I have collected data from Agri-officials from MOA-Saudi Arabia like their age, place of birth, current residence, level of education, area of specialization and their perception regarding privatization of extension and advisory services on a 5-point Likert scale. I have collected data from 94 officials. I have entered data in SPSS as well. Apart from descriptive statistics (mode, median and frequencies), I have some confusion in my mind regarding inferential statistics. I am treating data as ordinal. Kindly guide me that what test I will use if I want to find the relationship b/w age (for example- I have 4 categories of age) and their perceptions (15 Qs). Which test will be most suitable in this case?
The measurement level of the variable (i.e. whether it is ordinal or continous etc) is not the only thing to take into account when deciding how to examine possible associations between the variables. You need to think about the nature of the association - is it linear, curvilinear or something else? If you expect a linear association and you have two ordinal variables then using the Spearman correlation (or the gamma) is a good choice. If you believe the association is non-linear then you could use a chi-square test as well. When you only have 94 respondents it is a little bit difficult to use factor analysis but otherwise that might be a good suggestion as Jacek mentions (even if the requirements often ar interval-scaled variables unless you use special kinds of correlations).
- Badges
- Science topic
- Similar topics
- Agricultural Science