In: Math
which statistical analysis to use for a survey of 5 questions with four choices each strongly positive, positive, neutral and negative
This type of Survey is a Likert Scale. The most commonly used measures to evaluate such data are as follows:
MEASURE | LIKERT SCALE |
Central tendency | Mean |
Variability | Standard Deviation |
Associations | Pearson's R |
Other Statistics | ANOVE, t-test, regression |
Essentially, there is a long standing dispute on the two valid ways of interpreting Likert Scale data. Parametric test such as 2-sample t -test assume a normal continuous distribution. Non-parametric tests, such as the Mann-Whitney test, do not assume a normal or a continuous distribution. However, there are concerns about a lower ability to detect a difference when one truly exists.
Over the year researchers have argued that median should be used as a measure of central tendency . Similarly it has been contended that frequencies, contingency table, chi square test, Spearman Rho assessment or Mann-Whitney test should be used instead of parametric test which require interval data.
In my personal opinion, since you have a very less sample size, any non parametric test as mentioned above, as well an expert qualitative review from your experience would be sufficient. For further analysis, I would require the purpose of the study of questionnaire. Usually it is preferable one designs the test before designing the questionnaire to facilitate the data collection process.