How can I prove hetero/homogeneity
Hi guys, I'm writing from the view of a very junior medical author and my current paper discusses the effectiveness of a teaching technique. It's been suggested I test for heterogeneity of outcome between groups. I have no statistical knowledge and don't even know what raw data I would need. below is the relevant data. How would I go about doing this?
Thanks from a guy who doesn't even know where to start, James
|Teaching about communication as a manager, advocate or leader? 90% effective |
|Teaching team capacity and positive working relationships? 91% effective |
|Demonstrating team roles including leadership? 90% effective |
|Allowing you to link availability of resources to the need to use resources efficiently? 86% effective |
|Demonstrating appraisal, assessment and review of colleagues? 66% effective |
Re: How can I prove hetero/homogeneity
The first thing (and probably the most important first steps) is to figure out what data will quantify the attributes you need to answer your question.
So you will need to quantify effectiveness in all categories.
You will have to quantify leadership attributes, resource effeciency, relationships, team capacity and so on.
So as a first question for you, do you have ways to quantify these things in the context of your question? If not, then you will need to supply more information on these attributes so the right metrics can be chosen.
Even if you choose great metrics, understand that they may unreliable and the process from which they are obtained might carry bias.
But this is the very first step you need to consider and we can take it from here.