« Global Warming - My Growing Sense of Unease | Home | The Doom and Gloom Speech on Consulting for a Living »

Say “No” To Kinky Data

By crooky | March 5, 2008

(Note: sorry about the late night posting. My wife and kids both have the flu and I’ve been playing nurse tonight. I look terrible in the outfit, FYI)

I was in a meeting with a client today and had an excellent opportunity to show off my research methodologies expertise. The client was talking about a focus group that I had proposed doing back in the Fall of 2007. The original purpose of the focus group was to get some qualitative feedback on a proposed curriculum for a new graduate degree. One of the people on my client’s project committee was concerned about price elasticity of the proposed program.
Kinky Data

Now, I’m no economist but I do know that to determine elasticity of the price of a good or service, you need a good sample of a lot of different people’s opinions on the data point or your data will get too “kinky”.

So, I jumped in and noted that a focus group is not the best way to collect that kind of data and suggested a re-survey of a sample group we already talked to but didn’t ask too many questions about price. They saw the logic and in the end, agreed with me.

My point: a little bit of research methodology knowledge goes a long way.

A little aside: When I was studying Economics as part of my grad school program, I remember the lecture on Kinked Demand well. The instructor was trying to explain that when your sample size is too small, you get “kinky” demand data and it is only when you get a sample of thousands of people that you see a nice, smooth curve of a normal demand function. People in the class weren’t getting it so I piped up and said “look guys… it’s like this. When only a few people are doing something, it’s kind of kinky. Once everyone starts doing it, it’s normal.”

Everyone got that question on the test right at the end of the semester and they all laughed quietly when they saw the question about kinked demand curves.

*********************
Aaron “Crooky” Cruikshank is the Principal and Founder of Friuch Consulting. He has written professionally about science and technology for ten years.

Share/Save/Bookmark

Topics: Research Methodologies |

Comments