Likert scale & surveys – best practices

November 20, 2007 at 8:45 am 222 comments

Rensis Likert himself

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 >>

“The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information”>>

Glenn

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

The value of checklists and evaluation: 7 reasons The magical number seven, plus or minus two

222 Comments Add your own

  • 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

    Reply
  • 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

    Reply
  • [...] an earlier post on best practices for likert scale questions, I made reference to an article “The Magical Number [...]

    Reply
  • 4. anila bardai  |  December 3, 2007 at 5:22 am

    please send article or formate of questionier on likert scale

    Reply
    • 5. aklilu  |  August 19, 2012 at 8:02 am

      please send example

      Reply
  • 6. 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

    Reply
    • 7. aklilu  |  August 19, 2012 at 8:04 am

      I need additional example of article to classify knowledge in to good and poor based on mean score

      Reply
  • 8. writing open-ended questions « intelligent measurement  |  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 [...]

    Reply
  • [...] “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 [...]

    Reply
  • 10. 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

    Reply
  • 11. 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.

    Reply
  • 12. 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.

    Reply
  • 13. 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

    Reply
  • 14. 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

    Reply
  • 15. 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

    Reply
  • 16. 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.

    Reply
  • 17. 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

    Reply
  • 18. 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

    Reply
  • 19. 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

    Reply
  • 20. 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

    Reply
  • 21. 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

    Reply
  • 22. Kisha  |  May 19, 2008 at 8:56 pm

    I am completing my dissertation. I was told I would have better success if I found a survey comprable to the one I am attempting to create. Does anyone know of a site that provides a listing of various Likert scale surveys?

    Reply
  • 23. Glenn  |  May 20, 2008 at 1:13 pm

    Hello Kisha,

    I am not sure you will be able to find a survey that precisely matches the one of your dissertation, but you can certainly check the scales you use with scales used by others:

    http://www.gifted.uconn.edu/siegle/research/Instrument%20Reliability%20and%20Validity/Likert.html

    http://dataguru.org/ref/survey/responseoptions.asp

    Glenn

    Reply
  • 24. rachana  |  May 30, 2008 at 5:36 am

    Hi ,your article help me a lot in my market research first step. i need your immediate help for my research.i want to learn all the interpretation of factor and cluster analysis.this is my humble request to you.and i hope you will definitely help me.
    Thanks.

    Reply
  • 25. Ruby  |  June 6, 2008 at 12:00 pm

    Could you please help me out with the advantages of using a four point likert scale as opposed to using the five point one.

    Reply
    • 26. Cosmos Dzikunu  |  December 9, 2009 at 2:24 pm

      Please what are the advantages of four point Likert Type scale of Five point

      Reply
  • 27. donny  |  June 28, 2008 at 9:47 am

    hi glenn,i’m doing my final assignment now about service quality in a retail store. i’m little bit confuse about using likert scale. I’m confuse whether to use 5 or 7 Likert scale.
    Can you mention some advantages and disadvantages of both 5 and 7 likert scale?
    thx before..

    Reply
  • 28. John Dawes  |  June 30, 2008 at 7:10 am

    I recently published a study that used a ‘split sample’ experiment to see if the number of scale points has any effect on survey data. I split respondents into three groups, one answered using a 5-point scale, another group used a 7-point scale and the third, answered using a 10-point scale. I “re-scaled” the data using a simple bit of arithmetic, and the three scales produced almost identical results in terms of mean scores and variation about the mean. If anyone would like to look at a draft copy of the study, and does not have access to the journal, you can email me. Otherwise, the reference is Dawes, John. “Do Data Characteristics Change According to the Number of Scale Points Used ?” International Journal of Market Research Vol 50, No.1, 2008.

    Reply
    • 29. Josh  |  June 17, 2010 at 3:40 pm

      Hi! John, will be grateful if you can send me your published study. I am planning to do a survey on the satisfaction of the staff on the service provided by our administrative srvices department. Am trying to undersatnd the differences in using the 5, 7 or 10 poni scale. Any views on using a 6 point scale? What are comparative advantages or disadvantages in using these odd points or an even points scale?

      Reply
  • 30. Glenn  |  June 30, 2008 at 8:31 am

    Many thanks for those comments John & Donny. From what I’ve read, studies show there is little difference in reliability in using 5 or 7 point scales – studies on the whole tend to favour 5 over 7 – but not all agree. Most agree that more than 7 point scales are difficult for respondents to handle. Personally, I use 5 point scales often in surveys as I find 3 points too little and 7 points too many.
    Glenn

    Reply
  • 31. Asante bright Owusu  |  July 14, 2008 at 3:34 pm

    could U please send me more examples of the allpication of the likert scale ?, I am on my Mpil thesis and I would like to use it to achieve one of my objectives.
    thank you

    Reply
  • 32. Naveen  |  July 21, 2008 at 4:51 pm

    Dear Glenn,
    Thanks a lot for your time and kind information on measuremnts. I have a doubt too. My present research questionnaire is on the the following
    1.Existing woring conditions of employees in XXX Spinning Mill – Satisfactory or Unsatisfactory.
    2.Positive or negative change in lives of employees after employment in the Spinning Industry.

    My research covers both those questions and I plan to use questionnaires. My doubt is can I ask questions regarding both those areas in one questionnaire with the first half having Yes/No Questions and the second half having Linkert Scale.

    If my questionnaire uses both the scales, i.e. Yes/No Questions and Linkert Scale, will I still be able to do analysis such as Chi-Square etc.

    Thanks a lot in advance.

    Reply
  • 34. Glenn  |  July 23, 2008 at 6:56 am

    Dear Naveen,
    The use of different types of scales in surveys is recommended (such as likert scale and yes/no – a binary scale).This use of both scales shouldn’t be a problem to undertake such analysis as Chi-square.

    Glenn

    Reply
  • 35. ong  |  July 27, 2008 at 6:54 am

    Dear Glenn,
    I am in the process of writing my chapter 3 thesis, I don’t how to write why I will used 5 point Likert scale. my study is regarding the student and teacher perception on the effectivenss of clinical teaching. could kindly give some suggestion TQ

    Ong

    Reply
    • 36. Gillian  |  May 2, 2009 at 8:53 pm

      I am interested in the references used for this Likert article.
      Thank you in advance!
      Gillian

      Reply
  • 37. Glenn  |  July 30, 2008 at 8:32 pm

    Dear Ong,

    Simply put, although still debated, studies show that scales of 5 -7 are the most accurate (as compared to smaller or larger scales). You can read more in this article:

    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

    Reply
  • 38. Cultural issues in evaluation « intelligent measurement  |  August 1, 2008 at 3:36 pm

    [...] and usefulness of the evaluation and to be careful in the use of questionnaire types (such as the Likert scale) which may be misunderstood in some [...]

    Reply
  • 39. lina  |  August 4, 2008 at 10:29 am

    Dear Glenn

    Your article and responses were very informative. Could you kindly clarify a doubt i have. I am doing a comparitive case study using questionnaire that has likert scale questions. How could i analyze the likert scale questions.

    Another doubt is regarding sampling- could i generalize a dispropotionate stratified sampling to the population? What is weight?

    I would be much grateful if you could answer my questions.

    Leena

    Reply
  • 40. musonge mpah  |  September 2, 2008 at 11:16 am

    please i need to know the importance of linkert,and guthman scales used in research methods

    Reply
  • 41. Aziz  |  September 12, 2008 at 9:03 am

    From a marketing perspective, is there any literature to suggest the benefit of a 5 point as opposed to a 4 point scale. The assumption I have is that if you’d like to win over a group, it will be those in neutral that are more likely to be won over. Do you have any study to back this assumption?

