Skip Nav

DataIsBeautiful

Step 1: Write a winning dissertation proposal

❶It's the years you spend developing your unique contribution to the field, the code or the experiment that demonstrates that contribution, and eventually collecting the results that verify it. I know Papers3 offers an export option to BibTeX but it doesn't produce consistent results.

phd dissertation help how many pages

You have Successfully Subscribed!
How to Write Your Best Dissertation: Step-by-Step Guide
Want to add to the discussion?

Sometimes that takes 50 pages. Other times it takes Other than that, this is neat! Kyle, thanks — I figured I was missing something super simple — its been a long day. So the colors are meaningless then? Yes, the colors are useless. They convey no information that is not already transparently conveyed. For that matter, the box plot itself is thoroughly obsolete. As for the topic of the post, my math dissertation was 88 pages.

But his supervisor, Tarski, would not give him his PhD. He thought the thesis was too short. So there is a lower bound. The chart would be more useful if the majors were ordered by median or average dissertatin length.

But it is interesting to note that the more mathematical and objective the mayor, the less pages needed. Out of curiosity, Nathan, what was your dissertation on? That, of course doubles the plage length. Better, what was the word count in the dissertations? Mine started out much longer, but as I started conforming to target journal format standards, I would say the overall length was reduced by a third or so. Is this true outside of ecology as well? This is very interesting.

A more helpful plot would count theses by words, instead of pages, as a measure of content. I only took a few history classes in college, but I probably bought a dozen books that started out as PhD theses. They were way too dense to be best sellers. This is true, but if it's truly interesting research and the writer is good at storytelling , it gives you a hell of a base to start with and write some required reading and a bestseller.

At that point you're not writing for a wide audience but for an academic one, which means the language is vastly different than that which a wide audience is used to.

Also, you're likely examining esoteric things like historiography with the assumption that your readers again, an academic audience has a foundational knowledge on the theoretical framework. The average person doesn't have that foundation and would be completely lost reading a history dissertation. What's your area of study? I'm just starting analyzing data for my diss I'm in English rhetoric and composition , so not too far from the top!

I think most people in my program have written around pages, which seems reasonable. My boss just did his dissertation for CS, less than 50 pages, no problem. It's not about pages, it's not grade school, and most of the multiple-hundred page dissertations are full of charts, graphs, images, etc. The bonus is that I've already published a lot of my work, so it's mostly written up already! I'm a comp sci undergrad. What are you doing your research in?

I'm curious what is going on in the field at a PhD level. There's a ton going on in computer science right now. I study artificial intelligence and evolutionary computation. Here's a short-ish research statement by me: Reading about your research is super interesting! Keep up the good work! I'd like to add that my largest LaTeX file is about 30 pages of abstract algebra notes. It took a really long time to type, and that was pretty much transcribing notes.

I can't even imagine what doing pages of math research is like. My dad has a PhD in Physics and did everything on a typewriter. He also walked to and from university in the snow, up hill both ways. It'd still be pages of published LaTex - though I'd be impressed to see someone write an entire dissertation in one. I need to learn LaTex. I've seen MS Word used, and it freaks me out every time. LaTeX is the most common sane way to do it, and I try to encourage every undergrad I know to get started on it earlier rather than later.

AIAA actually has its own LaTeX template that you have to use when writing any papers you intend to submit to any one of their conferences or publications. I think other disciplinary organizations probably have theirs too.

Recently though, Markup Markdown is gaining a lot of traction because there are some document processors out there right now that combine its simplicity with inline LaTeX for mathematics. YMMV of course, but this is my personal observation at least in my particular subdiscipline of Aerospace Engineering. I've given it a try myself and I have to admit, it's growing on me.

I'm particularly partial to Pandoc. Using Word is likely to throw me into fits of rage and result in damaged computers. I dislike MS Word as much as the next sane person and I love LaTeX as much as the next computer scientist , but it sort of make sense for people who already know how to use it to simply use MS Word. If you don't have big mathematical equations to format, why would you bother with learning LaTeX when you know that the journal editors are going to go through your paper and simply copy paste the content and edit it themselves?

What we need to do in my humble opinion is push Markdown as a standard format for "normal text with images" and require it at university. And create good editors for it, with enough features to compete with MS Word. I thought that markdown was more for web pages, and not for actual papers. I definitely see the appeal of using it for web pages no floating figures, lists are easier, etc, etc , but why should you use it for actual documents? Markdown owes much of its recent resurgence in academia to the development of Pandoc -- a multi-format document converter.

The link explains the great features of this thing, but I'll elaborate a little bit myself on why it's so appealing. For starters, Markdown is readable in plain-text. This takes the headache out of collaborative publications. Unfortunately there are still a fair number of academics out there who aren't LaTeX-aware, or are outright hostile towards it due to the very verbose syntax.

