Welcome Guest ( Log In | Register )

Outline · [ Standard ] · Linear+

Economics Statistical Analysis for Dissertations?, R, Python, Stata, SAS, etc

views
     
TSExoflare
post Nov 25 2017, 04:45 PM, updated 8y ago

Enthusiast
*****
Junior Member
869 posts

Joined: Nov 2017


Hello all!

I was just wondering, but when it comes to writing dissertations, I was wondering if most universities in Malaysia teach students the specifics of processing high-dimensional or large-scale datasets?

I perform econometrics/statistics consulting, and a client of mine recently asked for some help in processing data for his dissertation with a statistical package (as he had limited experience with transforming variables, performing or interpreting multidimensional data regressions, and needed assistance on that front), and I was wondering if there's maybe a gap between what PhD level research professors want from their students, and the availability of instruction in producing statistically valid and professionally written reports, which I would like to fill.

I provide assistance in SPSS/PLS/EViews/STATA and help coach students in statistics/econometrics both through refresher courses, as well as through individualized help for specific projects.

Let me know about the specific requirements of your project, and email me at victortanws@uchicago.edu!

This post has been edited by Exoflare: Jan 26 2018, 02:21 PM
zemega
post Nov 25 2017, 06:46 PM

Enthusiast
*****
Junior Member
723 posts

Joined: Jul 2008
Is that client a phd candidate? For the candidate sake, he should do it himself, else he won't be able to answer the questions on viva. For your sake though, go for it. Statistics are important in determining whether their research is sound or not.

Just remember, besides doing the analysis for the candidate, be sure to make small write up of what you did and the significance of the output.

Besides, just because he's a PhD candidate, doesn't meant he is tech savvy or knows programming well. If given a small data set, he probably can do it manually.
TSExoflare
post Nov 25 2017, 10:43 PM

Enthusiast
*****
Junior Member
869 posts

Joined: Nov 2017


QUOTE(zemega @ Nov 25 2017, 06:46 PM)
Is that client a phd candidate? For the candidate sake, he should do it himself, else he won't be able to answer the questions on viva. For your sake though, go for it. Statistics are important in determining whether their research is sound or not.

Just remember, besides doing the analysis for the candidate, be sure to make small write up of what you did and the significance of the output.

Besides, just because he's a PhD candidate, doesn't meant he is tech savvy or knows programming well. If given a small data set, he probably can do it manually.
*
Yes, it was quite the interesting engagement!

Unfortunately due to client confidentiality I can't discuss specifics, but basically he provided me with a dataset and asked me to create a series of models for several possible scenarios with different sensitivity analyses included.

In the deliverable, I ended up justifying the methodology that he'd used for compiling the data and speaking about the choice of the models that he wished to use in the context of performing his research, transforming variables, providing context and documentation as to why each thing I did had a logical basis to it, as well as providing an appendix that contained the output.

If it's a really small dataset you can probably do it in Excel, but if you want to do certain things like time series analysis or regression analysis with multiple variables (there were dozens), you might find that a little difficult, to say the least!

At the end of the day it was his data, and he had a sense of what he needed from the analysis, which was not difficult to pick out - The main issue was just that he didn't have the technical skill to perform the analysis because his school didn't provide coverage of the technical tools and didn't provide student access (Stata in particular is quite expensive), so it seems quite understandable that he didn't have that skill on hand and wanted to outsource, IMO.


Blofeld
post Nov 26 2017, 10:44 AM

Look at all my stars!!
*******
Senior Member
4,697 posts

Joined: Mar 2012
Universities always provide workshops for that.

I think the student is just lazy.
TSExoflare
post Nov 26 2017, 12:35 PM

Enthusiast
*****
Junior Member
869 posts

Joined: Nov 2017


QUOTE(Blofeld @ Nov 26 2017, 10:44 AM)
Universities always provide workshops for that.

I think the student is just lazy.
*
Haha! Well at any rate, it's been good work experience so far, and I'd like to continue with more clients, so I'll be happy to hear out anyone who needs to get work of this nature done ^^
zemega
post Nov 26 2017, 01:36 PM

Enthusiast
*****
Junior Member
723 posts

Joined: Jul 2008
QUOTE(Exoflare @ Nov 26 2017, 12:35 PM)
Haha! Well at any rate, it's been good work experience so far, and I'd like to continue with more clients, so I'll be happy to hear out anyone who needs to get work of this nature done ^^
*
Don't worry, you will have a lot of clients, you just have to know where to advertise. As research gets more complicated and involve more and more data just to make that necessary distinction, the more your service is needed.

Just to add, every semester there will be new clients.

This post has been edited by zemega: Nov 26 2017, 01:58 PM
TSExoflare
post Nov 29 2017, 07:28 AM

Enthusiast
*****
Junior Member
869 posts

Joined: Nov 2017


QUOTE(zemega @ Nov 26 2017, 01:36 PM)
Don't worry, you will have a lot of clients, you just have to know where to advertise. As research gets more complicated and involve more and more data just to make that necessary distinction, the more your service is needed.

Just to add, every semester there will be new clients.
*
Right, I think that's true! My services are typically called in when students have significant trouble with specific things that they can't conveniently resolve - Advertising sounds good! Just not sure where I'd start, but any help would be helpful, and I'm willing to offer commissions ^^
v1n0d
post Dec 27 2017, 03:38 PM

Another roof, another proof.
*******
Senior Member
3,197 posts

Joined: Mar 2007
From: Kuala Lumpur, Malaysia


There's definitely room for running boot camps on statistical software. The more packages you can teach, the better. Since you're from an economics background, I presume you're well-versed in both SPSS and E-Views?
TSExoflare
post Dec 27 2017, 11:52 PM

Enthusiast
*****
Junior Member
869 posts

Joined: Nov 2017


QUOTE(v1n0d @ Dec 27 2017, 03:38 PM)
There's definitely room for running boot camps on statistical software. The more packages you can teach, the better. Since you're from an economics background, I presume you're well-versed in both SPSS and E-Views?
*
I'm much more experienced with Stata than Eviews, but can teach both along with SPSS ^^ Let me know if that's something you're potentially interested in doing!
Mr.dupe
post Jan 5 2018, 07:06 PM

Getting Started
**
Junior Member
91 posts

Joined: Sep 2012
In no circumstances, a PhD student should have someone else to do his work. That will be considered plagiarism. But as V1nod suggested, he can attend bootcamp to help out with his work

 

Change to:
| Lo-Fi Version
0.0155sec    0.40    5 queries    GZIP Disabled
Time is now: 29th November 2025 - 06:34 PM