Preparing for Biostatistics

[espre]

I will be entering an MPH program this summer, and I'm preparing for biostatistics. I'm interested in knowing how similar the core statistics courses are between schools. I located the Hopkins online course and am working through it, but will that necessarily cover the material that will be taught at other programs? I'd also like recommendations for other resources and textbooks.

Also, what about epidemiology? Is that similar between schools? My MPH focus will be maternal-child health, so I will probably not be taking extra statistics or epidemiology courses.

Thank you!

---

Members don't see this ad.

 

[Wiingy]

The biostatistics courses have been much gentler than I expected. For the non-biostatistics programs, much of the focus is on practice instead of theory (no derivations and almost no calculus). At UMich, the first biostat course starts with the normal distribution and ends with multiple linear regression. The second course picks up where the first left off, and ends with logistic and poisson regression.

I'd imagine that most biostatistics courses are similar at the different schools - if you have the email of students currently attending the program you'll be attending this summer I'd just ask them for tips about your specific program. Hopefully the info helped.

 

[espre]

Any more suggestions? I'm really hoping that people already enrolled in MPH programs can provide some information on preparing for biostatistics and epidemiology.

 

[Nibor6000]

My intro biostats course was exactly as Wiingy described. For me, intro biostats was much easier than intro epi, but that was because the epi instructor couldn't write a test to save her life.

If you have taken any kind of stats class, or if you can add, subtract, multiply, and divide, you'll be fine.

 

[cakes]

I'd be surprised if the approach to the intro classes were very different, but the professor can make a difference for a lot of people. My intro epi course wasn't any harder than biostats, but I wasn't a fan of that professor so much. hard and easy are very relative terms most of the time.

If you're looking for resources to prepare for a class you're already registered for, you could probably just email the professor. Find out what text book they're using. I'm not sure why you really want to compare to other schools at this point, but you could just buy a few different texts (or visit the library) to broaden your knowledge on any subject. I actually did that (hang out at the library) for my epi class because I really didn't like the text she chose for the class.

 

[espre]

I'd be surprised if the approach to the intro classes were very different, but the professor can make a difference for a lot of people. My intro epi course wasn't any harder than biostats, but I wasn't a fan of that professor so much. hard and easy are very relative terms most of the time.

If you're looking for resources to prepare for a class you're already registered for, you could probably just email the professor. Find out what text book they're using. I'm not sure why you really want to compare to other schools at this point, but you could just buy a few different texts (or visit the library) to broaden your knowledge on any subject. I actually did that (hang out at the library) for my epi class because I really didn't like the text she chose for the class.

Thanks for the information. I'm not enrolled in a course presently; I don't know yet which school I'll be entering - not all decisions have been announced yet. I've been using the Hopkins online material to prepare, and was wondering if this would be very similar to that found at other schools.

Any other strategies for preparing for biostatistics? What material did people find most confusing or difficult?

 

[JMM051]

I have it narrowed down to two schools and was able to find the text they use online at a used book store for $6! I figured getting acclimated to the text they use will help when instruction starts in the fall.

 

[CameraMan]

I have it narrowed down to two schools and was able to find the text they use online at a used book store for $6! I figured getting acclimated to the text they use will help when instruction starts in the fall.

Honestly, the biostats courses they teach in MPH programs are a bit of a joke. Same courses they use for nursing students and social research students.

No calculus. In fact, besides basic probability (e.g., odds of hitting TTH or HHT in a row on a coin toss), you won't have to do anything by hand. I took an undergrad stats course years ago and was able to test out of my biostats 1 course. I also took a few PhD level stats course (i.e., missing data analysis, social network analysis, multilevel modeling) and can tell you the difference is just night and day. The biostats courses in MPH programs is just intended to give you enough information that when you read a study that talks about "cluster random control" or the type of distribution, you understand. It's not meant to be a course where you can just sit down and start writing papers at the end of it. That said, I'm currently a TA in the biostats dept right now so if you want to see some sample notes, PM me. (You can also find quite a few of them online, which is where I go to find a foundation for mine.)

 

[cakes]

espre - gottcha. If I were you, I'd weigh my decision heavily on courses in my concentration and worry about the core courses after.

Honestly, the biostats courses they teach in MPH programs are a bit of a joke. Same courses they use for nursing students and social research students.

