Beth Schultz

Ask & Ask Again: Questions Matter in Modeling

NO RATINGS
View Comments: Newest First | Oldest First | Threaded View
MDMconsult
User Rank
Prospector
Re: Asking the right questions
MDMconsult   11/30/2012 6:26:35 PM
NO RATINGS
@Lyndon_Henry I agree, results should accept and help to narrow ideas from being "too broad" Numerous issues can be identified through these surveys. Work with the most problematic issues such as security challenges, internal problems, etc al. It is important to be able to prioritize, act on survey results with needed improvements.

Lyndon_Henry
User Rank
Blogger
Re: Asking the right questions
Lyndon_Henry   11/29/2012 5:27:18 PM
NO RATINGS
..

Beth asks


...so are you suggesting surveying business users/managers about what business problem they're hoping the analytics will help them address? Or are you speaking more generally of surveys, for customer satisfaction, etc.?


 

Well, certainly, surveying the potential users (managers, administrators, planners, etc.) of the results would be a terrific idea.

But I had in general public surveys more in mind — I've had more experience with those.

In either case, you need to try to define the key issues as narrowly as possible.  Otherwise, I think you get survey results that are so broad and mushy they are close to useless (except lots of planners and dcisionmakers like to get results that are so broad and mushy they can interpret them in ways to bolster their own preferred goals).

And if this is supposed to feed into the development of a model, you could really end up with some junk and perhaps a mess.

 

BethSchultz
User Rank
Blogger
Re: Asking the right questions
BethSchultz   11/29/2012 3:22:40 PM
NO RATINGS
@Lyndon_Henry -- so are you suggesting surveying business users/managers about what business problem they're hoping the analytics will help them address? Or are you speaking more generally of surveys, for customer satisfaction, etc.?

Lyndon_Henry
User Rank
Blogger
Asking the right questions
Lyndon_Henry   11/28/2012 11:05:41 PM
NO RATINGS
..

Beth's article asks


Do you ask enough questions during your modeling process? 


 

Where I've found asking (formulating) the right questions to be important has been in surveys.  Not that they've all led to analytical models, but they have influenced decisonmaking, and some have fed into the modeling process.  Learning to ask the right questions is also helpful in learning how to identify the critical variables that are needed to make good decisions and build more effective models.

 

Pierre DeBois
User Rank
Blogger
Re: Answer the question, question the answers
Pierre DeBois   11/27/2012 1:25:02 PM
NO RATINGS
Callmebob, very true.  I like how this seems to really force introspection into the value of what is being modeled. Modeling data that may take too long of an arc to answer relative to the business need could benefit from this structure.

Callmebob
User Rank
Master Analyst
Answer the question, question the answers
Callmebob   11/27/2012 1:15:06 PM
NO RATINGS
Here's where I think that the data wizards need to have a bit of a marketing mind. In fact, they should work in consortium to create models that frame marketing's questions that answer their tasks to identify target markets, market research, product development, marketing mix, and monitoring results.

Information Resources
More Blogs from Beth Schultz
It's been a fun three years, but now it's time to say goodbye.
Take inspiration from Christopher Columbus as you pursue your analytical journeys.
The "big" in big data is no reflection of the size of the organization embracing its potential.
Whether you're an undergrad, a graduate student, or an analytics professional already, the same best-practices advice lives large.
Satellite data can help solve puzzles, from the lofty to the mundane.
Radio Show
A2 Conversations
UPCOMING
James M. Connolly
See How Data is Revolutionizing Healthcare


2/25/2016   REGISTER   0
ARCHIVE
James M. Connolly
The Analytics Job and Salary Outlook for 2016


1/28/2016  LISTEN   16
ARCHIVE
James M. Connolly
See How Analytics Drive Change in the Retail World


1/7/2016  LISTEN   103
ARCHIVE
James M. Connolly
All Analytics Conversations: Forecasts for Analytics in 2016


12/18/2015  LISTEN   3
ARCHIVE
James M. Connolly
Don't Make This Mistake With Big Data


12/11/2015  LISTEN   4
ARCHIVE
James M. Connolly
Understand the Difference Between Data Science and Analytics


11/23/2015  LISTEN   30
ARCHIVE
James M. Connolly
Shape the Next Generation of Data Scientists


10/20/2015  LISTEN   69
ARCHIVE
James M. Connolly
Analytics Best Practices: Plan for the Human Factor


10/14/2015  LISTEN   117
ARCHIVE
James M. Connolly
See How Your Analytics Initiative Measures Up


9/24/2015  LISTEN   50
ARCHIVE
James M. Connolly
Use Mobile Analytics to See the Big Picture


9/1/2015  LISTEN   83
ARCHIVE
James M. Connolly
Hire and Manage a Great Analytics Team


9/1/2015  LISTEN   152
Information Resources
Quick Poll
Quick Poll
About Us  |  Contact Us  |  Help  |  Register  |  Twitter  |  Facebook  |  RSS