As we know, there are a different kinds of Likert based questionnaires. And I need to know the best techniques to analyze such questions for inferring a good and reasonable conclusion.
Any help is appreciated.
In order to infer justifiable conclusions from Likert responses, it is the question itself that needs attention. So to demonstrate that your Likert-based conclusions are solid, you will need to reference back to the quality of the question(s). The question needs to be non-ambiguous and matched to the available Likert scale the respondent will use when answering the question. In this way the Likert responses should be reasonably consistent from one respondent to the next. In order to demonstrate that, you will probably have to test the questions across a cross-section sample of respondents before rolling out the final questionnaire. The cross-section could take account of differences in age and other demographic differences, with the aim of showing your target audience are consistent in how they understand and then respond to the question(s).
Chris Kelly , First of all, thank you for your participation.
And next, you raised a good point for testing the questions across a cross-section sample of respondents which have to account socio-demographic difference. How can we test it? is it observational? or is there any significant statistical packages which can perform it? I guess you are focusing on cross tab tests like Monte carlo simulation, fisher exact test and chi-square test......am I right? How will you make sure of that your cross-sectional checking was fantastic? (Can you add more on this cross-sectional checking,please)
Mollalign
The answers to those questions will depend on the aims of your research. I.e. the demographic to which you are aiming your survey (eg old people, young people, business people, retirees, immigrants etc). As to testing question and response validity, I would start with a pilot test to a limited sample of respondents which ideally matches the demographic group(s) you are aiming at. Then study the pilot test results using thoughtful reflection to assess questionnaire ambiguity and Likert scale reliability. Hopefully that will result in improvements to the questions and Likert scales which in turn increase the usefulness of your final responses.
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning