Compartilhar via

((rating>=1) && (rating=<9)) || ((rating>=1) && (rating<=5)) ?

  • Artigo

We've been evaluating speaker performance, quality of sessions, satisfaction with events and almost every other measurable/metricable activity for the past few years on the One to Nine scale. And believe it or not, the Average bar is not 5.0 but 7.0 especially for Speaker ratings and overall satisfaction. 8.0 is good, 8.5 is Top 10. Obviously on a scale of 1-9, it's next to impossible to get a 9.0.

Recently i've been seeing a lot of other Microsoft events that started using the 1-5 scale. Those of you who attended MIX06 would know. I can see the merit of using the 1-5 scale especially when it is much simpler to understand. Although one might argue that a 1-9 scale provides more detail, I can safely say that is not true. How many people can't actually tell the difference between 3 and 4. or 6 and 7? usually its 1-2 for bad, 7-8 for acceptable, and 9 for impressive/excellent. With that in mind, wouldn't using a 1-5 scale be sufficient?

So here lies the rub,

How are we going to link the new 1-5 scale to the old 1-9 scale? you can't simply divide or multiply scores.

for example, a great speaker on the 1-5 scale might score 4.5. If you scale up by 2 and then multiply by 0.9, you'll get 8.1, where if you rate that same presentation on a 1-9 scale, the speaker might score 8.4 - 8.5.

so, what do you all think? 1-5 or 1-9?

Comments

  • Anonymous
    April 14, 2006
    Actually, I was taught to NEVER use an odd number of choices in a rating. An odd number allows an answer in the middle -- meaning the person didn't have to take a "side" ("no" answer).

    Always use an even number of choices in a rating, this forces the person to take a "side" on the issue.
  • Anonymous
    April 14, 2006
    Of interest: the five point scale is often used in questionnaires, to gauge someone's level of agreement. See http://en.wikipedia.org/wiki/Likert_scale
    "A typical test item in a Likert scale is a statement, the respondent is asked to indicate their degree of agreement with the statement. Traditionally a five-point scale is used, however many psychometricians advocate using a seven or nine point scale."