Hi, I would like to ask — is the use of descriptive analysis, for example to explore user perceptions, sufficient at the PhD level? The analysis only involves frequencies and percentages, without using measures such as standard deviation or mean. My research questions are fully addressed through descriptive analysis alone.
Thank you Dr Jochen Wilhelm
, appreciate your time to answer my question. I'll rethink about my method approach. Thank you once again
If there truly is next to nothing that is known about your topic, then a purely descriptive study might be justified.
FYI the standard deviation is a descriptive statistic, and it is often useful to know how wide the variation was in your sample.
Thank you Prof David L Morgan for responding to my query. However, how important is it for me to consider the wide variation within my sample? I often observe that descriptive analyses typically include measures such as the mean, median, and standard deviation. As I understand it, standard deviation is primarily used to assess the normality of the data, whether it is normally or non-normally distributed which subsequently guides the selection of either parametric or non-parametric tests (such as the Kruskal-Wallis test, t-test, Chi-square, etc.).
My question is, can descriptive statistics stand on their own without the inclusion of standard deviation? Is the descriptive data alone insufficient? If I do not intend to conduct any inferential statistical tests, is it still necessary for me to report the standard deviation in my thesis?
You can choose to report for not report any set of descriptive statistics that you choose, but the standard deviation has nothing to do with whether your data are normally distributed.
It depends on your field and what you are assessing, descriptive statistics alone may not be sufficient for a PhD thesis
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology