Jump to content

cyberwulf

Members
  • Posts

    839
  • Joined

  • Last visited

  • Days Won

    9

Reputation Activity

  1. Like
    cyberwulf got a reaction from rubotomy in Stats program by tiers?   
    Stats and biostats departments should really be ranked separately. 
     
    For stat, my rankings would be:
     
    TIER I
    Stanford; UC - Berkeley; Harvard; Chicago;
     
    TIER II
    UW - Seattle; CMU; Duke; UW - Madison; UM - Ann Arbor; NC State;
     
    TIER III
    Wharton; Cornell; Columbia; Minnesota; UCLA; UNC; Yale;
     
    For biostat, I would say:
     
    TIER I
    Harvard; Johns Hopkins; UW-Seattle;
     
    TIER II
    Minnesota; Michigan; UNC;
     
    TIER III
    UC-Berkeley; Penn; Emory; Columbia; UCLA; Brown; UW-Madison;
  2. Like
    cyberwulf got a reaction from endthestory in Before you start agonizing over your personal/research statement for stat or biostat, read this.   
    As an addendum, the one main exception to the "personal statements aren't important" rule is applicants who have an unusual academic record/background compared to the typical stat/biostat grad applicant. This could include, for example: 
     
    - Applicants who have been out of school for several (5+) years.
    - Applicants who are changing fields.
    - Applicants who lack prerequisites.
    - Applicants whose academic performance was affected by serious personal/medical circumstances (e.g., one semester of terrible grades due to death of parent, major illness, etc.)
     
    If you fall into one of the first three categories, we want to know why you chose statistics/biostatistics and how you think your background has prepared you for success in grad school. If the last category applies, it's important for us to know since it provides needed context for interpreting your previous academic performance.
  3. Upvote
    cyberwulf got a reaction from BL4CKxP3NGU1N in Hours per week for statistics /biostatistics phd?   
    90 hours? I'm not sure I believe that for one week, let alone a whole summer. In fields like stats and biostats, where virtually all of your work time involves intense thinking or doing (unlike lab-oriented fields where a decent chunk of "work" time is waiting for experiments to finish), it's just not possible for most humans to put in more than 50 or so productive hours per week. In grad school (at a top program), I was probably putting in about 40 hours/week spread across 7 days; some weeks less, others a bit more but rarely more than 50 and certainly never exceeding 60. I don't think I was way outside the norm.
    If you have to put in anything near 90 hours per week on a regular basis to be successful in a program, then I would argue that you're in the wrong program (or, at the very least, should be looking for a different advisor).
  4. Like
    cyberwulf got a reaction from 21andstressed in M.S. in Biostats Profile Evaluation for Fall 2022   
    I predict you're going to get in everywhere for a Masters. You would likely be competitive for PhD admission at most of the places on your list.
  5. Upvote
    cyberwulf got a reaction from LeoStat in Biostatistics PhD profile evaluation - Fall 2022   
    @bayessays Yes, most biostat programs admit 50% or more domestic for funding (training grant) reasons, but the applicant pool is usually >70% international. So the probability of being admitted is lower for international students. 
     
    Also, this profile isn’t particularly impressive for a domestic biostat applicant in 2021; solid performance at a decent undergrad followed by average performance at a decent MS program is borderline for admission to a top 10 program.
    (By the way, @LeoStat, please don’t take this the wrong way. Your results could be quite good with strong letters, and I hope my sense of your chances is overly pessimistic!)
  6. Upvote
    cyberwulf got a reaction from LeoStat in Biostatistics PhD profile evaluation - Fall 2022   
    I'm going to be a little bit of a downer here and say that I wouldn't bet on getting into a top 10 biostat program. The applicant pool has gotten insanely deep, and even places outside the top 5 can now afford to accept only the top handful of international students who apply. At our (top 10) program last year, I would estimate that we saw 30-50 international applicants with profiles at least strong as yours; we admitted fewer than 10.
  7. Like
    cyberwulf reacted to trynagetby in Best PhD programs for Causal Inference   
    Harvard, Berkley, UW Stats all have at least 3 very top/rising star type people doing causal inference research. An important distinction you have to make is whether you want to do "classical" causal inference  (propensity scores, average treatment effects, instrumental variables, potential outcomes framework) or "modern" causal inference (dags,judea pearl causal discovery,  reinforcement learning, adaptive designs etc...). Both are pretty hot right now but the flavor of research is extremely different.
     