No calculus. In fact, besides basic probability (e.g., odds of hitting TTH or HHT in a row on a coin toss), you won't have to do anything by hand. I took an undergrad stats course years ago and was able to test out of my biostats 1 course. I also took a few PhD level stats course (i.e., missing data analysis, social network analysis, multilevel modeling) and can tell you the difference is just night and day. The biostats courses in MPH programs is just intended to give you enough information that when you read a study that talks about "cluster random control" or the type of distribution, you understand. It's not meant to be a course where you can just sit down and start writing papers at the end of it. That said, I'm currently a TA in the biostats dept right now so if you want to see some sample notes, PM me. (You can also find quite a few of them online, which is where I go to find a foundation for mine.)

Well then, it seems there are some differences in approach. I attended the lower intro class for a number of reasons and we did everything by hand. We learnt how to use some basic computer programs, but only after going though all the concepts. We touched on topics much beyond basic probability as well. I didn't find it hard, but it was a perfect transition back into school after 5 years.

I certainly wouldn't call the class I took a joke (it's really much too subjective anyway), but it's true that the instructor is pretty nasty at what she does. It all has to do with where you as a student are coming from in terms of what will be challenging or not. Chance are there will be an option for a lower level geared towards students not concentrating in epi or stats and those who have never taken a class in statistics before.

 

[CameraMan]

I'm honestly a little surprise they made you do t-tests (or any sort of nonparametric by hand). That is really quite rigorous--I'm really glad I didn't have to do any of that. I guess the philosophy of our department was that we needed to know how to do these things in a normal setting (e.g., how to collect data for a study, input the data, run diagnostics on the distribution, and run meaningful tests).

In the classes I had to TA for, the diversity of students ranged from MPH students to undergrad (and graduate) nursing students to dental and med students to career changers who haven't been in school for a decade so when we made assignments or lectures, the target is not for them to be able to do any of it by hand since they'll never have to. The target was simply for them to know how to do it in R or SPSS, know what the computer was "magically" doing for them behind the scenes, know when to use each test appropriately, and how to understand the diagnostics.

A large part of some class sessions was just reviewing a long and tedious public health article and going over the statistical tests, why alternatives weren't use, possible weaknesses, distribution issues, etc. The book we used (Biostats by Rosner) was referred to only for the theory, math, and sometimes for sample problems.

Also, I didn't mean to make it sound like you would only be doing basic probability. You should really be finished with that lecture within the first week or two, but knowing how to do it backwards and forwards will help you the rest of the way.

 

[cakes]

Maybe it depends on whether you are taking public health classes at a medical school or going to a school of public health? CameraMan, where did you take those courses? The BU program is only mph students with a few ma/ms students. However the ma/ms students are probably getting the masters in either epi or biostats and rarely take the lower level intro course anyway.

Either way, a t-test or ANOVA with a small sample isn't bad. We only used small sample sizes when learning the building block of the statistics analysis. The professor I had authored the textbook we used and liked it: Essentials of Biostatistics in Public Health.

 

[abcabc1]

Honestly, the biostats courses they teach in MPH programs are a bit of a joke. Same courses they use for nursing students and social research students.

No calculus. In fact, besides basic probability (e.g., odds of hitting TTH or HHT in a row on a coin toss), you won't have to do anything by hand. I took an undergrad stats course years ago and was able to test out of my biostats 1 course. I also took a few PhD level stats course (i.e., missing data analysis, social network analysis, multilevel modeling) and can tell you the difference is just night and day. The biostats courses in MPH programs is just intended to give you enough information that when you read a study that talks about "cluster random control" or the type of distribution, you understand. It's not meant to be a course where you can just sit down and start writing papers at the end of it. That said, I'm currently a TA in the biostats dept right now so if you want to see some sample notes, PM me. (You can also find quite a few of them online, which is where I go to find a foundation for mine.)

I agree. The Biostats courses offered to general MPH students and, in some cases, even Biostats MPH students, are not that rigorous. They re more general and, as they are easier, they also provide the students with less marketable skills. The statistics courses in MS and PhD programs are much more serious and Math heavy, with the PhD courses incorporating an added level of theory, proofs, etc beyond those in the MS.