Hi,
I am a novice, first time researcher who started a research project with very little experience in designing questionnaires.
- My survey consists of 5 parts or sections with each section having an average of 5 questions.
- Some sections had branching questions so not all respondents would be presented with that section if they did not qualify for it.
- I also had a qualifying question in the survey, that would kick respondents out if they answered NO to the question which caused my survey pool to reduce from 200 to about 78!
- All of these design elements really left me with data that was too finely distributed along the 5-point scale - unless I combined the top box score (% responses) the neutral option had the biggest piece (% respondents)...i.e. the neither agree or disagree option is my mode for almost all questions.
To publish the research, is it statistically legitimate for me to consolidate my 5-point into a 3-point, i.e. consider the 2 top box scores on both the agree and disagree ends of the scale, to draw meaningful results from my data? Do you have any other suggestions on how I can about doing my analysis? Thank you!!
I found benefit from this forum and learned from several other threads that most of my survey questions fall into "likert scale data" instead of a "likert type data" see: https://joe.org/joe/2012april/tt2.php
Which means I have to take a different perspective on analyzing this data...
Hey
I *would not* consolidate a 5 point scale to a 3 point. But if you have 5 sections, and if each section is meaningfully different from each other, you can calculate means per section so that you have 5 values (I understand that some participants will have less than 5 though?)
The means can be then used as-is for analysis (more informative), or you can create groups by median split, depends on your question.
Shira Meir Drexler Thank you. Each section is meaningfully different, it probes probable barriers to low physician engagement with process improvement in healthcare. And the sections are connected to my overall research question.
- Badges
- Science topic
- Similar topics
- Biological Science
- Histology
- Sectioning