    Reply
  • 42. Glenn  |  September 18, 2008 at 2:34 pm

    Dear Aziz, I have not read anything that supports the argument that you put forward – there is a difference of opinion as to whether 4 point or a 5 point scale works better.
    Glenn

    Reply
  • 43. Glenn  |  September 18, 2008 at 2:36 pm

    Dear Musonge, please see above for information on the Likert scale. A quick search on the web will bring up many resources and explanations about the Guthman scale.
    Glenn

    Reply
  • 44. Glenn  |  September 18, 2008 at 2:39 pm

    Dear Leena,

    Apologies, I somehow missed seeing your comment on my blog.

    To compare Likert scales are very easy – but they must be “comparable” – for example using the same scale, e.g. poor, good, excellent. etc. If not it is difficult to compare. For the weighting, I’m really not the right person to ask – sorry!
    Glenn

    Reply
  • 45. Stuart  |  September 23, 2008 at 2:07 am

    Can you comment on using a mixture of 5pt and 7pt scales in the same survey? Any concerns…

    Thanks

    Reply
  • 46. Glenn  |  September 23, 2008 at 12:48 pm

    Dear Stuart, the main issue with using 5 and 7pt scales is if the questions using these scales need to be compared. If they don’t then it’s not a problem. If they do it is possible but more of a hassle as you have to make some calculations to make them comparable and it doesn’t allow to make a quick comparison. I would generally recommend using the same number of points in a survey (but not necessarily the same response options) – although that’s sometimes not possible.
    Glenn

    Reply
  • 47. Preeya  |  October 7, 2008 at 10:08 am

    Hi Glen,

    You’ve said above, ‘However, studies are inconclusive on if a “don’t know” option increases accuracy of responses.’

    Can you please provide me with some references I can look up to investigate this further?

    many thanks

    Reply
  • 48. Glenn  |  October 9, 2008 at 1:28 pm

    Hello Preeya,

    This is discussed in the following article:

    1) 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

    Reply
  • 49. Toyin Taiwo  |  October 21, 2008 at 6:57 am

    I am writing up my thesis and my supervisor requested I state the reason for using a 4-Likert scale. I find the article and comments posted here very useful and educative.

    Reply
  • [...] written before about survey respones and the use of “don’t know” as an option on a Likert scale. [...]

    Reply
  • 51. ambrina  |  November 12, 2008 at 9:11 pm

    Hi thanks fot the very helpful information, if possible could you please give me references. Thanks again
    Ambrina -x-

    Reply
  • 52. Glenn  |  November 13, 2008 at 9:27 am

    Thank you Ambrina, I’ve just sent them to you.

    Glenn

    Reply
  • 53. arnie manlapas  |  December 3, 2008 at 1:24 am

    i would want to know your references for all the information you have cited above. i’ll be needing them for my thesis. if you got more about Thurstone Scaling. please send me the list of references or i would appreciate more if you could send me copies of research journal articles about them. thanks

    Reply
  • 54. Joan  |  December 4, 2008 at 4:02 am

    Could you pls give me the references about the Likert scale for all the information that you’ve cited above?
    Thanks for sharing the valuable information.

    Reply
  • 55. Glenn  |  December 4, 2008 at 9:08 pm

    Thank you Joan & Arnie, I’ve sent you the references.
    Glenn

    Reply
  • 56. Michelle  |  January 23, 2009 at 3:01 pm

    This is very useful information! Many thanks! It’s possible you could send me the references as well? I need them for my thesis…
    Michelle

    Reply
  • 57. Dan B  |  January 27, 2009 at 4:02 pm

    Hi Glenn,
    thank you for a very informative and useful piece. I just wondered if you had ever found any information or knew anything regarding the order of the scale – should the scale be positive to negative left to right or the reverse? Is there any evidence advocating the use of one approach over the other? Anybody?
    Thanks,
    Dan B

    Reply
  • 58. Glenn  |  February 2, 2009 at 7:51 pm

    thanks Michelle, I’ve sent them to your email.

    Hi Dan, That’s a question I’ve often asked myself – but I can’t find any *hard* academic reference on it (although I do remember reading something somewhere supporting negative to postive (left to right) but I can’t recall where it was).

    I’ve written about this earlier to another reader, I quote:

    “I believe it’s better to go from the negative to the positive, left to right. It seems more logic to me and some automated survey software mark your answers and calculate the responses for graphs on this basis, e.g. that the first point is the lowest. But I’ve had clients argue that it should be the opposite way around – put positive to negative, left to right – as people will click on the first point by default – which I personally don’t believe. I’ve never found any academic reference supporting either way but looking at all examples in academic articles, 95% are written as negative to positive, left to right.”

    Reply
  • 59. Shynnette  |  February 3, 2009 at 7:15 am

    Hi Glenn
    I am doing a research and my adviser told me to use Likert scal but I don’t know how to analyze the data using the scale. it’s about effectiveness. hope you can help give directions to my study. thanks in advance.

    Reply
  • 60. jewel  |  March 10, 2009 at 12:47 am

    i’m working on my thesis and i have to compare data from 3 point and 6 point likert type scales. it was suggested that i recode the 6 point to 3 point does anyone have any idea of how to do this and what research supports this? thanks

    Reply
  • 61. Glenn  |  March 11, 2009 at 8:14 am

    Hello Jewel, That doesn’t sound very difficult as all you have to do is halve the data of the 6 point (e.g. 6 becomes 3, 5 becomes 2.5 ,etc). Only issue will be that you will have half points that would not exist in a three point scale – putting in question the comparable nature of the data. You may want to ask a statistics expert their opinion.

    Reply
  • 62. David McBride  |  March 14, 2009 at 11:25 pm

    Hi Glenn, great site, I happened upon it when looking at research methods for my MBA dissertation. Intuitively I wanted to use a six point scale to force people to adopt a position, even if that is a mild one. My theory is that everyone has an unspoken opinion, however mild on matters of importance to them. The uncomfortable feeling of having to make a positive or negative statement makes people consider their responses more carefully. The advantage of having a central starting point of 0 with +VE and -VE values is considerable. Social desirabilty bias must also be factored in as individuals are influenced by the desired norms of their social, cultural and political groupings.

    Your blog seems to focus on 5, 7 or even 3 point scales but 4 and in my view a 6 point scale is worth a look.

    Is there any good academic literature that you can point me to?

    Thanks for the great website

    Reply
  • 63. Glenn  |  March 18, 2009 at 1:25 pm

    Hi David, thanks for the compliments. As you state I favour more the 3 or 5 point approach in order to produce a middle point. But as you propose – to select an even number – 6 – to force people to adopt an opinion – is another point of view on the subject of attitudes and scales. There are some, like myself that believe that people simply may not have an opinion on certain issues and shouldn’t be forced to choose. Yet there are others that argue that people do have *hidden* opinions on everything..

    This is probably the best academic article I’ve read to date which sums up these issues and will lead you to other resources in this area:
    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

    Glenn

    Reply
  • 64. Steve  |  April 7, 2009 at 10:22 pm

    Hi Glenn, i am looking at the weaknesses of using a 5 point likert scale in a school survey. Basically, i need as much info against likert scales as possible! Also, any references you could could provide would be great!

    Thanks

    Steve

    Reply
  • 65. bilal  |  April 13, 2009 at 10:14 am

    Hi Glenn, i am working on my MS thesis using a survey. I have used five point likert scale. Now i am confused how to treat the data. it will be treated as Ordinal or interval. As different people have different opinions in this regard. If you can please guide me or give me any refrence it will be very helpful for me.
    Thanks
    Bilal

    Reply
  • 66. genevacom  |  April 15, 2009 at 7:18 am

    HI Steve, That’s a good question, the main disadvantage of the Likert scale is matching the points of the scale to the feelings of the people responding – the problem is that people sometimes are not able to accurately place their sentiments on a scale at a particular point. But then again, a scale is better than a simple yes/no (you have more choice).

    Reply
  • 67. Glenn  |  April 16, 2009 at 6:47 pm

    Hi Bilal,

    That really depends upon what is in your scales – as they can be either interval OR ordinal.

