we then judge those who did the “wrong” kind as inferior to those who did the “right” kind.

    It’s all one-upmanship. Many monolithic dissertations could really benefit from external peer-review by a journal and anonymous reviewers. Many loosely-connected papers treated as chapters could benefit from stronger binding. I don’t believe there’s any unbiased way to declare one method better than another.

    Like

  7. Dave Says:

    If that were the sole purpose of the dissertation, then it would indeed be inferior. But that’s not it’s purpose

    Plus, most theses are available for download now too. I know mine is.

    Like

  8. zb Says:

    “Judging inferiority or superiority (and always finding ourselves on the superior side, yes?) seems like a hallmark of insecurity. ”

    Love this.

    I agree that there is real value from aspects of both kinds of D’s.

    But, on the other hand, I think that when I see a PhD recipient who did not manage to publish anything during their 4+ years, I have no reason to believe that they will every accomplish any publishable work. Could this view be altered by specific information about what they were able to learn during that period (and why this will eventually result in contributions to scientific knowledge)? Probably, but it would not be an easy task.

    Now, I don’t know much about what value a PhD is supposed to provide to a non-research career, so presumably those hiring PhDs for other goals could judge based on other criterion.

    My view may not answer the question of whether a student should be *allowed* to graduate without any publications, but it is certainly relevant to whether a student should decide to graduate without any publications.

    Like

  9. Dr24hours Says:

    This is also very field-dependent. In many engineering disciplines, conference presentation remains more valuable for a GS than publication.

    Like

  10. Lee Says:

    Echoing what Dave said, electronic thesis and dissertations (ETDs) can be considered published by scientific journals. In fact I’ve had others comment that they read my dissertation prior to the work being completely published.

    Like


  11. I think the monolithic dissertation makes more sense in the humanities because they are more a book-based rather than paper-based culture and more than a few dissertations in history, comparative literature and so forth have gone on to be published books and be read by thousands. This generally doesn’t happen in the sciences. Instead nobody reads it except the advisor and the committee. If that. Everybody has stories where they or a friend claimed to include a statement in their dissertation along the lines of “anyone who reads this sentence can claim $20 from the author” and were never taken up on it.

    Like

  12. drugmonkey Says:

    zb-
    I am aware of individuals with no doctoral pubs who went on to publish appropriately over years as independent faculty. Also individuals with several first-author glamor and baby glamor pubs who never published a thing as independent faculty. Nada.

    Accounting for the way publication success paves the way for future jobs, grant money and factoring that out… I reject your supposition.

    Like

  13. odyssey Says:

    If the purpose of a dissertation is to highlight publications along with an intro chapter, why not just require students to publish a review along with primary research papers and do away with dissertations? <— facetious comment alert

    Like


  14. @Dr24hours
    Yes and no. I was a postdoc in a Computer Science department (working on phylogenetic algorithms) and so I was exposed to the CS/Math/Engineering way of doing things. Yes, “conference presentations” are important, but they are very different from those in biology. They aren’t just talks but include a peer reviewed paper component. These are collected in “Proceedings of the 12th International Conference on Obscure Research”, etc. and are consulted and cited by people who never attended the conference — they are basically considered journals, and some conferences are considered more prestigious than others.

    Like

  15. jojo Says:

    Both a badly integrated stapled-paper dissertation, and a monolithic, unpublishable one are bad. Not sure why it’s particularly important to decide which is worse.

    Like

  16. Dee Says:

    Devil’s advocate and a little off topic. Since dissertations are readily available for download/ consumption, should references there-in count towards Impact Factors? (I have wondered about this for a minute)

    Like

  17. DJMH Says:

    Inferior as doorstop, perhaps.

    Like

  18. dr24hours Says:

    @badger

    Oh yes, I know. I’m a systems engineer. My first paper in the Winter Simulation Conference is finally happening this year. And yes, it’s a peer-reviewed paper being published in the proceedings. Similar in concept to a journal article. But still a slightly different world.

