Likert scale & surveys - best practices
November 20, 2007
I’ve been looking into the best practices for using the Likert scale type of question, probably the most widely used response scale featured in surveys - often used to measure attitudes and other factors (e.g. ”Excellent” to “Poor”). Created by Rensis Likert (pictured above) in the 1930s, his original scale featured five points. Over time, there has been many discussions and disagreements focused on one central question: What works best with the Likert scale to give you the most accurate responses?
I have read a number of studies on this question (sorry, I don’t link to them as they are all books or academic journals (that require a fee) but if you are interested write to me and I’ll give you the references) and the following are the points that most (but not all) scholars agree on:
More than seven points on a scale are too much. Studies show that people are not able to place their point of view on a scale greater than seven. So go for seven or less. What is the perfect number? Studies are not conclusive on this, most commonly mentioned are five, four or three point scales.
Numbered scales are difficult for people. For example, scales that are marked “1 to 5, with 5 being the highest” result in less accurate results than scales with labels such as “good” or “poor”. If numbered scales are used, signposts are recommended (e.g. put “poor” above 1, “satisfactory” above 3 and “excellent” above 5).
Labelled scales need to be as accurate as possible. Commonly uses labels such as “often” or “sometimes” often result in inaccurate responses. As these terms mean different notions of engagement from person to person, culture to culture (not to add the complexity of translating these terms). Scholars recommend using time-bound labels for frequency measures such as “once a week” (although problems of correct recall are also an issue). In addition, studies show that people find it difficult to differentiate between “very good” and “good” - better to use “good” and “excellent”.
And that’s it! Basically, there are inconclusive results on the use of a middle or neutral point (e.g. four point vs. a five point scale). Some scholars advocate a five point scale where respondents can have a “neutral” middle point whereas others prefer to “force” people to select a negative or positive position with a four point scale. In addition, the use of a “don’t know” option is inconclusive. I personally believe that a “don’t know” option is essential on some scales where people may simply not have an opinion. However, studies are inconclusive on if a “don’t know” option increases accuracy of responses.
Further information on the Likert Scale:
Examples of commonly-used Likert Scales >>
More examples of commonly-used Likert Scales >>
“Designing a useful Likert Scale” (pdf)>>
Glenn
Entry Filed under: PR Evaluation, Training Evaluation. .
19 Comments Add your own
Leave a Comment
Some HTML allowed:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>
Trackback this post | Subscribe to the comments via RSS Feed
1.
Michael Blowers | November 20, 2007 at 9:01 am
Hi Glenn, this is really interesting and ties in with some reseach I have playing with over the years regarding the spotting of tone for media coverage. I have always defaulted to the accepted groupings of ‘positive’, ‘neutral’ and ‘negative’, but have always been slightly uneasy allocating a neutral rating to a media item which just mentions a company’s name - this must have a value? A post on my blog regarding this is at: http://mediaevaluation.blogspot.com/
All the best, Michael
2.
Glenn | November 20, 2007 at 9:17 am
Hi Michael, I see what you mean. After reading your post, it did make me think about the concept of neutral rating of a media item - actually it’s so commonly accepted that no one’s challenged it before! But it goes back to two points: is there ever really a neutral point on an issue and is the mere mention of a company in terms of visibility a “positive” result in itself?
Requires more thought…
Glenn
3.
The magical number seven,&hellip | November 27, 2007 at 7:28 am
[...] an earlier post on best practices for likert scale questions, I made reference to an article “The Magical Number [...]
4.
anila bardai | December 3, 2007 at 5:22 am
please send article or formate of questionier on likert scale
5.
Glenn | December 3, 2007 at 7:53 am
Hello,
Please consult the links at the end of my post - clicking on these links will lead you to plenty of articles and examples of the likert scale.
Glenn
6.
writing open-ended questi&hellip | December 11, 2007 at 7:23 am
[...] previously written about best practices for using likert scale questions in surveys, I’d like to say something in favour of using open-ended questions. An open-ended [...]
7.
Likert scales, frequency &hellip | March 11, 2008 at 9:36 am
[...] “never”. etc. However, these scales often provide inaccurate responses as I’ve written about before. Why is that so? Well, describing frequency differs enormously from person to person. This [...]
8.
Elham | March 12, 2008 at 6:45 pm
please send more examples of three point scale and the adavantage & disadvantage of 3 or 5 or 7 scales
9.
dian | March 19, 2008 at 8:19 am
hi glen, in my thesis i use four point scales, could you tell me what is the name: “likert with four point scale”. .. i had read many books but so far i cant find the name yet. would you help me..
thank you.