    Ordinal scales are where the order matters but not the difference between values. Typically scales of satisfaction, preference or agreement. For example, you might ask people to rate their level of happiness on a scale of 1 to 10. A score of 7 means more happiness than a score of 5, and that is more than a score of 3. But the difference between the 7 and the 5 may not be the same as that between 5 and 3. The values simply express an order.

    Interval scales are where the order matters but the difference is meaningful in “real terms”. Typically scales of salary, temperature or age. Here if someone is paid $50,000 and someone $45,000, that’s a “real” difference of $5000.

    So how you analyse the data and what techniques you use depends upon the above differences.
    Glenn

    Reply
  • 68. Lickertskalor och datanivåer « Contextually  |  May 6, 2009 at 12:37 pm

    [...] bra och seriös bloggpost finns på Likert scale & surveys – best practices Vi har stöd i vårt resonemang om att mer än 7 är för mycket, att det är svårt med siffror [...]

    Reply
  • 69. Mark  |  May 12, 2009 at 9:47 am

    Glenn

    I would also appreciate a list of references.

    Thanks
    Mark

    Reply
    • 70. Glenn  |  May 12, 2009 at 5:51 pm

      Thank you Gillian and Mark, I have sent you the references.
      Glenn

      Reply
  • 71. Daily Links « Jason’s Blog  |  May 15, 2009 at 3:03 am

    [...] Likert scale & surveys – best practices From intelligent measurment blog [...]

    Reply
  • 72. ida damayanti  |  May 27, 2009 at 5:05 am

    Hi

    I need to finish my magister psycology research on student career choices of Holland hexagon personality theory. I faind verry difficult to make an item instrument in Likert Scale. I would be appreciate if you could help me to send an coppy of that item pleasse.

    cheers
    Ida

    Reply
  • 73. ida damayanti  |  May 27, 2009 at 5:20 am

    hi
    I need to finish my magister psychology research on student career choice. I need to make an item instrumen of Likert scale to do this research. I would be apreciate if you could help me to do this please

    cheers
    ida

    Reply
  • 74. Sam  |  July 10, 2009 at 1:28 am

    Hi Glenn

    I am developing a questionnaire for children in hospital and is planning to use Likert scale. I read in this site that for children, the number of anchoring points should be limited. Can you direct me to a reference to support this view?

    Thanks. Sam

    Reply
  • 75. mona  |  July 15, 2009 at 1:09 am

    Dear Glenn,
    I would like to ask if you know any Likert Scale regarding Sexual Behavior of Adolescents?

    Reply
  • 76. saleh  |  July 26, 2009 at 7:48 pm

    Hi Glenn

    I would also appreciate a list of references.

    Many Thanks
    Saleh

    Reply
  • 77. dawn  |  July 30, 2009 at 9:44 pm

    Hi Glenn,

    I’ve noticed that you’ve stated in responses that you don’t know much about weighting and such, but do you happen to know of any articles or websites that may help? I recently completed a survey for my dissertation which used questions and scales that had already been tested (and thus validated), and the scales used varied (5-point, 7-point). However, we need to compare results based on these scales and I’m looking to find out more about how to do so. If you have any information that may help, it would be greatly appreciated, but I understand if you don’t!

    Many thanks,
    Dawn

    Reply
  • 78. Jon  |  August 17, 2009 at 8:17 pm

    Glenn –
    I think I may be asking the same question as Dawn. Do you know (or do you know of any references that could explain) how researchers generally assess a “significant” difference in Likert scale scores? We are doing a study to look at differences in satisfaction scores for patients in a hospital – some of whom receive an intervention and some of whom are “control” patients receiving standard of care. I am trying to calculate a goal for the number of patients to enroll but I can only get that by deciding what kind of difference I will consider “clinically significant”. Any ideas about where to look?
    Thanks,
    Jon

    Reply
    • 79. Glenn  |  September 17, 2009 at 6:29 am

      hi Jon, if I understand what you are asking, it is a good question – when is “satifisfaction” at an acceptable level – 80%, 70% ? I know that the standards quality people SGS (www.sgs.com) consider that objectives are not met when satisfaction falls beneath 70% – in other words, 30% on the negative side of a Likert scale.
      Glenn

      Reply
  • 80. sam  |  September 27, 2009 at 5:35 am

    Your article on the likert scale is the answer to my prayer…thank you. I would also like the list of references. TQ

    Reply
    • 81. Glenn  |  September 30, 2009 at 8:07 pm

      Thanks Sam, I’ve sent you the references by email.
      Glenn

      Reply
  • 82. asrab  |  September 29, 2009 at 8:46 am

    Hi Glenn,
    Thank you for your informative article. I would like to know what the importance of the central point in Likert scale.
    I would also appreciate a list of references.

    Reply
    • 83. Glenn  |  September 30, 2009 at 8:12 pm

      Hello Asrab,

      I’ve sent you the references by email. The central point of the Likert Scale is a hotly debated issue.

      On one hand, some argue that the central point is essential as it allows people to have a neutral position when asked to rate and issue, as some people may not have an opinion on an issue.

      On the other hand, others argue that a Likert Scale should not have a central point, as people should have an opinion on all issues – or be forced to take an opinion. Otherwise, too many prefer to take the neutral middle point that does not help to determine tendencies.

      Glenn

      Reply
      • 84. Caroline  |  November 24, 2009 at 8:38 pm

        Hi,
        I hate to ask and be rude but i was wondering if you could send me the references for this aswell as I’m really atruggling to find many relevant articles on the databases,
        thanks for your time,
        Caroline

      • 85. Glenn  |  November 25, 2009 at 8:50 am

        Hi Caroline, no problem, I’ve sent them to you by email.
        Glenn

  • 86. Shrly  |  October 2, 2009 at 8:34 am

    Hi hi,

    This is a very good discussion.
    Could you kindly send me the references as well?

    I have a question. If I have historical data on a 4pt scale and new data on a 5pt scale in a longitudinal study, and I want to make comparison with the newly collated data with the historicals, any suggestions as to how best to do this? Any pros and cons in terms of the approaches? Thanks.

    Shrly

    Reply
  • 87. Likert scales « assessCME  |  October 5, 2009 at 10:41 am

    [...] you create your next Likert scale question, check out this post.  How big should your scale be?  How should it be labeled?  Turns out there’s a science [...]

    Reply
  • 88. asrab  |  October 7, 2009 at 6:29 pm

    Hi,

    Thank you for sending the references

    I have diffeculty in reporting the results of Likert scale.

    I used 4-point scale in grammatical judgment task: 1 ( completely unacceptable), 2 (unacceptable), 3 (acceptable), 4 (completely acceptable)

    1 and 2 indicte rejection, 3 and 4 indicate acceptance. when analyzing data mean scores below 2 indicate rejection and means above 2 indicate acceptance. is this way correct because I read in some reseaches that used 5-point scale that the mean scores above 3 were reported as acceptance and means below 3 were rejection.

    I would be much grateful if you could answer my question.

    Reply
  • 89. Silvia  |  October 14, 2009 at 8:02 pm

    Olá, Glenn
    Eu estou a realizar o meu prjecto e tenho uma duvida que gostaria de perguntar.No meu questionario é usada uma escala de Likert de 4 pontos que vai avaliar a frequencia dos sintomas vivenciados pelas pessoas. No entanto, não sei qual sera a tradução ingles-portugues da seguinte escala: 1:Yes, definitely 2: Yes, sometimes 3: No, nor much 4: No, not at all
    Opções possiveis: 1: sim, sem duvida 2:sim, por vezes 3: não, raramente 4: não, de modo nenhum
    Ou: 1: sim, sempre 2: sim, por vezes 3: não, raramente 4: não, nunca
    Se me pudesse ajudar nesta traduçao para portugues agradecia imenso e se tiver alguma referncia bibliografica que me possa indicar seria muito bom mesmo.
    Fico aguardar a sua opinao.
    Com os melhores cumprimentos Silvia

    Reply
  • [...] with Gellis Communications and in our discussions the use of Likert Scale in surveys. As I’ve written about before, the Likert scale (named after its creator pictured above) is widely used response scale in [...]