    Like

  19. Grumble Says:

    The doctoral thesis which consists of three published papers stapled to a thin Intro is inferior to the monolithic dissertation if, and only if, one places the monolithic dissertation on top of the doctoral thesis which consists of three published papers stapled to a thin Intro.

    Like

  20. bsci Says:

    Perhaps I’m in the minority here, but I think an important purpose of scientific PhD training is to show someone can independently conceive a scientific experiment, plan and execute the experiment, and communicate the findings to others. This doesn’t require an accepted publication, but it does require writing up something that is interesting and clear enough to be worthy of publication.

    If one is communicating science, why should a student waste time framing results into a monolithic dissertation rather than publish studies as they are completed in more accessible formats? I’m aware some dissertations are accessible for free or reasonable prices online, but with nothing like the ease to find and access like PubMed and PubMed Central. How often does anyone here actually identify and read other people’s dissertations? I’ve done it, but not often. At least in humanities, a monolithic dissertation is essentially the draft of a book to shop around to publishers, but that’s rarely true in science.

    My judgement of “inferiority” is based on what I see as the desired outcomes. If the goal is to widely communicate one’s research, getting research through peer review is superior to not. If the goal is to get grants for a postdoc or faculty position, a track record of peer-reviewed publications is better than not. If the goal is to get an industry job, showing your research was acceptable to independent experts is better than not.

    I’m all for a dissertation including some solid introduction chapters. I found the background reading that went into these chapters to be much more useful than the words I put on the page. I also wouldn’t consider it appropriate to delay awarding a PhD merely because the introductory writing that connects some good research projects is merely mediocre.

    Like

  21. bob Says:

    I’m totally with you!
    Corollary: A tenure “dossier” consisting of papers, grants, etc. stapled to a thin introduction is inferior to a habilitation (you know, German-style 2nd dissertation).

    What, you don’t think assistant professors would also learn something by having to publish a book on their research?

    Like

  22. odyssey Says:

    So, a bunch of you honestly think a dissertation is nothing more than a repository for published/publishable data accrued during the PhD? Really? It shouldn’t consist of anything more than that? Or have any value beyond that? Really?

    Really?

    Like


  23. @odyssey

    I sure don’t. Wrote mine in 2008. Had to do full write up for all lab work I did, published or no. I had a few publications- was NOT allowed to “staple together” and call that a thesis. My PI insisted on a very full Introduction as well.

    For the projects that were published, I had to go into WAY more detail than was in the publication. I kinda thought that was the point of a dissertation….

    Like

  24. Joe Says:

    A thesis should have a real introduction (on the order of 20pp) setting up the problem and the state of the knowledge as well as the rationale. A thesis should have a real discussion (on the order of 10pp) describing the accomplishments made and putting the work in context as well as describing future directions envisioned by the student for carrying the investigation forward. In between should be three chapters of peer-reviewed published work or publishable work.
    If you add on appendices including co-authored pubs and small investigations that have not been completed as well as strain lists (or your field-specific equivalent), then the thesis becomes a highly useful document for your lab.

    Like

  25. bsci Says:

    @odyssey, You haven’t made a clear case to me what a modern scientific dissertation is supposed to be that a record of publications + an introduction & conclusion isn’t? Either you’re saying students should devote time to writing a book rather than publishable papers or you’re saying they should invest weeks or months of extra time rewriting their papers into a book format. What benefits does this extra hurdle give students and the scientific community? If you want to make sure students understand how their work fits into the larger scientific picture, faculty have oral exams and many opportunities for informal discussion to test a student’s understanding. What exactly do you think the value of a book-style scientific dissertation is?

    Like


  26. @ bsci

    My dissertation (2008) = 160 page book.

    Yeah. They are supposed to be that way. Re-phrasing the previous work is an excellent exercise in how to “write”.

    Please note I have a BA, not a BS. I actually had to take college level English classes.

    Also, my dissertation is uploaded to my website, so feel free to look. (Although I think most of the graphics didn’t make it, but I got the original if you wanna see ’em.)

