Educational Measurement/ Quant. Methods in Education

[Graham17]

Just wondering if other people who applied to, or are currently enrolled, in these types of programs have any thoughts on the different schools that offer such programs etc.

Edited January 15, 2013 by Graham17

[edgirl]

I'm doing the master's program in QM at Penn GSE this year. It's a slightly weird mix of courses (e.g., regression/ANOVA, test construction, policy research), and depending on your stats/quant background it may well feel too introductory, although it is possible to replace core courses with higher-level ones if you've already taken equivalent classes. I think it's a better fit for people who have sort of broad education interests, are thinking about working as policy researchers or analysts, and want to be able to understand/do solid statistical analysis for such research.

 

I think Columbia's programs are the most quantitatively rigorous and require the most quant background. But of course TC is expensive and offers kind of crappy financial aid. BC's ERME programs are also more straight quant, I think, than Penn's QM, though like Penn there is a broader focus on things like test construction rather than just a series of analysis courses. Feel free to PM me if you're thinking about Penn at all.

[missgord]

i totally disagree.  i think that upenn, BC, usf, UNC greensboro and TC all have less straight quant courses. they're more applied, in that they train people to become quant analysts of social science data and they care about your substantive issues. the programs that are more technical (aka, expect your dissertation to be a simulation and expect to take 0 substantive courses) are maryland, michigan, boulder, washington

[jc14]

I like this topic. I'm an applicant for the Fall 2015 cohort as well. 

I think TC is very rigorous in terms of course work. The program I'm talking about in particular is their M.S Applied Statistics program. I've also found Vanderbilt's QM program and NYU's Applied Statistics in SS to be just as rigorous. Of course this is all just based on my research. And my impressions are mostly just based off of the curriculum.

@missgord - what makes you say washington is more technical/rigorous quantitative work. I've been admitted to the UW Measurement/Statistics program, but I'm on the fence because I feel like it's not rigorous enough. Same with BC's ERME program. 

I'm very curious as to what other people think - both current and prospective students.

[nashville0808]

Columbia's MS in applied statistics is not as rigorous as you think. All courses are not difficult and focus on applied analysis. They are weak in theory. This may be true to most of the quantitative programs within schools of education.

[missgord]

I agree with @nashville0808 - they are weak in theory.  But, they are not meant to be theoretical programs.  That is what the MS in Stat is for.  These are applied stat.  So, it's just very different.  We are meant to use the formulas that the statisticians create to do research to influence policy.

 

@jasonchang15 have you attended any of the programs you're talking about? i dont know anything about NYU's program but I think vanderbuilt's program is MUCH harder than TC's program.  MUCH more challenging.  But, it really depends on what you want to study.  Imagine your dissertation.  Do you want to...

 

a.) propose a new model for a statistical theory and use simulations to defend it?

b.) use advanced quant to apply it to real social science data to influence policy decisions while simultaneously discussing the method?

 

if you want a, go to TC or Vanderbuilt (but Vanderbuilt would be a better fit).

if you want b, go to TC.  I don't think Vanderbuilt would permit you to do this.

 

The more methodological programs will not permit you to take substantive courses in the social sciences or any qualitative courses.  The more applied programs will require it.  That is the difference, and it is a very important difference.  Looking at the professor's research interests and publications will also help you understand.

 

Also, U Washington will definitely be rigorous enough.  I applied there as a PhD applicant and had a phone interview with them last week.  Halfway through the phone call we both realized it wasn't the right fit.  She expected a dissertation like example A, and I want to have a dissertation like example B.  I was rejected a few days later, and it was a good call on their part - I wouldn't have accepted.

 

Similartly, I had a very mathy friend at TC (MS applied stat) who was totally bored and frustrated by how easy the courses were and the requirements to take substantive courses when he wanted to take math courses.

 

I was accepted at ERME and will be attending, since they fully support example B research.  That doesn't mean that they aren't rigorous - it's an incredibly program and I know I will learn a lot about methods - they just have a different focus.

Edited March 2, 2015 by missgord

[nashville0808]

@missgord - I think people on this thread are interested in measurements, evaluation, and assessment (psychometrics) rather than policy analysis. Statistical methods used in the former area is quite different from ones used in the latter.

 

What I'm frustrated with courses offered in the former field (I have a good math background) is that instructors do not explain how formulas were developed. They introduce formulas from somewhere and explain how to use them. They don't use mathematics or prove theorems. So you will end up using them without knowing them in depth. You will always wonder where these formulas come from in your career unless you receive rigorous theoretical training. It depends on your satisfaction. If you just want to do applied stuff and don't care about theories, probably you don't need to know the details. A problem with this is that when someone develops new theories, which affect applied stuff, you won't be able to understand them until someone explains them to you. I don't feel good about this if I work as a researcher or a professor. If you go to a program in measurement, evaluation, and assessment, I would strongly recommend that you concurrently take courses at a stat department. I ended up taking many courses from stat/math departments.

[missgord]

@nashville I don't completely disagree. Where have you studied? Often your PhD experience is what you make of it. Plenty of schools will offer independent studies with professors that require complete theoretical work. EDMS at Maryland is one of them. The regular courses taught by the departments are taught to a wide range of students in the college of education so they can't become too theoretical because that is just not what that audience needs. The theoretical background is important but if you want to focus completely on theoretical you should do a degree in statistics or quant pysch instead of this field. Like I was saying, certain programs in this field are more theoretical than others and you should be aware of that and choose one based the compatibility of your research interests. These programs offer applies stat, evaluation and measurement. So, there is a variety under one umbrella. That is why I'm curious about where you've attended school - I want to know if you were in a less theoretical program (as I suspect).

[naomi6]

The great thing about TC statistic program is that you can be expose to the theoretical along with the applied  by taking supplemental courses from the department of Statistics at CU.