## Simplifying the Likert Scale

written by: Mayflor Markusic โข edited by: Elizabeth Stannard Gromisch โข updated: 10/17/2014

Teachers are always involved in educational research. In interpreting the results of research in which the method involved the use of questionnaires, one commonly used psychometric scale is the Likert Scale. Is this scale appropriate for your current research?

• slide 1 of 3

The Likert Scale is a popular format of questionnaire that is used in educational research, especially in the field of special education. It was invented by Rensis Likert, an educator and psychologist, who advocated an employee-centered organization.

Since the inception of this psychometric scale, there have been several versions based on the number of points in the scale. That is, the Likert scale can be four-point, five-point, six-point, and so on. The even-numbered scale usually forces a respondent to choose while the odd-numbered scale provides an option for indecision or neutrality. Below are the two Likert scales, the four-point and the five-point.

• slide 2 of 3

### The Likert Item

In making questionnaires that use the Likert Scale, not just any type of question will suffice. In fact, there is a format that must be followed in formulating the questions. The questions should ask for an agreement or disagreement. Below is an example of a Likert item:

The Bush Doctrine is an effective foreign policy.

Based on the item, the respondent will choose a number from 1 to 5 using the criteria below:

1 โ strongly agree

2 โ somewhat agree

3 โ neutral/no opinion

4 โ somewhat disagree

5 โ strongly disagree

There are questions that appear to be similar to the above, but are not real Likert items. Here are some examples:

How often do you visit the zoo?

1 โ Never

2 โ Rarely

3 โ Sometimes

4 โ Often

5 โ Always

The above question or target statement is a Likert-type item because it does not seek agreement or disagreement.

How old are you?

1 โ Below 18

2 โ 18-25

3 โ 26-35

4 โ 36-50

5 โ Above 50

The question above is neither a Likert-item nor a Likert-type item. It is actually considered as an ordered-category item.

Very informative 1 2 3 4 5 6 7 8 9 Not informative

The above question is considered as a discrete analog scale. It is similar to the item below:

Gregarious 1 2 3 4 5 6 7 8 9 Reclusive

When the teacher is putting together the target statements in the questionnaire that is supposed to follow the format of the Likert Scale, each item should be examined to determine whether it is actually a Likert item or one of the non-Likert items described here.

There is, however, an exception to the rule of agreement-disagreement in the Likert Scale. Below is an example of an acceptable Likert item:

How do you feel about the Principal's performance this year?

1 โ strongly approve

2 โ somewhat approve

3 โ neutral/no opinion

4 โ somewhat disapprove

5 โ strongly disapprove

• slide 3 of 3

### A Likert Interpretation

After creating the questionnaire and collecting the answers of the respondents, the educational researcher moves on to the task of interpreting the scores for each question. There are four possible general landscapes. For example, here is a Likert item:

The No Child Left Behind Act needs to be repealed.

1 โ strongly agree

2 โ somewhat agree

3 โ neutral/no opinion

4 โ somewhat disagree

5 โ strongly disagree

If there were ten respondents, the four possible landscapes would be the following: In the first column of responses, the researcher finds that there is a consistency of responses with the majority of the respondents favoring 4 (somewhat disagree).

In the second column of responses, the researcher also finds a consistency of responses but there is complete polarization. The respondents are divided into two opposing opinions about the NCLB Act.

In the third column of responses, the researcher finds that there is no consistency and there is no polarization.

And in the fourth column of responses, there is polarization, but there is also no consistency of responses. Whenever there is no consistency of responses, no clear conclusions can be drawn.