    Like

  27. Odyssey Says:

    A dissertation should be a complete synthesis of the project(s) described therein, including full experimental details, all data – positive, negative and inconclusive* – and it should tell a story (or stories). There should be clear delineation of which data were collected by the student, and, if included, which by others (co-authors on any papers that have been published for example). The dissertation should clearly point the way towards future work, and provide sufficient details to allow others to not have to follow the same false paths and reach the same dead ends the student had to endure. Most of all it is an exercise in which the student demonstrates that they understand the big picture beyond just gathering data for papers. That they understand how their projects come together to make a more complete story, how that story fits into the existing knowledge base, and how to write all of that in a clear manner.

    Papers do not include full experimental details. Nor the negative and inconclusive data. They don’t outline precisely who did what (how often is the student sole author?). Papers rarely outline long term future directions. They do not include the details of false paths and dead ends. And they do not demonstrate that the student understands the big picture.

    Without all of the above the dissertation committee is not receiving a complete description of what the student has done and whether they have truly earned a PhD. IMHO.

    * But obviously not data from experiments clearly screwed up.

    Like

  28. qaz Says:

    I always say that there are two types of theses. It can be a hoop to jump through so you can go on to bigger and better things or it can be a magnum opus that sets you up in a field for a decade or more. There are advantages and disadvantages to each.

    Personally, my thesis turned into a book that was read by every incoming graduate student in the hundreds of labs that work in my field for more than five years. (I still hear that some PIs give their incoming student the book to read “to catch them up” to that year.) The thesis of the student who came after me in that lab was cited hundreds of times as the best description of his field for at least a half dozen years.

    I also know of students who stapled papers together, went off to industry and “made their millions”.

    Both are fine. It depends on your goals. And on quality work.

    Like

  29. bsci Says:

    @Allison, Re-phrasing the previous work is an excellent exercise in how to “write,” but why should this exercise be a hurdle towards a PhD? If you want to do this exercise, why not have people rephrase publications for a lay audience? That might be an even more useful exercise.

    @Odyssey, That’s a nice list but I’m again not sure why this all needs to be placed on a dissertation. Listing full experimental details and all data is a wonderful idea. I believe that’s called a lab notebook that is kept with the lab. I good lab notebook with some summary pages will be more useful than anything someone writes in a dissertation. If a student was a lousy record keeper, they obviously need to summarize their work somewhere, but I’m not sure why a dissertation needs to be that place. Relevant negative results should be part of publications already. If negative results are conveniently ignored, that’s a problem that extends well beyond what is or isn’t in a dissertation.

    If there were interesting side projects that were dead ends, I agree those could be in a dissertation, but I don’t see why they are an essential element.

    Listing what work the student actually did is important and I think that was a requirement for me. It required a few paragraphs of text in each chapter.

    Some more ideas for future work is nice, but that doesn’t expand a dissertation much beyond the papers stapled together.

    The focus on the big picture is also confusing to me. Isn’t this already in the introduction & discussion sections of most papers? We all agree that this can be expanded a bit in a dissertation introductory chapter, but it doesn’t need to be a masterpiece.

    Your view also makes it seem like members of a dissertation committee never interact with a student until they read the thesis. I still think a committee member will better evaluate if a student understands their own work and the bigger picture through conversation rather than by having them write 50 extra pages of text.

    Like

  30. DJMH Says:

    ” or it can be a magnum opus that sets you up in a field for a decade or more”

    But many biomedical science PhDs change fields significantly from their grad work to postdocs, or at least they do if they want to get significant new training out of their postdocs, so then you’d have written a magnum opus on stuff you may not be focusing on for the next decade anyhow.

    Like

  31. qaz Says:

    DJMH – exactly. It depends on your goals.

    It’s not just biomedical science PhDs. Lots of engineering PhDs just need the letters to get the job they want. On the other hand, there are lots of biomedical science PhDs who don’t change fields or who change fields less than they think they will. (A lot of the “fields” are broader than many grad students understand.)

