Thursday, December 1, 2011

On A Scale from 1 to 10

I'm a big fan of quantitative data. I'm such a big fan that I frequently find myself figuring out ways to translate qualitative questions into ones that solicit quantitative answers.


I think it's because quantitative answers tend to be much more useful than qualitative answers.

What do I mean by "useful?"

The semantics (meaning) of qualitative answers tend to vary from person to person. In fact, they tend to vary from time to time with just one person.

For example, consider the qualitative answer: Great.

How's your day going? Great!

How was dinner last night? Great!

Hey, your mother-in-law just showed up at the reception desk and she says she needs to talk to you. Great!

Great is a great word, but it doesn't mean a heck of a lot; more accurately, it means so many things that it ends up meaning nothing. The same goes with terms like: OK, fine, confident, happy, interesting, nice, a lot, special... What in the world do these terms actually mean? Nada!

Would you describe yourself as confident or lacking confidence? Are you generally happy or unhappy? Did you find the presentation interesting? How'd you sleep last night? How much effort did you put into this assignment?

Who knows what an affirmative, negative or one-term response actually means? I don't.

My response to the ambiguity of qualitative answers is to translate my questions into ones that solicit quantitative responses. Rather than asking someone, "Would you consider yourself to be a confident person?", I do the following...

I'd like to ask you to assess your level of confidence. On a scale of 1 to 10 (10 being extremely confident and 1 lacking all confidence), how would you rate your confidence level?

Someone might respond, "Uhhh... I don't know. I mean, how can I put a number on that?", which pretty much answers the question; We're talking somewhere between 0 and 1.

Someone else might respond, "Oh, I'm definitely an 8, maybe a 9."

This is a great start. We now have a quantitative assessment of confidence level. However, the number is still questionable because we didn't calibrate the scale. What does 1 mean? What does 10 mean? To calibrate, I do the following...

OK, you're an 8. I'd like to calibrate your rating based on people you know. Who in your life would you consider to be the person with the least self confidence? (Again, the willingness to answer says a lot).

Um... I guess it would be my brother Ernie. Yeah, definitely Ernie. He's the king of second-guessing.

Great. We're going let Ernie represent a 1. Now, who's the most confident person you know?

Hmm... That would have to be my aunt Sue. She was in the Marine Corps and became a Master Sergeant. She's the most confident person I've ever met.

Cool. So, if your brother Ernie is a 1 and your aunt Sue is a 10, how would you rate your confidence level?

Uh, what did I say before, 8?

Or maybe 9.

Well, I gotta say that I'd be a lot closer to Ernie than to Sue. So, I guess that makes me more like a 5 or a 4.

Now we have a quantitative assessment that's been calibrated to the assessor's frame of reference. If the assessor's desire is to build confidence, we have a baseline measurement against which we can gauge progress. With two tangible models anchoring the end-points, it's easy to get back on the quantitative track if we ever drift into less useful qualitative analysis (e.g., I totally lack confidence or I am amazingly confident).

This method of assessment can be applied to pretty much anything that requires assessment of state or progress. You'd be amazed at how many qualitative assessments (attraction, goodness, happiness, satisfaction, desire, passion, effort, skill) can be transformed into quantitative ones and how enlightening it can be to come up with a calibrated number.

What would you like to quantify?

Happy Thursday,

1 comment:

  1. Talking with Jonathan this morning, I realize that I failed to clarify a significant point regarding my method of quantifying qualitative: The point of the exercise is not accuracy, it's perspective.

    If you spend more than a minute identifying your end-points and assigning a number, then you're probably over-thinking the question and leaning a bit to far into precision and accuracy.



Read, smile, think and post a message to let us know how this article inspired you...