    Reply
  • 91. Zhao  |  November 5, 2009 at 10:27 pm

    hi Glenn,

    I designed a questionnaire with five-point likert scale. Each question is related to design statement. I want to identify which design elements is potential problems based on the participants’ perception. Could you tell me which analysis approach is right to select questions as serious problems from the questionnaire? I try to use Frequencies table, however, sometime it is really hard to decide whether the question is a problems. for example, 26% agree; 50% neutral, 24% disagree

    Reply
    • 92. Glenn  |  November 9, 2009 at 9:36 pm

      Dear Zhao,

      There are many options. From my experience, what can be done is to examine and compare the percentages of each result per point on the scale, or alternatively group negative (point 1,2)/positive (point 4,5)/neutral (point 3) points to see if any patterns emerge. Another option is if you have several items measuring the same design feature you can use the Cronbach’s Alpha test of reliability. In other words, by increasing the number of questions on the same item, it often increases the reliability of the results.
      Glenn

      Reply
  • 93. foibles  |  December 5, 2009 at 7:31 pm

    glenn, thanks for running down the subtleties of the Likert scale. Each Q-type has it’s own kinds of inherent issues. I’ve been looking around and Not every online survey tool handles multiple styles of questions. Zoomerang does. The text import feature that auto-configures questions is pretty handy too.
    http://snurl.com/zoomtypes

    Reply
  • 94. maria  |  January 13, 2010 at 10:40 am

    I am really enjoying your blog! Thank you very much for the effort!
    Could I also have the references?

    Also, regarding the answers, your say “I’ve never found any academic reference supporting either way but looking at all examples in academic articles, 95% are written as negative to positive, left to right.” Is this an aproximization? I agree with you but my superviser does not, I would like to be able to argue this point with references, if possible.

    Thanks again,
    Maria

    Reply
  • 95. Willy  |  January 26, 2010 at 9:11 pm

    hi glen,
    I’m currently in the process of doing my thesis proposal on
    issues and concern on Motorlaunch Operators. Can you suggest
    what instrument will I used. Can you sent me a sample if
    you have one. Thanks and GOD Speed. Willy

    Reply
  • 96. Brandie  |  February 3, 2010 at 5:38 pm

    Hi Glen,

    I’m gathering information regarding the minimum number of response levels for a Likert scale. Could you please send me the references for your statements above?

    Reply
  • 97. Colin  |  February 15, 2010 at 10:53 am

    Glenn

    Interesting comments on the Likert scale – could you also send me the references you used?

    Many thanks

    Reply
  • 98. khaled Hamuda  |  February 18, 2010 at 6:33 pm

    could you send to me the advantages of qoestionnaire of 6 points scales, and any refferences talk about this kind of questionnaire

    many thanks

    Reply
  • 99. Edward  |  March 7, 2010 at 12:49 am

    Hi glen

    could you send any information preferably journal articles on using worded scales over numbers e.g.

    1 2 3 4

    or

    agree disagree

    and which method is better, i need this infomation for my dissertation thanks

    Reply
  • 100. Glenn  |  March 7, 2010 at 6:34 am

    Thank you Edward, that point on numbers vs. labels is covered in one of the artlces that I just sent you

    Glenn

    Reply
  • 101. Judith  |  March 9, 2010 at 11:57 am

    Hi Glen,

    I would like to compare whether a 6-point likert scale differs from a 9-point likert scale for the same type of questionnaire. The questionnaire is originally developed to measure values on a 9-point scale (validated), but we also used the same value-instrument but only on a 6-point likert scale (not validated). We would like to know what kind of weighing procedure (if any) is needed to use the 2 measures in further analyses. Maybe you have example of articles about this?

    Thank you,
    Judith

    Reply
    • 102. Glenn  |  April 16, 2010 at 7:06 pm

      hello Judith, i think you’d be better to ask a statitician. My feeling would be that the by transforming the nine points to six points you will get more accurate results (as nine point scale is too broad) but by doing so I believe you would have to reconsider if the scale is valid or not.
      Glenn

      Reply
  • 103. ronica  |  March 10, 2010 at 7:00 am

    hi! do you have any standardized questionnaire or an internationally accepted questionnaire regarding students satisfaction using a likert scale? thank you! :))

    Reply
    • 104. Glenn  |  April 16, 2010 at 7:07 pm

      Sorry I am not aware of any standardised questionnaires on student satisfaction – would be a good idea!

      Reply
  • 105. Mary  |  March 12, 2010 at 1:45 am

    Who would I need to contact for permission to use the Likert survey? Need to use a pre-test and post-test for research purposes. Thanks, Mary

    Reply
    • 106. Glenn  |  April 16, 2010 at 7:08 pm

      the Likert scale has no copyright attached to it. It is a good idea to pre-test it.

      Reply
  • 107. P. K. Au  |  March 13, 2010 at 5:09 am

    I used 5 options of Liker scale for the Consumer response to advertisements in the HOE model. What are the limitations? Will the variations be not good, too small, or inadequate to show the analysed outcomes?
    As I have used this 5 point Likert Scale, I don’t want to change but I am afraid my supervisor will challenge me why I used it instead of the 6 options? How will I reply? Please enlighten.
    Thank you very much indeed for your help.

    Reply
    • 108. Glenn  |  April 16, 2010 at 7:10 pm

      Studies show that a five point scale is slightly more reliable than a six point scale. see this article:
      1) 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

      Reply
  • 109. P. K. Au  |  March 13, 2010 at 5:12 am

    I used the 5 options of Likert scale for the consumer response to ads as to prove linkage of HOE model. Will the variations be too small, no good, too narrow, or no meaning at att? My supervisor will me on why I used this 5-point? As I have used it, I don’t want to change to 6-point as time is of a factor. How will answer my supervisor? Please enlighten. Thank you very much indeed for your answer. Have a nice day!

    Reply
  • 110. quenny  |  March 27, 2010 at 7:49 am

    Sir, if my moderator and panel ask me why i choose 4 likert scale for my questionnaire to measure attitude, any recommend for me to answer for this question?

    Reply
    • 111. Glenn  |  April 16, 2010 at 7:00 pm

      I would say that four scale is useful if you want a good split between
      negative and positive. Plus four is a scale that people can deal with
      (more than seven is too much)

      Reply
  • 112. MC  |  April 11, 2010 at 11:19 am

    Dear Glenn

    I am in the process of researching and writing up my methodology for my MSc dissertation i would be grateful if you could send me teh references for the studies that you have looked at. Many thanks

    Reply
    • 113. Glenn  |  April 16, 2010 at 7:01 pm

      Hello, I’ve sent them to you.
      Glenn

      Reply
  • 114. Rob  |  April 26, 2010 at 3:01 pm

    In your research did you run into any pros or cons regarding the order of statement responses? Is the order of Strongly Agree to Strong Disagree better or worse than Strongly Disagree to Strongly Agree?

    Reply
  • 115. Stephie Birkner  |  May 1, 2010 at 6:36 pm

    Hi Glenn,

    great to found your short reader on Likert scales and the attached blog!

    Could you be so kind and send me your reference list you mentioned above.

    Thanks a lot
    Cheers Stephie

    Reply
  • 116. Markus  |  May 14, 2010 at 9:41 am

    Hello Glenn,

    thanks for the posts about the Likert scale! Regarding this, I’ve a question:

    I’m using 5-point Likert scales for most of my questions, however I also want to use a 3-point Likert scale for some questions.

    Is validity influenced in any way by using two types of scales within one survey?