  8. Upvote
    cyberwulf got a reaction from SteelBite in Before you start agonizing over your personal/research statement for stat or biostat, read this.   
    Every admissions season, many students applying to statistics and biostatistics programs are intimidated by the task of writing personal or research statements. Indeed, there is an entire sub-forum on GC dedicated to SoPs (Statements of Purpose) where there is much hand-wringing over how to craft the perfect text.   But while I can't speak for disciplines outside of the statistical sciences, I can confidently say that in stat and biostat, the evidence strongly suggests that personal statements have little impact on admissions.   I've written several posts about this in the past; here's a summary of why you should stop worrying so much about a 1-2 page essay:   1. Mathematical ability is best assessed through academic records and test scores (and to a lesser extent, letters), so it is generally quite easy to order students on this important trait.The pool of students applying to statistics and biostatistics departments isn't particularly deep, so that a major concern of even excellent departments is whether applicants can handle the requisite mathematical coursework and exams.

    2. Very, very few applicants have meaningful statistical research experience before starting graduate school. As a result, many students end up working on dissertations in areas entirely different than they were initially interested in... and this is totally OK!

    3. Funding in most (but not all) U.S. stat/biostat programs is allocated at the department level to the strongest incoming students, so applicants aren't typically "matched" to potential advisors who agree to fund them*. Rather, the department projects the total number of positions available and then tries to recruit up to that number of students. Once the students are on campus, they are then either assigned to a position or (ideally) have some choices available to them.   Given points 2 and 3, declarations in the personal statement such as "I am very interested in studying [X] with Professors [u,V,W]" usually carry little weight. They typically translate to: "[X] is a hot topic which I know very little about but sounds interesting, and I see on your website that Professors [u,V,W] list [X] as a research area." Which, again, is JUST FINE, since that's essentially all most people can credibly write.