    Like

  32. The Lab Mix Says:

    Lots of comments here already, but I agree with Odyssey on this one.

    To me, a doctoral thesis (at least in biomed) should essentially be the story of your training, warts and all. Provide the details between the experimental lines. Show the struggles, but also the victories in your project(s), and fully explain how you managed to go from one to another. Papers by default only show the shining jewels of your work, but realistically those results can be a small percentage of your labor. The thesis is the record of that labor. Of all the troubleshooting, of all the assay development, and the dead ends navigated. Ideally all this doctoral work produces papers, but that isn’t the direct point.

    Mind you, I trained in the UK, where PhDs are much shorter. So for 6+ year degrees, maybe expecting some kind of paper production is more warranted.

    Like

  33. magma Says:

    Very divergent opinions here.

    I think there is a strong bias for all of us: people who wrote a monolith convince themselves it was worthwhile, people who stapled papers together don’t want to think that they took a shortcut.

    I am definitely all for stapling papers (published or not) together, which is also what I did for my thesis. Perhaps this is also field dependent. In my area of Earth Science, the student is generally the driver of the work and does much more than just collecting the data. Typically the student is also (co)corresponding author. So the papers/chapters are really those of the student. As for negative results and full details: that is what the supplementary is for and anything that deserves to be in a thesis, deserves to be in a supplementary.

    Like

  34. odyssey Says:

    @bsci:
    Listing full experimental details and all data is a wonderful idea. I believe that’s called a lab notebook that is kept with the lab.

    Listing full experimental details for the data presented in the dissertation is a part of telling the story. That doesn’t in any way detract from the importance of good lab notebooks.

    Relevant negative results should be part of publications already.

    All I can say to this is your field must be very, very different to mine and those related to mine. It is very, very difficult to publish negative results in mine. And let’s not forget about the inconclusive data…

    If there were interesting side projects that were dead ends, I agree those could be in a dissertation, but I don’t see why they are an essential element.

    These are included both as a compendium of what others should avoid, and as a demonstration to the dissertation committee that the student has learned how to recognize such things and has learned from the experiences.

    Listing what work the student actually did is important and I think that was a requirement for me. It required a few paragraphs of text in each chapter.

    Not important. Imperative. Most biomedical research papers have multiple authors. Even if the student is first author that is only an indication that they contributed the the largest piece to the publication. And largest does not necessarily mean majority. Also, multi-author papers are often multi-authored. The student needs to be able to write up the work in their dissertation themselves.

    Some more ideas for future work is nice, but that doesn’t expand a dissertation much beyond the papers stapled together.

    “Nice”? If a student cannot do this they have not earned the degree. This is the hallmark of an independent researcher. Becoming an independent researcher is the whole point.

    The focus on the big picture is also confusing to me. Isn’t this already in the introduction & discussion sections of most papers?

    The big picture outlined by a student should go well beyond the abbreviated “why you should care about our paper” stuff included in manuscripts. Again, this is a hallmark of the independent researcher.

    Your view also makes it seem like members of a dissertation committee never interact with a student until they read the thesis.

    Not at all. But most committee members are more invested in other things the other 99% of the time, and most meetings with them do not cover the work of the student in its entirety. The dissertation is the place for that.

    Like


  35. @bsci
    Agree lay audience summary would be helpful.

    However.

    Considering the horrid prose I regularly read in biomedical literature from native English speakers, yeah- I’d say learning how to communicate properly is an issue.

    It’s not like science writing is the Great American Novel, you just need to be clear to your intended audience. What I read in today’s literature lacks clarity, not to mention concision. And basic grammar. (Paragraph breaks people! They are your friend!)

    More importantly, the dissertation should be used to continue your lab projects. So: full, tidy experimentals; all the data you ever collected- published and unpublished; and a thorough introduction so your PI can recall, 20 years from now, exactly what you did and why you did it.