    Thanks, regards from Austria!
    Markus

    Reply
  • 117. nataly  |  May 22, 2010 at 2:26 am

    Hellow Glenn
    I used 6 point likert scale before and after a course. But the scale’s category seems to have 4 positive and 2 negative. Can you explain me the possible bias? I used the same questionnaire before and after a course. Can you advice me refrences for this issue?
    thanks a lot
    Nataly

    Reply
  • 118. nataly  |  May 22, 2010 at 4:55 pm

    My study is about students’ skills development. To measure the development I used a questionnaire 6 point Likert scale and here is the problem I need your opinion:

    1. the scale points are:
    1- I do not agree at all
    2- I do not agree
    3. I tent to agree
    4. I agree in general
    5. I greatly agree
    6. I agree absolutely

    In my study I used the same questionnaire twice “before and after the treatment”. I found significant differences in the mean total score between “before and after the treatment”

    I now realize that my scale is miss-alignment (2- not agree 4- agree).

    Is the data I got remained valid because of the “before and after” design (two measurements with the same questionnaire)? How can I explain it?

    Can you help me please?

    Thanks and Best Regards
    Nataly

    Reply
  • 119. Ben  |  June 4, 2010 at 9:33 am

    Hi Glenn, I’m doing an assignment on customer loyalty. For my questionnaire survey, my supervisor says I have to use previous questionnaires that has already been tested. Here’s the problem, I have checked everywhere possible and I can’t find anything useful.. The factors being tested in the questionnaire are product quality, product taste, price, promotional efforts (the 4Ps) and the dependent variable is Customer Loyalty.. Pls could you give me some leads on where I could find previous questionnaires regarding this research..

    Thank You…
    Ben

    Reply
  • 120. Markus  |  June 4, 2010 at 10:03 am

    Ben, take a look here:

    *) Bearden, W. O., & Netemeyer, R. G. (1999). Handbook of marketing scales: Multi-item measures for marketing and consumer behavior research (2. ed.). Thousand Oaks, Calif.: SAGE Publ.

    *) Bruner, G. C., & Hensel, P. J. (1992). Marketing scales handbook: : a compilation on multi-item measures. Chicago, Ill.: American Marketing Association.

    Both will give you an idea and further references. Cheers! Markus

    Reply
  • 121. Ben  |  June 5, 2010 at 6:39 am

    Thanks a lot Markus.. Will do just that.
    Cheers!

    Reply
  • 122. Glenn  |  June 6, 2010 at 2:14 pm

    Hi Markus and Ben,

    Thanks for those tips Markus. Ben, I would also suggest you look at the main survey websites (Zoomerang, Survey Monkey, Prezza, etc. ) as often they have examples of surveys (marketing including) online.

    You may also want to point out that the companies are certainly doing surveys of this nature all the time but most surveys would remain proprietary and they would not necessarily be willing to share them.

    Good luck!
    Glenn

    Reply
  • 123. Lora Feld  |  June 6, 2010 at 3:01 pm

    I want to use a Likert scale for my master’s thesis. My adviser says “Likert scales can be valid research tools if they have undergone appropriate testing for validity and reliability”. She dropped that on me and I’m supposed to figure this out. I thought Likert scales were pretty standard: agreement, frequency, importance, quality, likelihood. Is it the questions you ask that have to have validity/reliability? If so, how do test THAT? Thanks.

    Reply
  • 124. Ben  |  June 17, 2010 at 4:00 am

    Thanks a lot Marcus and Glenn. I’m will check the main online survey sites Glenn, thanks again. Lora, wish I could be of much help now but I’m still an undergraduate so I guess your thesis is way above my level. I’m currently doing my undergraduate thesis now.
    Cheers!!!

    Reply
  • 125. Ben  |  June 17, 2010 at 1:48 pm

    Hi Glenn,
    The online survey sites only allow users to conduct surveys by creating online questionnaires and analyzing them through the website. These sites do not to exactly allow users to extract questions from previous questionnaires. Thanks again for the sources.
    Cheers!!
    Ben

    Reply
  • 126. beenish  |  July 21, 2010 at 4:49 pm

    hi glen iam masters in sociology would u please guide me how can i earn for me through internet

    Reply
  • 127. Michelle  |  August 28, 2010 at 9:52 pm

    Can you tell me how to make a likert scale with positive and negative for a questionaire in excel?

    Reply
  • 128. Kate M  |  September 7, 2010 at 9:09 am

    Hi Glenn,

    I am currently writing my dissertation for my MPH and am using a 5 point scale in a questionnaire for adolescents. Your comments above are really useful as to why I should use a 5 point scale but do you have a reference that I could use in support of using this with adolescents or even source materials that have recommendations for adolescents?

    thanks

    Kate

    Reply
  • 129. munthakrao  |  September 15, 2010 at 4:23 pm

    Hi Glen,
    My survey for my MBA dessertation 24 questions involved a mixture of 5,4,3 scale. Which were unavoidable. So what should be my approach to analysis these questions?. Can you suggest some references or site which can explain me as to how to approach the analysis part?

    Reply
  • 130. Glenn  |  September 15, 2010 at 7:20 pm

    Kate, sorry I know of no literature on using likerts for adolescents, but there is this study on using the likert scale with children: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

    You might find some references within the article

    Reply
  • 131. Glenn  |  September 15, 2010 at 7:21 pm

    Munthakrao, you can make some mathamatical calculations to try and make the scales comparable, you should consult a statistics expert – that’s not me!

    Reply
  • 132. Jason  |  September 15, 2010 at 8:30 pm

    Glenn,
    Do you have a couple references that note that using number scales are more difficult?

    Reply
  • 133. Glenn  |  September 15, 2010 at 8:58 pm

    HI Jason,

    Yes this article indicates that number scales are less understandable for people than those with labels (e.g. “good”, “poor”, etc.):
    1) 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

    Reply
    • 134. Jason  |  September 16, 2010 at 2:51 pm

      Thanks very much, Glenn!

      Reply
  • 135. Greg Yelland  |  September 20, 2010 at 7:41 pm

    Hi Glenn,

    This is a great resource. I’d appreciate it if you could tell me what, if any, the impact of changing scale item values (i.e., Strongly Agree =1 or 3) has on the reporting. Also, are you aware of any “best practices” with regards to Likert scale values?

    Reply
    • 136. Glenn  |  September 21, 2010 at 7:49 pm

      Hi Greg,

      I believe you are asking about whether scales should go left to right or right to left? If so, here is something I wrote earlier to another query:
      I’ve written about this earlier to another reader, I quote:

      “I believe it’s better to go from the negative to the positive, left to right. It seems more logic to me and some automated survey software mark your answers and calculate the responses for graphs on this basis, e.g. that the first point is the lowest. But I’ve had clients argue that it should be the opposite way around – put positive to negative, left to right – as people will click on the first point by default – which I personally don’t believe. I’ve never found any academic reference supporting either way but looking at all examples in academic articles, 95% are written as negative to positive, left to right.”

      In the post above, I list the key best practices that I know of.
      Glenn

      Reply
  • 137. vie  |  September 21, 2010 at 7:34 pm

    could u give me a 5 point likert scale fitted for theobservation of headlice.thank you…

    Reply
  • 138. Glenn  |  September 21, 2010 at 7:47 pm

    Hi Elvira, sorry not a specialist on headlice – not sure I get what you want to do, but do you mean infection level? Then you could ask something like, what is the current infection level of headlice; No infection, little infection, Some infection, A lot of infection, full infection. It’s not perfect, but now my head is itching…

    Reply
  • 139. Patricia  |  October 8, 2010 at 11:27 pm

    I am using a 1-5 scale. Most people answer questions with 4 or 5. For example, out a group of 20 people, 2 will rate 4 and 18 will rate 5. What would be an appropriate formula to use in an excel spreadsheet to quantify this information to give the average?
    Thanks for any assistance you can provide – math is not my natural forte.

