intelligent measurement

Likert scale & surveys – best practices – 2

November 27, 2011 at 2:07 pm 3 comments

i’ve written previously about the Likert scale and surveys – and received literally 100s of enquiries about it. A reader has now  pointed me towards this excellent article on survey questions and Likert scales that adds some interesting points to the discussion.

From my previous post, I listed the following best practices on using the Likert Scale in survey questions:

• More than seven points on a scale are too much.
• Numbered scales are difficult for people
• Labelled scales need to be as accurate as possible

And here are some further points to add drawn from this article:

• Be careful with the choice of words for labels:

“Occasionally” has been found to be very different than “seldom” but relatively close in meaning to “sometimes” (quote from article)

• Include a “don’t know” if for a point where people may simply not have an opinion:

“Providing a “don’t know” choice significantly reduced the number of meaningless responses.”

• People will respond more often to those items on the left hand side of the scale:

“There is evidence of a bias towards the left side of the scale”

On that last point, I always write my scales left to right – bad to good… This means that people may tend to select more easily the “bad” ratings. I haven’t found that to be the case (respondents often seem to be over-positive in their ratings I feel), but I stand corrected…

View the full article here>>

Entry filed under: Evaluation methodology, Evaluation tools (surveys, interviews..).