Which exploratory approach for paired binary data sets, multiple items in 2 categories, patterns in one category explaining patterns in the other ?

#1
Hello everyone,

Problem : We have collected a nice set of data but we don't know for sure how to deal with it ! (Yes... not the right way to proceed indeed, but thats the situation right now...)

Context : Our structure provides help to health professionals (physicians) for situations that they consider too complex or time consuming to take on on their own. When some time consuming tasks that doesn't need their expertise can be delegated, they call us and we take care of it under their directives. (We are a non-profit organization providing a free service to health professionnals).

In this context, for each situation that is addressed to us, we take note of :

- the main cause(s) of this request for support. The first category PRO.
This is a descriptive profile of the demand from professionnals : what factors leaded them to call us. The pro side dataset contains 4 non-exclusive entries whose responses are binary (YES/NO).

- the profile of the patient’s situation. The second category PAT.
It is a descriptive profile of the patients situation : what problematics are present in his situation. The patient profile data set contains 15 non-exclusive entries whose reponses are binary (YES/NO)

Here is how the data set is made :

Situation | PRO1 | PRO2 | PRO3 | PRO4 | - | PAT1 | PAT2 | PAT3 | PAT4 | PAT5 ...etc.

36 - - - - - | YES | no | no | YES | - | no | no | YES | YES | no | ...etc.

37 - - - - - | no | no | YES | no | - | YES | no | no | no | YES | …etc.

...etc. (250 situations)


In theory, PRO patterns respond to PAT patterns :

i.e. Professionnals have specific needs and express their demand (PRO) according to the profile of the situation (PAT) they have to deal with.


My questions are :

1- Which type of exploratory analysis can be undertaken with this data set ?

2- How to identify :

- - PAT Patterns (emerging consistent clusters) and their frequencies

- - PRO patterns and their frequencies

- - a causal link between PAT and PRO items (a YES response to a PAT item leads to a YES response to a PRO item)

- - a causal link between PAT and PRO Patterns (a YES response to a set of PAT items leads to a YES response to a set of PRO items)



It’s just to indentifies regularities and frequencies, optimize our operational process accordingly for the sake of the patients and the professionnals using our service, and to provide useful feedback to our funds provider.

I’m at a loss, my statistical knowledge is very limited, and i would be infinitely grateful if you could shed some light on how to process this promising data set in the right way ! (Name of the analysis/tool , recommendations.... anything really)


GG
 
Last edited: