Matrix/Rating Scale Question
A Matrix question is a closed-ended question that asks respondents to evaluate one or more row items using the same set of column choices.
A Rating Scale question, commonly known as a Likert Scale, is a variation of the Matrix question where you can assign weights to each answer choice. Rating Scales automatically calculate a weighted average for each answer choice in the Analyze Results section.
Creating a Matrix/Rating Scale Question
To add this question type:
- Drag and drop Matrix/Rating Scale into your survey from the BUILDER section of the sidebar.
- In the Rows fields, enter the items you want respondents to evaluate.
- In the Columns fields, enter the measurements you want respondents to use to evaluate the row items. Column choices should be applicable to every row item.
- Configure any additional options.
- Click Save.
Make This a Single-Row Rating Scale (Remove Row Choices)
Selecting this option does three things at once:
- Removes the Row fields. You will only have to enter question text and Column Choices.
- Applies the Use Weights option so a weighted average will be calculated in the Analyze Results section.
- Enables the option to add Question Skip Logic to skip respondents to a future point in the survey based on their answer choice.
Allow Multiple Responses per Row (Use Checkboxes)
Use checkboxes to allow respondents to select more than one answer choice per row.
Assign a weight to each column, transforming the question into a Rating Scale or Likert Scale question with multiple rows.
- Add an N/A column: Add a column of "N/A" options that respondents can select if none of the other options apply to them. The answers from respondents that select the N/A option will not be factored into calculating the rating average. This option will only appear if you use weights.
Forced Ranking (one response per column)
Only allow respondents to select one choice per column. This asks respondents to compare the row choices to each other, similar to a Ranking question. This option will only appear if you only allow one response per row.
Under the Edit, Options, and Logic tabs, you can further customize the question in the following ways:
- Add an "Other" answer option or comment field
- Require an answer, and define a specific range or limit of rows that the respondent must answer
- Change the question size & layout
- Sort, randomize, or flip rows
- Add question skip logic, only if Make this a single-row rating scale is selected
- Add question & answer piping
Learn more: Editing Questions
Before you send out your survey, preview your survey design to see what your survey will look like to respondents.
Depending on the respondent's screen size, large Matrix/Rating Scale questions with many columns may require respondents to scroll to the right to see all of the available column choices. Keeping Matrix/Rating Scale questions short and simple can improve completion rates. Read more about the impact of question types on mobile vs. non-mobile completion rates on our blog: How to Keep Your Survey Respondents Happy on Mobile
If you require respondents to answer a specific number or range of rows, make sure you customize the error message to reflect the requirement, so respondents will know how to answer the question correctly.
If you selected Forced ranking (one response per column), respondents will only be able to apply each column choice to one row item. If a respondent tries to select the same column choice for another row, their previous selection will disappear.
Your results will be presented in a table with the questions or statements listed down the first column, and the answer choices listed across the first row.
If you enabled the Use Weights or Make this a single-row rating scale option when you created the question, you'll also see the average rating for each question or statement you asked respondents to evaluate.
If needed, you can change the weight of each answer choice in the Design section of the survey, even after the survey has collected responses.
The rating average is calculated as follows, where:
w = weight of answer choice
x = response count for answer choice
x1w1 + x2w2 + x3w3 ... xnwn
If you chose to include an N/A column on your Rating Scale question, any N/A responses will not factor into the weighted average.
For this example, we'll use a 5-point rating scale question. The weights assigned to each answer choices are shown in parentheses:
Responsibilities are shared fairly among your team members.
Neither Agree nor Disagree (3)
Strongly Agree (5)
After collecting responses to the survey, the results will look something like this.
|Strongly Disagree||Disagree||Neither Agree nor Disagree||Agree||Strongly Agree||Total||Average Rating|
|Responsibilities are shared fairly among your team members.||20.45%|
The average rating of 2.84 indicates that the average sentiment among respondents is that responsibilities are not shared fairly among their team members. The average rating was calculated as follows:
(9*1) + (12*2) + (9*3) + (5*4) + (9*5)
125/44 = 2.84
The Depth display option allows you to choose between the following options:
- Weighted Average: Charts the average rating for each answer choice.
- Distribution: Charts the absolute number or percentage of respondents that selected each answer choice.
The chart types available depend on the question type, and the display options you configure. Learn more: Custom Charts & Data Tables