In choosing a mixed-method approach I would be guided by the nature of the research problem as well as resource availability and constraints  (funding, maximum length of funding, access to participants, etc). The research design depends largely on:

You may want to refer to the following for some good guide on mixed methods research:

I hope this helps. All the best in your quest.

I agree with Francis.

Triangulation method would give you different but complementary data for your research.You may want to check on 'Concurrent Triangulation Design'  which involves concurrent but separate collection and analysis of quantitative and qualitative data

Embedded design on the other hand would be suitable based on the premise that one data set is not sufficient for the study.You may want to consider the relevance of embedded design if you have different questions that need to be answered and where each question requires different type of data.

very useful information thanks a lot.

yes, read this:

http://www.sagepub.com/upm-data/31987_Punch_Final_Proof.pdf

http://www.antle.iat.sfu.ca/courses/iat834/resources/Creswell&Clark_Chap4&5.pdf

Mixed method research involves not only techniques but the researchers ability to switch frameworks in their minds, particularly if you are using qualitative and quantitative approaches.  Generally quantitative approaches involve the use of a predetermined framework of theory.  You want to describe, compare or contrast or test something believed to be true.  It is generally a deductive approach. 

In qualitative methods you are going in without any preconceived ideas of the person's reaction to an experience or event.  In your case you are assuming care taking of an HIV person is a burden and requires coping on the part of the caregiver.  In a true qualitative study, you would not make those assumptions.  You would need to go broader and only assume the caregiver would have an emotional reaction to care giving.  Phenomenological interviewing is good because you can ask about the experience of care giving without any direction implied.  I don't know for example parents taking care of children with HIV would classify their care giving as a burden.

What is an amazing approach is to do a qualitative approach that is inductive and open minded and use the data from that study to develop a tool that then can be used in a quantitative study.  The School of Nursing at the University of Arizona in Tucson is well known for tool development and that has been a recommended approach.

Good luck.  Definitely a worthwhile study.

thank you for your kindness..will go through with it.

I found this book very helpful to understand mixed methods for beginner... 

Thank you. 

It might be helpful to take a step back and consider your research questions. Your questions should lead you in a particular direction with regard to the choice of research design.

If, for example, you're interested in using qualitative methods to explore the reasons underlying the quantitative results or to enrich interpretation of quantitative results, then you're naturally led towards using an explanatory sequential design. If your qualitative strand is intended to explore the relevant issues and feed into the design of your quantitative instrument, you're naturally led towards using an exploratory sequential design. If you're mainly aiming to see whether and how the qualitative and quantitative findings converge or diverge from one another, then you might appropriately use a concurrent triangulation design. This is not an exhaustive list, though; it's well worth spending some time considering your research questions alongside the Creswell & Plano Clark text, Designing and Conducting Mixed Methods Research, to see what design is best aligned with your questions and aims.