Discussion: View Thread

This message has been cross posted to the following eGroups: Statistical Consulting Section and Quality and Productivity Section .
-------------------------------------------
We recently formed a Community of Practice ("The Analytics") of slightly over 40 staff from all across UTMB, engaged in one form or another of data analysis - numerical, statistical, graphical, predictive, modeling / simulation / optimization.

A sub-group of us was asked by an executive-level business intelligence (BI) committee to design and execute a "use case study" to recommend software programs (systems?) to augment Oracle's BI Enterprise Edition...typical candidates could be QlikView, Tableau, Spotfire, Minitab, JMP, SPSS, etc.

We'd like to design and administer a survey to assess the functions and needs of "The Analytics" and I'm having a devil of a time coming up with anything that I think would be useful.  I've come up with four categories: 

1. Types of analyses (a long list of various and sundry data analysis methods)
2. Types of visualizations (another long list, this one of types of graphs)
3. Types of data (clinical, financial, operational, HR)
4. Sources of data (e.g., Excel, Access, EMR, Oracle databases, Clarity, PeopleSoft, several external databases, ...)

-------------------------------------------
Wayne Fischer
Statistician
University of Texas Medical Branch
-------------------------------------------

Hi Wayne,

It is a great idea.  Brainstorm around "types of questions asked"  ie "asked to analyze data from rapid change quality project", asked to analyze data from "pre-post research design for process change"  asked for dashboard of routine collected quality metrics:
I know this isn't exactly the right question, but might lead you in the right direction.


-------------------------------------------
Dorothy Syblik
-------------------------------------------

Perhaps I am wrong, but when I read over the questions that you are contemplating asking it seems to me that you might be falling into a classic trap, i.e., you are focusing on what you do as opposed to focusing on what your clients need.  In my experience, admittedly heavily in the commercial world, I find that my clients have questions that they need answers to such as "Which initiative should we emphasize for the 5 year plan?" or "How important are these different features when viewed from the eyes of our prospective customers?"  Rarely do they care how I get the answer, or how the data are presented to them -- they rely upon me to select the right techniques and modes of visualization that will do the job.  The ones who do care about the techniques usually are quite ill-informed and typically spout the latest in buzz-word bingo as they attempt to "look smart."  

So, with that as background, my first thought is that I'd talk face-to-face with several of my clients and prospects and ask them what problems they have that they need to have solved.  Some of those problems will be outside of your team's sweet-spot, but others will likely be ones that you can help them with.

-------------------------------------------
David Mangen
Owner
Mangen Research Associates, Inc.
-------------------------------------------







Original Message:
Sent: 11-12-2013 09:42
From: Dorothy Syblik
Subject: ? Survey to determine data analysts' needs... ?

Hi Wayne,

It is a great idea.  Brainstorm around "types of questions asked"  ie "asked to analyze data from rapid change quality project", asked to analyze data from "pre-post research design for process change"  asked for dashboard of routine collected quality metrics:
I know this isn't exactly the right question, but might lead you in the right direction.


-------------------------------------------
Dorothy Syblik
-------------------------------------------

Thanks, David.  [Pardon my delayed response - your posting got hung up in the clearance process...]

As part of our study, we will be developing and administering a *second* survey - this one of our internal clients...those who use the results of our analyses.

And we've added a section to "The Analytics" survey that asks about the types of questions asked / solutions needed by our clients. 

-------------------------------------------
Wayne Fischer
Statistician
University of Texas Medical Branch
-------------------------------------------







Original Message:
Sent: 11-12-2013 10:09
From: David Mangen
Subject: ? Survey to determine data analysts' needs... ?

Perhaps I am wrong, but when I read over the questions that you are contemplating asking it seems to me that you might be falling into a classic trap, i.e., you are focusing on what you do as opposed to focusing on what your clients need.  In my experience, admittedly heavily in the commercial world, I find that my clients have questions that they need answers to such as "Which initiative should we emphasize for the 5 year plan?" or "How important are these different features when viewed from the eyes of our prospective customers?"  Rarely do they care how I get the answer, or how the data are presented to them -- they rely upon me to select the right techniques and modes of visualization that will do the job.  The ones who do care about the techniques usually are quite ill-informed and typically spout the latest in buzz-word bingo as they attempt to "look smart."  

So, with that as background, my first thought is that I'd talk face-to-face with several of my clients and prospects and ask them what problems they have that they need to have solved.  Some of those problems will be outside of your team's sweet-spot, but others will likely be ones that you can help them with.

-------------------------------------------
David Mangen
Owner
Mangen Research Associates, Inc.
-------------------------------------------