    4. Research potential *is* important, but the best source of information on this trait is letters of recommendation, not a one-page essay. In some fields, part of showing research potential is demonstrating that you have already thought of a reasonable project that will turn into a dissertation. Since (virtually) no one applying to stat/biostat has a "shovel-ready" dissertation idea, research potential is generally assessed using some combination of mathematical ability, creativity, and perhaps some exposure to lower-level research, all of which are best evaluated using other parts of the application.   I don't mean to denigrate the personal statement too much. There are a few key things to avoid (eg. rampant grammatical errors, aimless rambling, saying you have no intention of pursuing an academic career if you are applying to a PhD program) and of course there will be exceptions to every rule, but in general, as long as the PS is competent it probably won't affect your chances of admission significantly.   
  9. Upvote
    cyberwulf got a reaction from bayessays in Lower-ranked/unranked Biostatistics PhD Programs for Mediocre Students?   
    With strong letters you should be able to crack a program ranked in the 15-25 range for biostat programs (think Iowa, Vanderbilt, Florida, etc.)
    If you can shoulder the cost (or get a scholarship/fellowship), you could definitely get into a top-10 Masters program, where good performance could put you in a position for much better admissions results.
  10. Upvote
    cyberwulf got a reaction from nauhark in Qualitative Reputations of Top Biostatistics Programs?   
    Here are some things which come to mind when I think about these programs. Of course, this is based on my observations (and narratives from others) and hence entirely subjective:
    Harvard: Light coursework, heavy emphasis on research. Can be competitive. Many graduates end up in Harvard-affiliated non-faculty (or contract faculty) positions.
    Hopkins: Heavy coursework. Fun environment (at least inside the building). Most well-known advisors very "data science"-y. 
    UW: Heavy coursework. Fun student experience. Thin on advisors that don't do variable selection/machine learning. Students tend to graduate with a broad background but few papers.
    UNC: Moderate coursework. Strong connections to industry. Big program with high student-to-faculty ratio. Students can get "lost" and take a long time to graduate (or not graduate at all).
    Michigan: Moderate coursework. Big program. Well-rounded, particularly strong in genetics/genomics and causal inference.
    UC Berkeley: Light coursework. High flexibility, "choose your own adventure". Not affiliated with a medical school, so more emphasis on methods than application.
  11. Like
    cyberwulf got a reaction from stat_guy in Qualitative Reputations of Top Biostatistics Programs?   
    Here are some things which come to mind when I think about these programs. Of course, this is based on my observations (and narratives from others) and hence entirely subjective:
    Harvard: Light coursework, heavy emphasis on research. Can be competitive. Many graduates end up in Harvard-affiliated non-faculty (or contract faculty) positions.
    Hopkins: Heavy coursework. Fun environment (at least inside the building). Most well-known advisors very "data science"-y. 
    UW: Heavy coursework. Fun student experience. Thin on advisors that don't do variable selection/machine learning. Students tend to graduate with a broad background but few papers.
    UNC: Moderate coursework. Strong connections to industry. Big program with high student-to-faculty ratio. Students can get "lost" and take a long time to graduate (or not graduate at all).
    Michigan: Moderate coursework. Big program. Well-rounded, particularly strong in genetics/genomics and causal inference.
    UC Berkeley: Light coursework. High flexibility, "choose your own adventure". Not affiliated with a medical school, so more emphasis on methods than application.
  12. Like
    cyberwulf reacted to possumvibes in Fall 2021 Statistics/Biostatistics Applicant Thread   
    Just wanted to give a quick shoutout to @bayessays for being around and giving solid input/advice to a lot of people on here.
  13. Upvote
    cyberwulf got a reaction from PluvianSprite in Qualitative Reputations of Top Biostatistics Programs?   
    Here are some things which come to mind when I think about these programs. Of course, this is based on my observations (and narratives from others) and hence entirely subjective:
    Harvard: Light coursework, heavy emphasis on research. Can be competitive. Many graduates end up in Harvard-affiliated non-faculty (or contract faculty) positions.
    Hopkins: Heavy coursework. Fun environment (at least inside the building). Most well-known advisors very "data science"-y. 
    UW: Heavy coursework. Fun student experience. Thin on advisors that don't do variable selection/machine learning. Students tend to graduate with a broad background but few papers.
    UNC: Moderate coursework. Strong connections to industry. Big program with high student-to-faculty ratio. Students can get "lost" and take a long time to graduate (or not graduate at all).
