.png){kind=icon}
You will have often filled out questionnaires that use a Likert scale, where you are asked to give your level of agreement with a statement, such as the example below:
“I am really learning lots about experimental design from this book”
- Strongly disagree
- Disagree
- Neutral
- Agree
- Strongly agree
This is a categorical variable, but one that does not have clear boundaries between levels. You could imagine that you and I might feel exactly the same about the book but you would tick against ‘4’ and I would tick against ‘5’. Or you might find that if this question came at the start of a questionnaire you might choose ‘5’, but if it comes towards the end of a long questionnaire when you are feeling rather jaded about the whole enterprise you might choose ‘3’. But for collecting data on opinions, attitudes, and feelings constructing clear boundaries between levels of a categorical variable is generally very challenging.