I am conducting a survey to find out the views of sports coaches; what barriers and facilitators do they think impede or would enhance their practise etc.
the survey has 18 questions, the first 7 of those are tick-box responses related to the demographics of the respondent
the final questions are a mixture between Likert responses and ranking questions. There are also a couple of open questions “other, please state”, “is there any other information….”.
I am struggling to understand how I would analyse my survey (other than the thematic analysis of the open questions and descriptives). My tutors have suggested that other students have had interesting results using Bayesian analysis. whilst I understand the very basic principle, on the whole subjective uncertainty etc, I struggle with statistics at any level past the usual T test, ANOVA etc.
I'm wondering if anyone can give me any guidance as to where I might start?
One a similar journal does not outline how they analysed there like hurt or ranking questions other than saying they used SPSS. Another simply states that they used SPSS using chi squared test for descriptive statistics.
Hello Alex,
How to analyze a data set depends on: (a) your specific research question(s); (b) the nature of the variable(s) involved (e.g., how quantified; scale); and (c) how the data were collected. Unfortunately, I can't offer much advice without knowing all of these points, other than to say that summarizing "...the views of sports coaches..." sounds more like a descriptive study than anything else.
Yes, I've no doubt you've come across other examples that are inadequately explained (e.g., "used SPSS"); people frequently leave out key details in their write-ups, which makes the tasks of understanding what happened and how to judge it pretty difficult indeed.
The advice about Bayesian approaches is too vague to be remotely helpful as given.
If you can elaborate your query, I have no doubt that folks in the RGate community will be able to offer more focused recommendations. Barring that, I'd wager that there are competent statisticians at your institution that would be able to help you work your way through this.
Good luck with your work.
I suggest,
1. consider questions that represent your interest. Usually, 2 or 3 and test the association with demographic questions.
2. cross-sectional tables represent your work of test and estimate with p_values for both.
3. inference for categorical data available in many books, for example, "Alan Agresti"
Best!!
AN
- Badges
- Science topic
- Similar topics
- Statistics
- Survey