Markdown circumvents these difficulties. Secondly, BibTeX is getting outdated and cumbersome. It is not used by any of the major bibliography management programs out there, such as Mendeley, Zotero and Papers. I know Papers3 offers an export option to BibTeX but it doesn't produce consistent results. Pandoc, in the meantime, handles CSL format directly. I also consider Pandoc's citation syntax to be far superior to LaTeX, but that's personal opinion. Third, and perhaps most important, is that Pandoc is capable of handling hybrid syntax within the same document or workflow.

I mean two things by that:. It can parse in-line LaTeX within Markdown, which means that you can cherry pick only the specific styling or math tools that you need from LaTeX while preserving Markdown's simplicity everywhere else.

It can parse Markdown files alongside a LaTeX style sheet. And finally, because Pandoc is a local document processor just like LaTeX, the documents themselves can be hosted on code repositories like GitHub or BitBucket for efficient collaboration with robust version tracking, and then converted to identical PDF outputs by individual collaborators.

So in essence, it preserves the collaborative advantage of LaTeX by using the same "code development" paradigms that GUI word processors cannot provide. The end result is indistinguishable, but I find the workflow to be simpler. I realized that I made a mistake earlier to mention Mou when I really should have mentioned Pandoc instead. I like Mou a lot, but I don't use it as a processor. I use it as an editor with an incomplete quick-preview function. The actual document generation is done via Pandoc, so that's more pertinent to the discussion at hand.

Yes, but my understanding is that use of LaTeX in a field is directly correlated by how mathy the field is. It's standard in physics and CS as well, and people in those fields e. Standard in economics and statistics. I write everything in LaTeX or just plain text files anymore regardless of how much math or code they have. We use it all the time for papers and presentations. It's pretty much the expectation for graduate papers, and lots of profs use it instead of Powerpoint the Beamer package allows creation of presentations.

Everybody in my cohort has moved to google docs or LaTeX. Word is bloated and evil. The philosophy department at my alma mater uses LaTeX for everything. Most of the undergrads use regular word processors, but some of the ones that end up taking intermediate logic also end up using LaTeX for everything.

I was the only person in my department that I know of who used LaTeX to write my thesis. Everyone else used the Word template provided by the university. I was also the only person that I know of who didn't have my thesis returned by the graduate school for formatting revisions. Last semester I TeXed typed pages of stuff for homework. I'm in my first year of grad school for math. Admittedly, there are probably a bunch of half blank pages in there I just had Acrobat combine the pdfs to get the number of pages , but that's still more than pages in one semester.

I'm probably on track to do the same this semester. Of course, this is all homework stuff. How do they make a living while they're in school. If you do "cultural anthro" and hang out with head hunters, you can learn a few tricks There are lots of teaching opportunities for scholars with a Master's degree, which many obtain as a step in their PhD program.

Also, many obtain outside employment, sometimes unrelated to anthropology, and work on their dissertation in their spare time. You're usually spending at least 18 months to 2 years doing your research in the particular culture you're studying. It is relatively easy for that to end up being 8 years in an anthropology Ph. Average is about 6 some say 7.

Immersion takes time to get truly wonderful results and is a minimum two year commitment. If there weren't all the pre reqs in the first years we'd be much happier. Makes sense that the data-heavy majors are on the shorter side of things since they can convey a substantial portion of their premise with formulas and graphs as opposed to lengthy explanations of social phenomena for example. I, for one, would not have predicted that physics and chemistry would be at opposite ends.

I think at least some of this is random cultural drift. Makes sense to me. A physics dissertation would be mostly math, which is really information-dense, while in chemistry you have to include drawings and, depending on the research area, potentially an absolute shit ton of experimental data.

Let's not forget about specificities on writing out certain reactions were done in certain conditions such that it can be perfectly replicated by other chemists You can replicate chemical research from the dissertation alone?

I'm in physics, and there is no way in hell you could do that without all the preliminary information to papers, all lab books and access to the source code repositories The principle at play for most of us in the natural sciences is that any other person with similar training should be able to replicate your experiment. If your experiments are not reproducible, the findings are not real.

Graduating philosophy undergrads are beasts at the LSAT, but that's all I got off the top of my head. I combined mine B. Writing grants in fire prevention, discussing my moral obligation in risk vs reward situations, upholding fire code law , etc. I have a dissertation from a professor named David Enoch, he studied under a super-famous contemporary named Derek Parfit.

Enoch's dissertation at NYU is about pgs. Wittgenstein did it in 75 pages. There's a bumper sticker in that somewhere, surely. No one's finished a This is particularly important considering many faculties in the UK and Ireland will specify a word count, particularly in humanities and the social sciences this seems to be typically somewhere between 80k and k.

Sometimes not even your committee I wrote my 40 page senior seminar paper last semester for my history undergrad and it was miserable. The thought that I would need to do 7 times that to be remotely close to average makes me want to vomit.

