Your statistical approach should be considered even before you begin your thesis and while you write your literature review. Your statistical approach must consider:
-Your skill level.
-The topic you are discussing.
-Your hypotheses. If your hypotheses mention one variable causing another variable, correlation analysis may not be enough.
-Are you creating latent variables. If you are formalizing concepts that by themselves are not measurable, then you may find yourself dipping your feet in skills that, without our help, may be more than a standard student can handle.
- Model articles. Are there articles from other countries that attempted to take the approach you wish to take? What limitations did these authors face?
-The culture within your country. For a previously established questionnaire, reliability and validity may not exist because of local norms.
In other words, think about your methodology consistently and change it as information comes in.
Key words: thesis, topic, methodology, statistical, statistics discipline.