    If you’ve been keeping good records, this can all be done rather quickly. I wrote most of mine over the course of a weekend. (Remember I had a few publications already. Did have to do a few sections with all the unpublished data.)

    If you don’t wish to communicate, get a Master’s. (I’d actually like to see a degree that’s like a “technical PhD”– where skill in lab and doing very technical, short reports is the goal; something that’s useful in industry. This is kind of what many EU nations do- plus, they grade all the dissertations.)

    Also, ditto what odyssey said about negative results. Not everything needs to get published- but it must be recorded in a neat fashion so those that come after you do not have to “re-invent the wheel” in the lab, or repeat your mistakes.

    Like

  36. bsci Says:

    Looking back on the comments here, I think there is little disagreement on what skills someone should demonstrate before earning a PhD. The question is how many of these skills should be demonstrated in a dissertation vs dissertation + personal interactions, lab notebooks, lab manuals, group talks, peer reviewed papers.

    For example, if someone writes horrid prose in a peer reviewed paper, like Allison mentions, I don’t see how having them write distinct prose for a dissertation that will be read by fewer people than the original paper will be any better. This isn’t to discount the problem, but to question whether the dissertation is a useful place to address it.

    I also agree with odyssey that many of the things you list could go in a dissertation, but, if a student has kept good communal lab records and has those skills evaluated in other formats, I see no reason to also put them in a dissertation.

    Like

  37. jojo Says:

    It seems like a lot of posters that are against including full papers as thesis chapters are upset that doing so will prevent the student from learning how to write and synthesize. This confuses me. If the student already did the work and wrote the paper, they have demonstrated their ability to write and synthesize.

    I wonder if some of the confusion is over whether the included papers were actually written by the student. I was taught that it’s inappropriate to include in the dissertation work you did not do and did not write up. But at least in my field (evoecobio), you are expected to have a majority of papers/publishable units that you wrote, lead the work on, and did the majority of the work yourself. A paper which was completely the work of a student can and should be included as a complete thesis chapter, assuming it fits with the rest of the dissertation and appropriate context is included in the first and last chapters.

    On the other hand, If a big part of the dissertation work was a multi author paper of which the student only contributed a small part, the student should summarize only their work as a thesis chapter or part of a chapter, while of course doing their own synthesize/big picture summary of where their work fits in (NOT copy-pasting from others’ writing, as that’s plagiarism). In the case of such a student, the dissertation writing is an essential part of their training that they would get in no other way, and a stapled paper is absolutely inappropriate in this case.

    Maybe a student being the corresponding/first author is just so rare in some other fields that it never comes up? Is the dissertation is the first time the student ever wrote anything? If so that’s too bad, students should have the opportunity to learn how to write a journal article from start to finish, as well as a dissertation.

    Like


  38. @bsci

    I know a lot of scientists who became scientists explicitly so they did not have to write. To some folks, writing is quite difficult. This doesn’t mean they don’t have the learn the basics, but I can defiantly empathize with technical experts who fail to see the need for fancy-pants prose.

    All of this does point the the sheer variability in PhDs. This is good: different fields must have different standards, different schools and PIs do things different ways. Diversity helps science progress. A PhD is not a certification like an MD, and dissertations are supposed to be unique and original.

    Maybe we do need to start having different “levels” of PhD though- like they do in Germany. Over there, a Dr. rer. nat. = full blown dissertation with original writing; re-phrases the papers and clarifies things- all that fun “science communication” stuff. Just Dr. = staple the papers together & go get a real job in industry. 🙂 (I think this is how it works. Then they have this whole “habilitation” process if you wanna be a prof which I think goes a bit overboard.)

    Also, I’ve used my dissertation quite a bit for teaching. It’s quite handy- lots of figures and examples that would not fit into a paper, but that do help clear things up for students.

    @jojo is also bringing up some good points. If you’re in a field where huge big projects are the norm, your dissertation can clarify what exactly your role was, and highlights your specific work and contributions.

    Like


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: