Built a web app to visualise GRE & GPA averages and distributions based off grad cafe survey.

[juanmesh]

The search parameters are pretty much the same as grad cafe's.

https://grad-cafe-visualizations.herokuapp.com/

[hpgenomix]

this is sick

[svent]

Pretty cool. How specific is the search function? For example, does Texas A&M include all the UT Austin, UT Dallas etc. results? Does University of Arizona include Arizona State?

[juanmesh]

6 minutes ago, hpgenomix said:

this is sick

lol

[juanmesh]

2 minutes ago, svent said:

Pretty cool. How specific is the search function? For example, does Texas A&M include all the UT Austin, UT Dallas etc. results? Does University of Arizona include Arizona State?

I don't really know. I simply transfer the search parameters to the grad cafe site; so whatever they're using is what I'm using. But based on my experience, I'm guessing Texas A&M does not include the other schools.

[svent]

I see. Probably a good tool for most schools, but if there's a reasonable chance of results pulling in multiple schools, I'd analyze the results manually. (UA/ASU, TAMU/UT, UPenn/PSU, etc.).

[juanmesh]

24 minutes ago, svent said:

I see. Probably a good tool for most schools, but if there's a reasonable chance of results pulling in multiple schools, I'd analyze the results manually. (UA/ASU, TAMU/UT, UPenn/PSU, etc.).

If you were to search for "Pennsylvania", you'd get results from multiple schools.

The major issue is the survey results are entered by human beings so at times, respondents simply write "Penn".

I could very easily create an option to release the results (downloadable as a file) after some of the pre-processing I currently do. Is it something you'd be interested in?

[svent]

I agree. Sometimes people will enter UC San Diego, sometimes just UCSD. Some people have even mistakenly entered San Diego State. To be honest, I don't really have any use for it myself. I already know where I'm going, and don't expect to ever read the results page again.

[juanmesh]

22 minutes ago, svent said:

I agree. Sometimes people will enter UC San Diego, sometimes just UCSD. Some people have even mistakenly entered San Diego State. To be honest, I don't really have any use for it myself. I already know where I'm going, and don't expect to ever read the results page again.

Yeah, there's this idea that programmers have: trust nothing, absolutely nothing, that comes from a user.

All right, might still add it in case folks want to look over the raw numbers themselves. I built this when I kept looking through the results page. I'll probably go there once more to enter my results when my waitlist turns to a rejection at this one school.

[svent]

I don't know what QM is, but if you get into a software engineering job after school at some point, this looks like a nice project to have on your resume, regardless of how many people use it.

[juanmesh]

4 hours ago, svent said:

I don't know what QM is, but if you get into a software engineering job after school at some point, this looks like a nice project to have on your resume, regardless of how many people use it.

Maybe, feels really light but yeah, I'll definitely list it somewhere. I'm working on something that's a lot heavier, and I feel it's much more interesting. It visualizes data too but for actual classrooms - the Canvas LMS system.

QM is quantitative methods. I should be joining the Quantitative Research, Evaluation and Measurement (mouthful) program within the Ed. Studies program at Ohio State. My background is Information Systems, but don't plan to work in it. Hope to use the skills I've learned though, they're always a great plus, even if it's simply for productivity, like automating basic tasks.

Here's a related xkcd joke for "programmers"

[Warelin]

Any chance of you including Subject scores in the data? =)

[juanmesh]

2 hours ago, Warelin said:

Any chance of you including Subject scores in the data? =)

I could but I figured they're not reported very much, maybe I'll add it this weekend(?) Subject scores are typically within the 200 - 990 range, right? Or it's not always this way?

[Warelin]

3 hours ago, juanmesh said:

I could but I figured they're not reported very much, maybe I'll add it this weekend(?) Subject scores are typically within the 200 - 990 range, right? Or it's not always this way?

You are correct and I think it's always been that way.

[kaykaykay]

nice. department would be a nice addition!

[juanmesh]

11 hours ago, kaykaykay said:

nice. department would be a nice addition!

How would this work?

[Komugi]

On 4/13/2016 at 4:44 AM, juanmesh said:

The search parameters are pretty much the same as grad cafe's.

https://grad-cafe-visualizations.herokuapp.com/

Thanks a lot! That would help me choose which schools to apply to.

Edited April 28, 2016 by AhmedMA

[juanmesh]

5 hours ago, AhmedMA said:

Thanks a lot! That would help me choose which schools to apply to.

You're welcome. I'm not sure what you mean though. You mean you might consider it as a filter for which schools/departments your GRE scores might work at?

[Warelin]

On 4/27/2016 at 1:22 AM, kaykaykay said:

nice. department would be a nice addition!

But it already does that. Just type the school name followed by the department.

[Komugi]

7 hours ago, juanmesh said:

I'm not sure what you mean though. You mean you might consider it as a filter for which schools/departments your GRE scores might work at?

Yes, I think I should not apply to schools that accept an average gre score higher than mine, though as international application I am already not so sure which schools I should apply to. 

Is there a way I could know how many samples were used to generate the statisitcs for each search? For example, if I search for electrical engieering, the app says it used data from the most recent 20 pages of search results. Since some applicants submit their scores and others do not, I cannot simply multiply the number of submissions per page by 20 and expect to the result to be the number of samples used. Is there a way I can know the sample size used to generate the statistics?

Thanks

Edited April 29, 2016 by AhmedMA

[juanmesh]

@AhmedMA Well, I'll try to add that this weekend, maybe tomorrow. It shouldn't be so difficult for me to do this. Maybe a summary table below the graphs.

[juanmesh]

4 hours ago, AhmedMA said:

Yes, I think I should not apply to schools that accept an average gre score higher than mine, though as international application I am already not so sure which schools I should apply to. 

Is there a way I could know how many samples were used to generate the statisitcs for each search? For example, if I search for electrical engieering, the app says it used data from the most recent 20 pages of search results. Since some applicants submit their scores and others do not, I cannot simply multiply the number of submissions per page by 20 and expect to the result to be the number of samples used. Is there a way I can know the sample size used to generate the statistics?

Thanks

Done! I think.

[Komugi]

4 hours ago, juanmesh said:

Done! I think.

juanmesh,
This is just too awesome! Thanks! I really appreciate you taking the time to add the functionality. Now I can get a sense of how reliable my inferences would be.

[fernandes]

Good job mate! Did you use Python for doing it? 

Also, I have a suggestion: it would be nice to narrow/approximate the average calculated value to two decimals only.

Edited April 29, 2016 by fernandes
suggestion

[juanmesh]

5 hours ago, AhmedMA said:

juanmesh,
This is just too awesome! Thanks! I really appreciate you taking the time to add the functionality. Now I can get a sense of how reliable my inferences would be.

You're welcome.