Junior year I wrote a 35 page paper on food adulteration and the pure food and drug act and if it helped solve the problem. I thought it would be an easy short paper since it was just a 12 page max paper.

I went over and by page 35 I was only half way done with my argument and history of food adulteration.

I now know I can write a long fucking paper I had so many sources my bibliography took about 3 pages single spaced. Or they have an awful lot of quantitative data or have great editing skills. A long thesis is not necessarily a good thesis. I bet that one Natural Resources Science and Management major would feel pretty silly after seeing this. Seen this before and I remember from last time, someone pointed out that sociology and statistics have no outliers.

Economics is way down at the bottom, but applied economics is about two thirds of the way up. Can someone tell me what that means?

Economics PhD dissertations being short has more to do with how math intense they are than anything else--and not math as in the data tables kind of way, but more about math as in the real analysis kind of way. Saving trees since Galois wrote down Galois theory in a letter the night before he was shot in a duel. I've always heard the saying that given a 3 person committee only 3 people will ever read your thesis. It happens sometimes in History. I edited a bibliography for a former lecturer and he cited two unpublished doctoral theses; I've seen it happen in other works in the same area too.

Also, from seeing my mother's reaction to my funding proposals, I very much doubt she'd sit through any potential finished thesis of mine. For a lot of the numerical disciplines I'm in Aerospace , the publications you put out over the course of your research ultimately get distilled into your thesis. You could say that it just kinda writes itself over time. The difficult part of the work isn't really the writing bit. It's the years you spend developing your unique contribution to the field, the code or the experiment that demonstrates that contribution, and eventually collecting the results that verify it.

By the time you actually get to the point where you have to articulate those on paper, you already know what you're talking about extremely well and the explanation just kind of rolls off your fingers and onto the keyboard without too much trouble. I love writing -- even papers -- but there are lots of good reasons not to wade into the swamp that is academia.

Then you add some smart sentences to string them together into one theme. It gets easier as you understand your field more, but it's not one giant idea done at once, it's several small ideas put under an umbrella. In Australia that's known as "PhD by publication" as opposed to the more traditional "PhD by thesis" and is becoming very common due to the pressures placed on students by universities to publish.

If you want half a chance to snag an academic position after graduation you better have a solid number of peer reviewed papers published, in a high impact journal with a good h-index. I picture a lot of this in your daily life: Math and Computer Science papers aren't very long because typically algorithms are short. Like the algorithm for doing a back prop neural network is like 40 lines of code maybe, and those lines don't fill the entire page.

The rest of it is explanation, which is never a long explanation. Math is just so efficient in explaining itself because it assume, so much back knowledge. Same thing with economics. I imagine most economics dissertations have tons of graphs, formulas, charts, etc. Must have been a helluva paper. Rule of thumb is the shorter the better. John Nash's was 32 pages. And it had 2 citations, one of which was himself. The typical math thesis is based on a mathematical result that either is or will be published as a journal article of less than 50 pages.

The length of many math theses are dominated by the author's decision or sometimes their advisor's decision about how much background to include. Many mathematicians can "get away" with a 30 page thesis but choose not to.

Cool, didn't know that. And, like English majors say, if I had more time I'd have made it shorter. I'm masochistically pleased to see my biochem thesis looks to have been about 10 pages over the thickest biochem thesis in the given sample set. I was a math major, and I'd like to point out that one page of a written math paper can take a lot of time to work through. They often assume quite a bit.

Welcome to Reddit,

Main Topics

Privacy Policy

In my view, your task is to locate the minimum pages or words required. You may have a mentor who will answer a straightforward question. But if not, your first step is to look over as many dissertations as possible that have been passed by the appropriate committees, particularly in the last two years.

Privacy FAQs

A PhD thesis should have as many pages because of Formatting a thesis which is page long or consists of 40, words needs real hard work. And students, who have completed writing the document, feel weary of correcting the format of the document.

About Our Ads

As many PhD candidates are wont to do, Beck took the pressure of readying for his defense and channeled it toward an incredibly interesting (if entirely thesis-unrelated) side project. Apr 15,  · The data contained 2, records for students that completed their dissertations since The range was incredibly variable (minimum of 21 pages, maximum of ), but most dissertations were around to pages. Interestingly, a lot of students graduated in August just prior to the fall semester.

Cookie Info

phd dissertation help how many pages Apr 15, My advisor once told me about a student in Chemistry that produced a dissertation that was less than five pages. Dissertation persuasive essay writers workshop smechowski Phd Dissertation Help How Many Pages how to start an application essay for college albert camus essays online. How to Write Your Best Dissertation: · Complete Your Dissertation or Thesis in Two Semesters or Less phd dissertation help how many pages [Evelyn Hunt phd dissertation help how many pages Ogden] .