I have research project on comparing the effectiveness of two different class of organization (named 1 and 2) through a collected data from clients of that organizations.
1. Data collected for 50 variables including demographic variables e.g age, sex etc.
2. Two type of variables included Nominal and likert scale items about satisfaction of services provided by organizations to their clients.
Q1. Which statistical analysis would be appropriate for such data?
Q2. I have 2 hypothesis which statistical test should i apply on them?
(Hypothesis 1. one organization is good services than other)
(Hypothesis 2. both organization not effective in their local context)
My personal statistical study suggest me MANOVA is it right?
MANOVA is rarely the answer. It mashes up all your variables into a composite variable and then carries out a statistical test that is pretty much uninterpretable. I've yet to see a real-life question tested appropriately by MANOVA.
You should focus on a definition of 'good service'. What is the key measurement? Ideally, this should be a real-life indicator like time to recovery, rather than a 'soft' measure like rated satisfaction (which can be influenced by things like courtesy bias).
What are seconary indicators or performance?
You may have to face the reality that there is no single measure that defines a good service, and, instead, you will be comparing and contrasting the services in terms of a number of outcomes.
If you recruited participants from multiple centres of each type, you should stratify by centre, or adjust for clustering of participants within centres.
No it is not. Given Ronán Michael Conroy advice you should focus on your DV if it is Likert see the discussion here:
https://www.researchgate.net/post/Can_you_help_with_statistical_analysis_advice
And Here, esp if your variable is binary(eg Categorical DV)
https://www.google.com/search?q=What+type+of+regression+should+I+choose&rlz=1C1CHBF_enUS874US874&oq=What+type+of+regression+should+I+choose&aqs=chrome..69i57.20795j0j1&sourceid=chrome&ie=UTF-8
Your research question(s) determine the statistical model. For a guide, you might look at:
https://www.google.com/search?q=Montgomery+Design+and+analysis+of+experiments&rlz=1C1CHBF_enUS874US874&oq=Montgomery+Design+and+analysis+of+experiments&aqs=chrome..69i57.30438j0j1&sourceid=chrome&ie=UTF-8
And PLEASE PLEASE PLEASE next time do all of this prior to DATA COLLECTION,
REMEMBER;:PROPER PRIOR PLANING PREVENTS POOR PERFORMANCE.
See the attached as well. Good Luck, D. Booth
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude