Likert Scale was invented by and named after the american social scientist and psychologist Rensis Likert. Likert Scale is a psychometric survey question scale that uses two extreme polarities to measure respondent attitude towards a statement. This attitude is captured in the form of response options - with agreement or disagreement response option types commonly used.
Below is a Likert Scale example question using 7 points in response options (we will discuss several more examples and questions further below) :
It is important to note that while researchers commonly use the phrase “likert scale question”, in reality, Likert Scale is not, strictly speaking, a question - it is in the form of a statement. The scale attempts to merely capture the respondent’s attitude towards the given statement. The reason we call it a question, is because it fits into the category of commonly used survey questions.
The quality of a Likert Scale question is judged on how “symmetric” or “balanced” the response options are. A simple thumb rule to create a good Likert Scale survey question is to ensure that the distance between the two extreme polarities in a Likert Scale must be equal. This means that there are equal, symmetric and unbiased opportunities for the respondent to select a response option.
A Likert Scale question may consist of even or odd number of response options. There is academic argument for using both types, however, odd number of response options are used in most instances because at the center of an odd response option is “neutral”. This allows for a more unbiased response option capture.
Using Likert scale questions with even number of response options is often described as a “forced choice” method where a respondent is forced to choose at least one polarity, with no option to stay “neutral”.
Likert Scale is part of quantitative research techniques and is often used in combination with other questions.
Let's understand the different likert scale questions and their response options with some examples.
As we discussed above, Likert scale questions can be in even or odd number of response options. We will first look at odd response options and then even response options, followed by an analysis of the two types.
Likert Scale questions with odd number of response options have a “neutral” option at the centre of the scale. This means that a respondent can stay neutral when they do not have or do not want to express an opinion on either polarities of the question.
Likert Scale example for 5 point response options:
Q. Dogs are more friendly than cats.
Answer Options :
1. Strongly Agree
5. Strongly Disagree
In the above example, we used a five point Likert scale survey as an example, however, seven and nine points Likert scales are also used in complex research studies.
Likert Scale example for 7 point response options:
Let's use a slightly different question to look at Likert Scale example for 9 point response options:
Q. NASA needs more investment from the government.
Answer Option :
1. Strongly disagree
3. Moderately disagree
4. Mildly disagree
6. Mildly agree
7. Moderately agree
9. Strongly agree
Likert Scale questions with even number of response options do not have an option for a respondent to stay neutral, hence to term “forced choice” is frequently used to describe this method.
Likert Scale example for 2 point response options:
Now let’s use the same question with 4 and 8 even response options.
Example for 4 point even response options:
Example for 8 point even response options:
Also view a live Likert Scale example in a survey : Likert Scale Survey Template
Likert Scale is part of the Four Core Measurement Scales that are frequently used by researchers.
While there are many opinions on the which type of likert scale leads to better analysis, let us objectively look at the two scale types and what additional research has to say:
1. Even Likert Scale Response Options : Having an even number of response options will reduce the chances that respondents will use the neutral response available, forcing them to express an opinion rather than taking the “easy” way out when replying to the question.
2. Odd Likert Scale Response Options : Having an odd number of options gives the respondent the option to select the neutral option when they are more likely to truly feel neutral about the topic, instead of forcing a biased response or causing the respondent to simply skip the question because they don’t want to commit to a positive or negative responses.
A study from 1972 published in the Journal of Applied Psychology found that having odd or even items in your rating scale will depend on what you want to get out of the study. The researchers tested different surveys that used anywhere from 2-19 options in the Likert-style scales. They tested for three things: testing time (i.e., how long it took respondents to complete the study), proportion of scale used, and proportion of “uncertain” responses being selected. The surprises: anything over 4 items used in the number of options in the scale resulted in about 60% of the scale being used. The more items in the scale, the less the incidence of the neutral value being selected; the 5-item scale had 21% use of the neutral score; the 13 and 19 point response options Likert scale had only 3% use of the neutral score. And time to complete the study didn’t seem to be consistently affected.
Cultural effect on Likert Scale Measurement
Another item to take into account that had not occurred to me until I started doing the research for the question: cultural differences play a role in how scales are used by respondents. A 2002 study in the journal Research in Nursing and Health studied responses by three ethnic groups: Chinese, Japanese, and Americans. This study found:
1. Construct validity (whether the question and answer options are actually measuring what you intend it to measure) was better for Chinese and American respondents when there were only 4 points response options in the Likert scale.
2. Construct validity was better for Japanese respondents when there were 7 options in the Likert scale questions.
3. Chinese and Japanese respondents were more likely to select a mid-range option when the question had to do with positive emotions. Americans were more likely to respond positively to the same questions.
Generally speaking, studies have found repeatedly that culture affects how a scale is used. If you want to really geek out about this aspect of scale usage and how response styles can surface in studies, you can check out a more in-depth study published in the Journal of Cross-Cultural Psychology from May 2014 that used a personality test to compare against tests using Likert-style scales. The paper outlines a two-part study that really dove into different response styles (acquiescent response style, extreme response style, midpoint response style, and socially desirable responding). In short, they found that, if conducting a cross-cultural study, response styles should be taken into consideration when looking at the data.
You may also like to read : Multiple Choice Questions and Question Types.
Unipolar scales are more contoured, allowing users to instead focus on the absence or presence of a single item. The scale measures the ordinal data, but most of the times unipolar scales generate more accurate answers. An example of a unipolar satisfaction scale is: not at all satisfied, slightly satisfied, moderately satisfied, very satisfied, and completely satisfied.
A unipolar Likert scale question type indicates a respondent to think of the presence or absence of a quality. For example, a common unipolar scale includes the following choices: not at all satisfied, slightly satisfied, moderately satisfied, very satisfied, and completely satisfied. It is arranged on a 5 point scale. A to E. Also, Unipolar question types lend themselves where there is a maximum amount of the attitude or none of it. For instance, let’s say, how helpful was the apple pie recipe? Very helpful, somewhat or not at all. From there, we can safely assume there is something in between–like “sort of” helpful. Below is a Unipolar Likert Scale Example
A bipolar scale indicates a respondent to balance two different qualities, defining the relative proportion of those qualities. Where a unipolar scale has one “pole,” a bipolar scale has two polar opposites. For example, a common bipolar scale includes the following choices: completely dissatisfied, mostly dissatisfied, somewhat dissatisfied, neither satisfied nor dissatisfied, somewhat satisfied, mostly satisfied, and completely satisfied. That is a scale with 0 in the middle (-3, -2, -1, 0, 1, 2, 3). Below is a Bipolar Likert Scale Example.
The difference between a Likert Scale and it's sibling - the Semantic Differential Scale, is that likert scale seeks to understand the respondent's attitude towards a given statement, where as, semantic differential scale actively seeks out respondent "opinion" and not attitude.