How to Use the Likert Scale in Statistical Analysis

If you want to understand how effective students or parents think a teacher is across several different categories, a survey using a Likert scale can give you an answer. Instead of asking questions, a survey using Likert scales poses statements to participants and asks them to rate whether they agree or disagree. These ratings can be converted to numbers and used to generate a numerical evaluation.

A Likert scale is an approach to surveying in which participants are given a declarative statement and then asked whether they strongly disagree, disagree, are neutral, agree or strongly agree with the statement. These response categories also have numerical values, with "strongly disagree" equaling "1" and "strongly agree" equaling "5." Although five-point Likert scales are most common, it's possible to use any other number of points. The important thing is that the response categories form a continuum, with the first category being the point of least agreement and the last being the point of most agreement with the statement being tested.

In education, Likert scales are sometimes used to ask students and parents to rate teachers. For example, a teacher might hand out an evaluation form with a dozen Likert items like "The teacher seemed very knowledgable about the subject" or "The teacher explained the material well." Ideally, Likert scales should include equal numbers of positive and negative items, so an evaluation might include "The teacher seemed ignorant of the subject matter." Evaluation writers can include as many or as few items as they want on many different teaching topics, from preparedness to friendliness or professionalism to overall effectiveness.

To generate useful information from Likert scale data, you'll want to input the results using SPSS predictive analytics software or a similar statistical program. In SPSS' "variable view," create different variables for each item on your Likert scale. If you asked 15 questions, you should have at least 15 different variables. Then, in "data view," enter each respondent's answers to your questions column by column. SPSS accepts numerical data only, so you'll enter "1" for "strongly disagree," "2" for "disagree" and so on up the scale. When you've finished entering the data, go to the variable drop-down menu and select "recode variable." In that screen, you can take each of the negatively phrased items you included in the scale and reverse their values so all your items point in the same direction.

Once you have all your data entered and your variables recoded, you can calculate your teacher's average rating in each category or combine them all into an overall effectiveness score. To find results for each item, just go to the "statistics" drop-down menu, select the "descriptive statistics" option and choose the variable you want to compute descriptive statistics for. This will give the mean, median and standard deviation of ratings for a particular item. To calculate an overall score, create a new variable on the "variable" menu that adds all of the scale items together and then divides by the number of items. Then, run descriptive statistics on that variable by using the same process described earlier.