    Reply
  • 140. glenn  |  October 9, 2010 at 2:29 pm

    Patricia, it’s quite simple. There are usually two basis calculations done:

    Percentages: For example, if your total responses are 20, and 2 responded to 4 and 18 to 5, it makes it that 10% have responded to 4 and 90% responded to 5.

    Average response rate: Then you can also make the average response rate also- 18×5 and 2×4 divided by 20 = 4.9. that’s your average response.

    Glenn

    Reply
  • 141. mirely  |  October 26, 2010 at 1:18 pm

    Up until now, I have been running a repeated survey using a 7 point Likert scale and labelling only the poles: I disagree completely – I agree completely; very poor – exceptional.
    I felt labelling every single point in between made the survey look too cramped.
    It has now been suggested to me that I label each individual point from now on and must now argue my case. What do you think Glenn? What would be more correct? Could this affect a longitudinal analysis? Could I please have the links you mentioned?
    Thanks a lot

    Reply
    • 142. Glenn  |  October 27, 2010 at 3:23 pm

      Hello Mirely,

      The study I read showed that labelling every single point is more reliable than just the pole points (even if it does admittedly make your survey look more crowded). This is explained in this article:
      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
      Glenn

      Reply
      • 143. mirely  |  October 28, 2010 at 8:00 am

        thank you!

  • 144. Mike  |  November 12, 2010 at 12:03 am

    Need Help

    I have data from a survey comprised of several different measures each that used a different likert-scale (5- and 6- pt scales). I would like to factor analyze this data, but I am uncertain how to handle the data to run this analysis (e.g. do I need to transform the scores prior to the analysis?, etc.) Any help would be greatly appreciated.

    Reply
  • 145. Glenn  |  November 25, 2010 at 10:50 am

    It is possible of course but you need to consult someone who is good with statistics – to help you convert the data to ensure that it is comparable for your analysis

    Reply
  • 146. Ian  |  April 9, 2011 at 2:08 am

    Hi Glenn,
    Not sure if you are still monitoring this thread but here goes…

    I’m currently writing my dissertation and have found your likert-scale discussion very informative.
    I am particularly interested in the odd/even number of points debate; please could you send me any references which relate to this?
    Much appreciated!

    Reply
  • 147. Jen  |  April 14, 2011 at 5:06 pm

    The “Designing a useful Likert Scale” pdf link appears to be broken…

    Reply
  • 148. Ahmed  |  April 20, 2011 at 12:52 pm

    Hi! Michael
    Thanks for posting the above useful links. They helped me a lot. Could you please help me in explaining as to how to analyse 3 and 5 points scale with regards to measurement type (ordinal and interval ).

    Reply
  • 149. Ahmed  |  April 20, 2011 at 12:54 pm

    Hi Glenn!
    Sorry for posting your name wrongly as Michael.

    Reply
  • 150. glenn  |  May 2, 2011 at 1:27 pm

    Ahmed, all the information you would need to make such an analysis is any standard statistics textbook.

    GLenn

    Reply
  • 151. NUHU  |  May 18, 2011 at 10:03 am

    how valid and reliable is the use of 3points likert scale in measuring efficiency in fixed assets management or any other point

    Reply
  • 152. glenn  |  May 22, 2011 at 6:16 pm

    a 3 point scale is considered reliable and valid for most cases, but largely depends upon your question and response categories.
    Glenn

    Reply
  • 153. Shelly  |  May 23, 2011 at 7:43 pm

    Excellent Blog. I am presently doing my thesis and I am using a three point likert scale for the children. I would therefore like the reference for the point you made about likert scales and children. Thanks

    Reply
    • 154. Glenn  |  May 27, 2011 at 12:41 pm

      Thank you, I’ve sent them to you.
      GLenn

      Reply
  • 155. Aleks  |  May 26, 2011 at 5:32 pm

    HOLA…
    MUY BUENOS DIAS
    SI ME PODRIAN MANDAR LAS PAGINAS DEL LOS LIBROS DE DONDE SACARON LA INFORMACION
    CON EL CUAL ELABORAR EL SITIO MEDICION INTELIGENTE Q ES ESTA DIRECCION
    SOBRE EL TEMA DE Escala de Likert y encuestas – las mejores prácticas
    http://intelligentmeasurement.wordpress.com/2007/11/20/likert-scale-surveys-best-practices/
    PORFAVOR—YA QUE ME PARESE INTERESANTE LO QUE PUBLICARON…

    SIN PTRO EN LO PARTICULAR-ESPERARE LA INFORMACION QUE SE SOLICITO. QUE ESTEN BIEN Y SALUDOS

    GRACIAS…!!!!!!!!

    Reply
  • 156. Cheri Quinton  |  June 7, 2011 at 9:32 am

    Hi, you have some excellent info on your blog. I am doing my thesis on creating a likert-scale and was wondering if you could pass on your references.

    Thanks heaps :)
    Cheri

    Reply
  • 157. karun  |  June 29, 2011 at 6:46 am

    Hi glenn. I have used a 3-point likert scale for my MBA summer project survey. I’m using the data to run a “principle component analysis” in spss software. Is it OK to use a 3-point scale for this analysis..? Actually i am looking for advantageous points to quote in my project review, as my project is almost over (collected about 270 responses) & i cant go back to correct the scale. So can u please tell me statements that would support the use of 3-point scale..??

    Reply
  • 158. Glenn  |  June 29, 2011 at 3:02 pm

    hello Karun, there is no unconditional statement that I’ve seen supporting a 3 point scale. As I mention in my post above, most scales used are 3, 5 or 7. Please look at the following article:
    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

    Reply
  • 159. bienomuchoawesomo3000  |  June 29, 2011 at 11:22 pm

    como se llama muy bien un muy bonitas un muchos feo

    Reply
  • 160. Nana  |  June 30, 2011 at 6:01 am

    Hi Glenn, your blog is very informative and useful. The discussions are just what I need right now. Thank you..

    Reply
  • 161. Grace  |  August 13, 2011 at 12:49 am

    Dear Glenn:

    I have revised a scale that measures parents’ awareness and behavior (what they think and what they are doing). I would like to use a 5-point Likert type response that measure “FREQUENCY.” I am thinking
    1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Very Often, 5 = Always
    or
    1 = Never, 2 = Rarely, 3 = Sometimes, 4 = often, 5 = always

    Have you seen article that use a frequency type scale?

    Thank you!

    Reply
    • 162. Glenn  |  August 14, 2011 at 6:32 pm

      Hi They are both scales used often (or very often!) so you could comfortably use either of them. Glenn

      Reply
      • 163. Grace  |  August 15, 2011 at 3:56 pm

        Thank you very much, Glenn

  • 164. Doofie  |  September 2, 2011 at 8:50 am

    Brilliant article and interesting responses! Friedman and Amoo’s (1999) ‘rating the rating scales’ may be helpful for a lot of the questions posed above.

    Reply
  • 166. Tabita  |  October 10, 2011 at 8:04 pm

    Dear Glenn,
    Would appreciate your view on the following. We recently reverted back to a 5-point rating scale in our Performance Management process, with 1 = Poor, 2 = Needs improvement, 3 = Satisfactory (this is also known as the Target), 4 = Good, 5 = Excellent. Now, we are a little stuck in respect of handling absolutes. We have certain objectives, where the target is the expectation and you can’t do anything to exceed the target. Does one assign a 3 rating (managers then feel they are discriminated against as the overall score is linked to bonuses / increases) or does one assign a 5 rating in the case of absolutes. Your assistance will be greatly appreciated.
    Kind regards
    Tabita

    Reply
    • 167. Glenn  |  October 12, 2011 at 1:44 pm

      hi Tabita, I’m not sure I understand the problem – what do you mean by “absolutes”? People who score a 5?
      Glenn

      Reply
      • 168. Tabita  |  October 13, 2011 at 6:30 pm

        Dear Glenn
        Thanks for getting back to me. I’ll include an example – we have certain deliverables that managers will either deliver on or not, e.g. transformation. The managers will either support the organizations’ transformation targets or not. They can’t exceed the target. Now the challenge is that if the maximum rating they can achieve for said objective then is a 3 (which is our “target met” rating on the scale). This means that their maximum overall score will be less than someone who can potentially achieve a 5 rating on each objective. Which in turn would affect the % increase / bonus they can potentially earn.
        Your assistance will be greatly appreciated.
        Kind regards
        Tabita

  • 169. Glenn  |  October 13, 2011 at 6:42 pm

    Hi Tabita, I think it’s a strange approach that some can only achieve a maximum of 3 where as others can receive a 5. It seems for me if a manager supports the transformation target then it should be scored as a 5..