    Michigan: Moderate coursework. Big program. Well-rounded, particularly strong in genetics/genomics and causal inference.
    UC Berkeley: Light coursework. High flexibility, "choose your own adventure". Not affiliated with a medical school, so more emphasis on methods than application.
  14. Like
    cyberwulf got a reaction from humber in Qualitative Reputations of Top Biostatistics Programs?   
    Here are some things which come to mind when I think about these programs. Of course, this is based on my observations (and narratives from others) and hence entirely subjective:
    Harvard: Light coursework, heavy emphasis on research. Can be competitive. Many graduates end up in Harvard-affiliated non-faculty (or contract faculty) positions.
    Hopkins: Heavy coursework. Fun environment (at least inside the building). Most well-known advisors very "data science"-y. 
    UW: Heavy coursework. Fun student experience. Thin on advisors that don't do variable selection/machine learning. Students tend to graduate with a broad background but few papers.
    UNC: Moderate coursework. Strong connections to industry. Big program with high student-to-faculty ratio. Students can get "lost" and take a long time to graduate (or not graduate at all).
    Michigan: Moderate coursework. Big program. Well-rounded, particularly strong in genetics/genomics and causal inference.
    UC Berkeley: Light coursework. High flexibility, "choose your own adventure". Not affiliated with a medical school, so more emphasis on methods than application.
  15. Like
    cyberwulf got a reaction from AngelaBiostats in Qualitative Reputations of Top Biostatistics Programs?   
    Here are some things which come to mind when I think about these programs. Of course, this is based on my observations (and narratives from others) and hence entirely subjective:
    Harvard: Light coursework, heavy emphasis on research. Can be competitive. Many graduates end up in Harvard-affiliated non-faculty (or contract faculty) positions.
    Hopkins: Heavy coursework. Fun environment (at least inside the building). Most well-known advisors very "data science"-y. 
    UW: Heavy coursework. Fun student experience. Thin on advisors that don't do variable selection/machine learning. Students tend to graduate with a broad background but few papers.
    UNC: Moderate coursework. Strong connections to industry. Big program with high student-to-faculty ratio. Students can get "lost" and take a long time to graduate (or not graduate at all).
    Michigan: Moderate coursework. Big program. Well-rounded, particularly strong in genetics/genomics and causal inference.
    UC Berkeley: Light coursework. High flexibility, "choose your own adventure". Not affiliated with a medical school, so more emphasis on methods than application.
  16. Like
    cyberwulf got a reaction from az25340 in Qualitative Reputations of Top Biostatistics Programs?   
    Here are some things which come to mind when I think about these programs. Of course, this is based on my observations (and narratives from others) and hence entirely subjective:
    Harvard: Light coursework, heavy emphasis on research. Can be competitive. Many graduates end up in Harvard-affiliated non-faculty (or contract faculty) positions.
    Hopkins: Heavy coursework. Fun environment (at least inside the building). Most well-known advisors very "data science"-y. 
    UW: Heavy coursework. Fun student experience. Thin on advisors that don't do variable selection/machine learning. Students tend to graduate with a broad background but few papers.
    UNC: Moderate coursework. Strong connections to industry. Big program with high student-to-faculty ratio. Students can get "lost" and take a long time to graduate (or not graduate at all).
    Michigan: Moderate coursework. Big program. Well-rounded, particularly strong in genetics/genomics and causal inference.
    UC Berkeley: Light coursework. High flexibility, "choose your own adventure". Not affiliated with a medical school, so more emphasis on methods than application.
  17. Upvote
    cyberwulf got a reaction from Geococcyx in Qualitative Reputations of Top Biostatistics Programs?   
    Here are some things which come to mind when I think about these programs. Of course, this is based on my observations (and narratives from others) and hence entirely subjective:
    Harvard: Light coursework, heavy emphasis on research. Can be competitive. Many graduates end up in Harvard-affiliated non-faculty (or contract faculty) positions.
    Hopkins: Heavy coursework. Fun environment (at least inside the building). Most well-known advisors very "data science"-y. 
    UW: Heavy coursework. Fun student experience. Thin on advisors that don't do variable selection/machine learning. Students tend to graduate with a broad background but few papers.
    UNC: Moderate coursework. Strong connections to industry. Big program with high student-to-faculty ratio. Students can get "lost" and take a long time to graduate (or not graduate at all).
