Monday, December 05, 2005

Types of data

I was struck on Saturday by a moment of confusion. We were talking about surveys and how they measure people's perceptions; are they then not qualitative in nature?

Nope.

Surveys with numeric responses provide quantitative data (you can add them up).

When I teach measures in my work, I talk about ways to sort measures. One was is sorting by hard or soft. Hard measures look at behaviors; soft measures look at perceptions. So, while a customer feedback survey provides quantative data, it is soft in nature. A measure which looks at customer retention, well, that's a hard data. It, too, is quantitative.

When I teach, I tell people they need to have both hard and soft measures. I think measures are, by definition quantitative. So, what I'm probably saying is that quantitative measures can be either hard (measuring behavior or results: retention rates or achievement test scores, for instance) or soft (measuring how students feel about school or why a new teacher says he quits).

We can have lots of different sorts of data:

  • Quantitative vs. Qualitative


  • Hard vs. Soft


  • Leading vs. Lagging


  • Producer-focused vs. Customer-focused


  • Activity vs. Results


  • Outputs vs. Outcomes

The more of the different kinds we have, the more complete a picture can we paint.