    Reply
    • 170. Tabita  |  October 13, 2011 at 7:15 pm

      Thank you Glenn, appreciate your feedback!
      PS The others I refer to potentially achieving a maximum score of 5, would not be accountable for transformation, e.g. specialists.

      Reply
  • 171. Glenn  |  October 13, 2011 at 7:21 pm

    yes, the fact that such an achievement is stopped at 3 seems strange, and that it only applies to managers. Logically if you’ve achieved the task with success then it should be a 5. It may be worth comparing all main tasks to see what does it take to get a 5…

    Reply
    • 172. Tabita  |  October 13, 2011 at 7:47 pm

      Thanks :) We are planning on reviewing the performance contracts and will definitely keep that in mind.

      Reply
  • 173. Kofi  |  October 19, 2011 at 6:34 am

    Hi Glenn,
    I would be very glad if you can link me to articles ( Authors) who justify the use of a 5 point Likert scale. I want to use such a scale but my supervisors want me to justify why the use of a 5 point but not a 7 point scale. I have seen so many researchers using the 5 point but they do not justify why they chose the 5 point over the 7 point.
    Looking forward to hearing from you
    Kofi

    Reply
  • 174. Criado de Verdade  |  October 28, 2011 at 2:35 am

    Thanks! I cited this post in my research.

    Reply
  • 175. Paul deSousa  |  October 28, 2011 at 2:14 pm

    Great stuff. Thank you

    Reply
  • [...] written previously about the Likert scale and surveys – and received literally 100s of enquiries about it. A [...]

    Reply
  • 177. Nick  |  December 6, 2011 at 1:31 pm

    Hi Glen,

    Thanks for the upload. The article is a big help. I was wondering if you could send me the references that you mentioned. I’m doing a paper on how intrapreneurship effects employee engagement and I’m using the Likert scale for the data collection. The references would be a great help. Thanks in advance.

    ‘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)’

    Cheers.

    Nick

    Reply
  • 178. VS  |  December 7, 2011 at 1:52 pm

    Hi
    I am looking for a 5 point likert scale that uses faces (like small emoticons) e.g. very happy, happy, ok, sad, very sad
    Can you give me details of where to find this easily and asap, please!
    Thank you!
    Vick

    Reply
  • 179. glenn  |  December 12, 2011 at 12:39 pm

    Sorry Vick, I have nothing like that, you’d be better to search online – these scales do exists, see these examples:
    http://mewhoopee.blogspot.com/2009/02/evaluation.html
    http://www.5pointscale.com/cover_Incredible5PtScale_main.jpg
    http://www.paradigmcorp.com/blog/?tag=likert-scale

    Reply
  • 180. Chris  |  December 27, 2011 at 10:08 pm

    Dear Glenn,

    Could choices like yes /no / i dont know be analyzed using chi square?

    Reply
  • 181. Joanne  |  January 6, 2012 at 9:32 pm

    Hi there

    I wonder if anyone can help me. I am currently designing a survey with 5-point likert scales. The survey contains a number of items assessing attitudes and a number of items assessing the frequency of thoughts, feelings and behaviours. I know it is probably better to have the same response options within one measure, but a frequency and attitude format appeared more suitable for what I wish to measure (i.e. the endorsement of a number of social influences). Is it acceptable to include two types of response formats within one measure and will it affect the analysis in any way? If anyone could help or recommend links/academic sources I would be very grateful :-)

    Reply
  • 182. cee  |  January 12, 2012 at 9:36 pm

    Hi Glenn,
    Your article is great. I’m also doing a research project with children and am going to use the likert scale now. Would you recommend using a 3 point or 5 point scale?

    Could you email me your references too?

    Reply
    • 183. cee  |  January 12, 2012 at 9:37 pm

      Meant to say the children are between 6 and 8.

      Reply
  • 184. Glenn  |  January 18, 2012 at 8:25 am

    Hello Claudine,

    This reference will be of interest to you:
    “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

    As for 3 or 5 point scale, it really depends upon what you need to measure – how many degrees of difference you need. Studies are non-conclusive as to whether 3 or 5 are better.

    Glenn

    Reply
  • 185. pjkwisPam  |  January 19, 2012 at 10:47 pm

    You seem to have opened the flood gates for questions by writing this post. :-) My question: If I am utilizing a pre-validated measure that uses a 7-point Likert scale, would it invalidate the measure to use a 5-point Likert scale instead? My other constructs are using 5-point, and I would like to stay consistent if at all possible. Thanks!

    Reply
  • 186. glenn  |  January 26, 2012 at 2:52 pm

    Hi Pam, yes many question on Mr Likert! your main issue will be comparison of the data, but with some simple calculations you can make them comparable (looking at the mean for example). If you look up the stream of comments in June 2008, John Dawes explains how he did this.

    Glen

    Reply
  • 187. Lisa  |  February 9, 2012 at 3:59 am

    Glenn, I don’t know if Vicki still needs the info, but what she is referring to are Kunin Faces. Here is the 1955 article on its development: http://onlinelibrary.wiley.com/doi/10.1111/j.1744-6570.1955.tb01189.x/abstract

    Reply
  • 188. Glenn  |  February 9, 2012 at 4:45 pm

    Thank you Lisa for the information!
    Glenn

    Reply
  • 189. lynn  |  March 1, 2012 at 3:38 pm

    Can anyone recommend a book/article on how to analyze the likert scale data after the survey is closed? my data is 5-point from strongly disagree to strongly agree. Thanks!

    Reply
  • 190. gino  |  March 17, 2012 at 6:09 am

    could you please help me. It’s very needed for our research paper. do you know the limits of 3 point likert scale? i really need it asap.

    Reply
  • [...] November 2007 – Likert scale & surveys – best practices: This post is my most popular of all time – generating a record 181 comments! People just love to [...]

    Reply
  • 192. scott bayley  |  March 19, 2012 at 4:39 am

    The following article reviews the use of different types of Likert scales and discusses how to empirically assess if a given scale is functioning properly:

    Bayley, S. 2001, ‘Measuring customer satisfaction’, Evaluation Journal of Australasia, vol 1, no 1, pp. 8-17

    Reply
    • 193. Ain  |  May 10, 2012 at 6:22 am

      Hi Scott,

      I’ve been trying to find the article you referenced but to no avail. Is there any way you can direct me to it?

      Thank you,
      Ain

      Reply
  • 194. Antohny Bouyer  |  March 27, 2012 at 10:03 pm

    Can I use a likert scale for a yes/no survey

    Reply
  • 195. rechie  |  April 1, 2012 at 4:07 am

    hello,

    i just would like to ask what would be the best indicators to be used in measuring the intensity of media exposure of the students? help me please.

    Reply
  • 196. orangbaik  |  April 7, 2012 at 7:17 pm

    Hello,

    I am in need of help for my research. I administered my questionnaire using 5-point likert scale (1=very unprepared, 2=unprepared 3=quite prepared 4=prepared 5= well-prepared) to 500 participants. Now that, I am trying to analysis the data, I am putting them into categories in terms of mean. For example, below 2.5 is considered unprepared etc. My question is how do I determined the cutoff mean for these 3 categories? Do I take below 2.4 as underprepared, 2.5 till 4 as prepared etc. Thank you. Any advice will be much appreciated.