    Michigan: Moderate coursework. Big program. Well-rounded, particularly strong in genetics/genomics and causal inference.
    UC Berkeley: Light coursework. High flexibility, "choose your own adventure". Not affiliated with a medical school, so more emphasis on methods than application.
  18. Like
    cyberwulf got a reaction from Counterfactual in Qualitative Reputations of Top Biostatistics Programs?   
    Here are some things which come to mind when I think about these programs. Of course, this is based on my observations (and narratives from others) and hence entirely subjective:
    Harvard: Light coursework, heavy emphasis on research. Can be competitive. Many graduates end up in Harvard-affiliated non-faculty (or contract faculty) positions.
    Hopkins: Heavy coursework. Fun environment (at least inside the building). Most well-known advisors very "data science"-y. 
    UW: Heavy coursework. Fun student experience. Thin on advisors that don't do variable selection/machine learning. Students tend to graduate with a broad background but few papers.
    UNC: Moderate coursework. Strong connections to industry. Big program with high student-to-faculty ratio. Students can get "lost" and take a long time to graduate (or not graduate at all).
    Michigan: Moderate coursework. Big program. Well-rounded, particularly strong in genetics/genomics and causal inference.
    UC Berkeley: Light coursework. High flexibility, "choose your own adventure". Not affiliated with a medical school, so more emphasis on methods than application.
  19. Like
    cyberwulf got a reaction from tukey in Qualitative Reputations of Top Biostatistics Programs?   
    Here are some things which come to mind when I think about these programs. Of course, this is based on my observations (and narratives from others) and hence entirely subjective:
    Harvard: Light coursework, heavy emphasis on research. Can be competitive. Many graduates end up in Harvard-affiliated non-faculty (or contract faculty) positions.
    Hopkins: Heavy coursework. Fun environment (at least inside the building). Most well-known advisors very "data science"-y. 
    UW: Heavy coursework. Fun student experience. Thin on advisors that don't do variable selection/machine learning. Students tend to graduate with a broad background but few papers.
    UNC: Moderate coursework. Strong connections to industry. Big program with high student-to-faculty ratio. Students can get "lost" and take a long time to graduate (or not graduate at all).
    Michigan: Moderate coursework. Big program. Well-rounded, particularly strong in genetics/genomics and causal inference.
    UC Berkeley: Light coursework. High flexibility, "choose your own adventure". Not affiliated with a medical school, so more emphasis on methods than application.
  20. Upvote
    cyberwulf got a reaction from Euler17 in Under the current circumstances, do you expect the 2021 BIOSTAT admission + funding to be much harder to get?   
    I see biostat admissions getting more competitive, not because incoming classes are getting smaller (though in some places they may be, somewhat) but because of increased interest in the field of biostatistics due to COVID. Nationally, applications to schools of public health are up about 20%, and while a good chunk of that is in other fields (hello, epidemiology!) there's definitely a spillover effect into biostat. Anecdotally, we're seeing a higher proportion of applications from people whose profile can be summed up as "I'm a smart person who didn't intend to go into biostat but gee that sounds pretty cool so let's give it a shot".
  21. Like
    cyberwulf got a reaction from kingduck in ETS sent scores late   
    If you clicked "Submit" on SOPHAS before 12/1, that is generally considered as having met the deadline. After that point, you just need to make sure your materials are complete by the time reviewing starts, which at most programs is in January.
  22. Upvote
    cyberwulf got a reaction from kingduck in How to tell fourth letter writer I don't need them to write for me   
    No need to be fancy, just tell them you won't be needing a letter from them because you have a sufficient number from other faculty. Chances are, they'll appreciate that you're taking something off their plate. 
  23. Upvote
    cyberwulf got a reaction from kingduck in Auto-rejection Potential?   
    I don't know of any programs that uses an auto-rejection rule based on a hard GRE cutoff. So, regardless of the range in which they occur (within reason; obviously improving from a 145 to a 147 isn't going to make you more competitive), incremental improvements in the GRE score are likely to have similarly incremental effects on your chances of admission.
  24. Upvote
    cyberwulf got a reaction from pomegranateleaves in TIFU on my CV   
    Don't sweat it. At worst, you'll make someone reviewing your app chuckle.
  25. Upvote
    cyberwulf got a reaction from kingduck in TIFU on my CV   
    Don't sweat it. At worst, you'll make someone reviewing your app chuckle.
×
×
  • Create New...

Important Information

This website uses cookies to ensure you get the best experience on our website. See our Privacy Policy and Terms of Use