Robbie - you hypothesis is two-tailed i.e. perceived and actual care - so you would need data on both and aligning tests. As it stands you have to have two different data sets - patient's and nurses and that makes your study complex - as well as more ethically challenging. If 'burnout' is your area of interest - wouldn't it be more practical to just collect data from nurses? Correlation would be fine - depending on your collected demographic data of interest i.e. burnout and it's association with age, gender, qualification etc. Do also bear in mind that this is already a very well researched area of nursing practice.

Robbie,

I feel that adding another 2 varibles (which is not necessary if you are focusing on nurses burnout only) would make it very complicated. Please clarify..why do you need percieved patient care? Is it from nurses point of view??

Also, I think, nurses burnout is well resesrched and Isuggest you make it more specific to make it more resesrchable.

Robbie,

You can research the 5 variables. You may would want to change the design to represent this. For instance, a mixed-methods design can represent both qualitative and quantitative data with correlations between the variables. Perceptions can be done with open-end interviews such as burnout, and perceived patient care. A quantitative standpoint for patient care, nurse, and patient's standpoint can utilize surveys. Also, a qualitative study can be descriptive, interpretative, comparison, historical, explorative, case study, narrative, phenomenology, grounded theory and the list goes on etc; this depends on the specific focus you want to address. Quantitative studies used with the qualitative studies help to ground the study by triangulating the data.