10.
Glenn | March 20, 2008 at 1:51 pm
Hi Dian,
For a likert scale with four scales, there is no given name as far as I know. I’ve seen the likert scale vary from 2 to 10 points on a scale. But there is no particular name for for a four point likert scale - it’s a four point scale without a central point (an even number).
Glenn
p.s. Elham, for more examples of Likert scale, see the links at the end of my post above.
11.
Harry | March 23, 2008 at 1:49 am
Hi Glen,
The original scholars used 5 point likert scale but my boss proposed that I should use 7 point likert scale so that this appears consistent with other variables which I intend to measure. But at the moment still can’t find any back up for this. Do you have one?
Many thanks
Harry
12.
Zoe | March 29, 2008 at 5:26 pm
Your page on Likert scaling is very useful. I wonder if you have any information regarding Likert scaling for questions to be answered by primary aged children. I need some scales that they would understand. Thanks
13.
Glenn | March 31, 2008 at 7:25 pm
Hello Zoe,
That’s a good question. There are some studies that exist on using likert scales with children, notably the following one:
“Adaptation of Likert Scaling for Use with Children”.
Occupational Therapy Journal of Research, v5 n1 p59-69 Jan 1985
http://www.eric.ed.gov/ERICWebPortal/custom/portlets/recordDetails/detailmini.jsp?_nfpb=true&_&ERICExtSearch_SearchValue_0=EJ317880&ERICExtSearch_SearchType_0=no&accno=EJ317880
However, I have not read this study so I am not familiar with its recommendations.
From what I know, I would say likert scales for children should have a limited number of points on the scale (between 3 - 5) and use simple terms (good, bad, etc.) or graphics (smiley/sad faces) above each point. Numbered scales would work less well in my opinion and some studies recommend using a visual analogue scale (but not for children under 7). A visual analogue scale (used for health issues often) is where you ask the child to put a mark on a line with two end points (see this example, once you see it - all will be clear!
http://www.blackwellpublishing.com/specialarticles/jcn_10_706.pdf
And of course, it all depends on how much guidance you give a child - I guess in most cases you need to guide a young child through the questions.
I hope this information is of use,
Kind regards,
Glenn
14.
dian | April 4, 2008 at 9:00 am
hi Glen.
thanks a lot for your answer, it’s very helping me.
I have another problem, i’m sure that you will help me again. I face some trouble with making questioner used likert scale answer. My questions in the questioner bias with yes/no questions although i had try hard to make statements but the result are still similar with yes/no questions.
Could you give some tips for changing question into statement in questioner with ikert scale measurement?
thank you very much.
15.
Allen | April 11, 2008 at 11:17 pm
http://ourworld.compuserve.com/homepages/jsuebersax/likert.htm
A good article by Dr. Uebersax about the confusion around the concept of a “Likert scale”.
Have a nice day,
Allen
16.
Glenn | April 13, 2008 at 1:01 pm
Thank you for that Allen. It does look an interesting article with some important points - I will make reference to it in a future post.
Kind regards
Glenn
17.
Glenn | April 13, 2008 at 1:02 pm
HI Dian,
That’s a good one also! let me think about that I’ll get back to you…
Glenn
18.
Glenn | April 13, 2008 at 1:26 pm
Dian, further to your question, actually some questions are more suited to a yes/no response and others are more suited to a Likert scale. For example, “have you ever done XY?” is suitable for yes/no. but “how favourable are you for XY?” is more suitable for likert style format. The likert scale is more for where you can imagine responses on a scale of frequency/intensity. For some questions this may not be appropriate.
I think if you are working with questioners doing the surveying for you, then you just have to guide them further with the response model written on your survey.
Glenn
19.
Glenn | April 13, 2008 at 1:37 pm
Dear Harry,
Apologies, I just saw your comment now. Actually most studies are not conclusive on the difference between 5 and 7 points on a scale. In other words, both 5 & 7 point would provide you with accurate and reliable responses. In the article listed below, they conclude that as you add more points, a scale becomes more reliable - but only up to a certain point (higher than 11 is too much for most people). Consequently, they argue that 7 is slightly more reliable than 5.
I would recommend that you use a 7 point scale, particularly if you have other comparative variables that you are going to measure on a 7 point scale - it makes it much easier to compare and avoids having to re-calculate your results.
The article I mentioned:
Alwin, D & Krosnick, J, “The reliability of survey attitude measurement: The influence of questions and respondent attributes”, Sociological Methods Research, 1991; 20; 139
http://smr.sagepub.com/cgi/content/abstract/20/1/139
Kind regards,
Glenn