    Reply
    • 197. Glenn  |  April 16, 2012 at 6:46 pm

      Hi, Isn’t it quite simple – you should have three categories – 1-2 very/unprepared 3 middle 4-5 well/prepared. No need to use half point scales as your cut offs should match your scale.
      Glenn

      Reply
  • 198. Rasmus  |  April 18, 2012 at 9:34 am

    Hello!

    Thank you for the excellent information and following discussion, very helpful for my thesis. I would also be very greatful if you could e-mail me your references. Thanks in advance!

    //Rasmus

    Reply
  • 199. Priya  |  April 20, 2012 at 6:42 am

    Hi Glenn,Interesting to read these discussions. I am doing a study on percention of health providers to team based work using a Likert scale( agree to disagree).Since the set up is very rural ,it becomes extremely difficult to explain the Likert scale -The comprehension level of providers is really low.How do we work with that? Any alternatives you can suggest?

    Priya

    Reply
  • 200. Drabarb  |  May 4, 2012 at 1:52 pm

    I do a lot of surveys of school programs and we usually have numerous items for which we use a 3-point Likert scale to measure respondents’ perceptions of quality. We’ve tried:

    Needs improvement – adequate – exemplary
    Poor – fair – good
    Needs improvement – satisfactory – good

    We’re still not satisfied that the middle point on any of these is perceived as the middle point.

    We’re also trying to reduce respondent bias (e.g., a principal will tend to want their school to “look good.” So that means the lowest point selected is usually the mid point.We want the survey to be a chance to reflect on what needs reform, not to be perceived as a compliance check or a PR opportunity.

    Any suggestions?

    Reply
  • 201. khellaf boumkhila  |  May 8, 2012 at 2:15 pm

    hi. thank you very much, i would like to know:
    How can you analyse data from a Likert Scale?

    Reply
  • 202. Nurul Ain  |  May 14, 2012 at 2:49 am

    Hi Glenn,

    I am looking for readings that justify the 6-point scale that utilizes the “somewhat agree” and “somewhat disagree” ratings in the middle portion of the scale. The survey my project is working on is for teachers to assess certain policies of their working environment. Do you have any suggestions to what readings I can look into?

    Thank you.

    P/S: I’ve looked through the Alwin and Krosnick reading you referenced earlier but it is not very helpful in justifying the 6-point scale.

    Reply
  • [...] it started out as a customer service survey: A simple 10-question survey that implemented the likert scale, close-ended, and open-ended questions. I even created a Gantt Chart for my co-intern and I to [...]

    Reply
  • 204. Glenn  |  May 20, 2012 at 3:50 pm

    Drabarb, your situation is often seen in self-assessment – the selection is overwhelming positive. You could increase the scale to five points, that may partially resolve the problem. There are also certain statistic methods you can apply to counter a over-positive bias – you’d have to ask a statistics expert…

    Reply
  • 205. Glenn  |  May 20, 2012 at 3:51 pm

    Nurul,

    I am not sure that you will find a specific justification for the 6-point scale. Perhaps check this article:

    RATING THE RATING SCALES by Hershey H. Friedman, Ph.D.
    Professor of Marketing and Busines
    http://academic.brooklyn.cuny.edu/economic/friedman/rateratingscales.htm

    Reply
  • 206. Danielle  |  June 4, 2012 at 2:45 pm

    Hi Glenn,
    Excellent post. Concise and easy to understand.
    Could you please send me your list of references? Also, I noticed the article by Alwin and Krosnick is dated 1991. Do you have any research that is more recent?
    Thanks!
    Danielle

    Reply
  • 207. Alexis  |  June 21, 2012 at 8:00 pm

    I second Danielle’s comment. Great article, but I would love the references. Please send a list if you are willing.
    Thank you.

    Alexis

    Reply
  • 208. Morgan Howard  |  July 2, 2012 at 11:01 pm

    Quick Question.
    If using a MPAS, Likert scale where previous totals for the 19 questions is out of 95. However in the 19 i have asked some have response options of 2, 3, 4 or normal 5 point. Am I correct in thinking I can rescore as follows (to keep each response out of 5);

    2 point;
    1=1
    2=5

    3 point;
    1=1
    2=3
    3=5

    4 point;
    1=1
    2=2.33
    3=3.67
    4=5

    5 point as normal.

    Questionnaires have gone out and it is regarding SPSS really. Just wondering if you had any advice.

    Reply
  • 209. Glenn  |  July 4, 2012 at 6:46 am

    I think that seems the most logical approach. I can’t see another solution. Just keep in mind that you have a different distribution (i.e. on a 2 point converted scale, you can’t have a “3” in any conversion). All the more reason to use consistent scales across future questionnaires.
    Glenn

    Reply
  • 210. CardiffPostgrad  |  July 14, 2012 at 2:43 pm

    Hi Glenn,

    Very interesting post. I’m currently undertaking some research for my dissertation and I would love to know your references – I’m currently constructing my questionnaire so looking for sources to base my questionnaire development on.

    Regards

    CardiffPostgrad

    Reply
  • 211. meili  |  July 28, 2012 at 6:56 am

    hi glenn,

    would you please email me any reference saying that 11-point likert scale was a good choice, if any? Thanks so much.

    meili

    Reply
  • 212. marthauganda  |  August 27, 2012 at 9:21 am

    This is going to be very helpful for my research…

    Reply
  • 213. Abdisalam  |  October 26, 2012 at 12:52 pm

    Hello, Glenn,
    I’m doing MBA assignment IT security management practices and my adviser told me to use 5 Likert scale, could you please send me my email the best form of five likert scale and A.P style reference.
    Thanks,
    Reply
    Keinan

    Reply
  • 214. Will  |  February 1, 2013 at 2:44 pm

    Hi Glenn,

    Please could I have a reference for the point that numbered scales are more difficult to understand.

    Thanks,

    Will

    Reply
  • 215. koshyfamily8554K. Koshy  |  February 28, 2013 at 7:53 am

    Dear Sir,
    Would you please give me details on how to make us of Linkerts Scale in the Researches

    Reply
  • 216. michael  |  March 4, 2013 at 10:15 am

    Hello,

    You mention some studies showing that scales with over 7 points are often difficult. Can you shed any more light on that or provide a link to the reference please? I am wondering why this conclusion was come to.

    Thanks

    M Harris

    Reply
  • 217. khairia  |  March 17, 2013 at 9:14 pm

    Hi.
    I have used 5 and 7 scale in my project that was used in other studies. now I need to justify why I use two different scale? could you help me in this point?
    Many thanks in advance

    khairia

    Reply
  • 218. DeLaine Goito  |  June 14, 2013 at 2:35 pm

    I need to develope a three point scale using visuals of emotions to assist a special needs student to recognize his feelings. Do you have any recommendations?

    Reply
  • 220. Rose  |  July 25, 2014 at 3:08 pm

    This article is really a pleasant one it assists new net
    viewers, who are wishing in favor of blogging.

    Reply
  • 221. caveman  |  July 30, 2014 at 12:54 pm

    It’s going to be end of mine day, except before end I am
    reading this impressive piece of writing to improve my knowledge.

    Reply
  • 222. Karry  |  July 31, 2014 at 11:14 am

    I added half and half to my coffee in the 2nd week which is a major no, no on this diet and I stayed at the same
    weight for the week. Some doctors will tell you not to exercise at all while taking it.
    It aromatizes (changes) to estrogen, a molecule proven to increase fat storage and
    decreases muscle mass.

    Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Trackback this post  |  Subscribe to the comments via RSS Feed


Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 861 other followers

Categories

Feeds


Follow

Get every new post delivered to your Inbox.

Join 861 other followers

%d bloggers like this: