13th July 2009 at 11:55 pm #5842
I am doctoral fellow interested in knowing about researchers who have good background of implementing mixed method designs in mental health. Especially those how are also interested in establishing mental health services keeping in view a cross cultural perspective and under served communities17th August 2009 at 4:02 am #5850DebbieParticipant
I recommend looking up Paul Bolton’s work or Kenneth Miller, they do a lot of work among underserved populations such as refugees. Paul does a lot of work in regards to culturally-valid assessments and is a great researcher.
Debbie18th August 2009 at 10:54 am #5849Mark FreestoneMember
Two very involved and culturally-aware academics in the UK are Dr Paul Godin and Professor Peter Beresford, both of whom are particularly interested in enabling service users to give their own voices and expertise to Mental Health research. I found Paul Godin’s recent mixed-method study of forensic inpatient units to be inspiring and there is even a DVD of some of the interviews available via the Forensic Mental Health Research Programme.
Personally I am quite involved in using mixed methods in small n service evaluations and happy to discuss this!
–Mark20th August 2009 at 3:27 pm #5848
Thanks Debbie. But how to contact these researchers? Are they on Methodspace? Or any website if they have?20th August 2009 at 3:42 pm #5847
Thanks for reply. Actually i am interested in seeking some consultation for my ongoing study which is actually an effectiveness study. I adapted Barkely’s manual for managing disruptive behaviors associated with ADHD for PAkistani population. And being the first study of its kind we experienced certain challenges, including the drop-out rate of almost 43.6% in our group sessions. We couldn’t do Intend To Treat Analysis because of NOn Equivalent Control Group in our study. And to handle this high drop out rate some experts in Maryland University suggested me to try mixed model approach with maximum milkelihood approach. Can i have some good references to understand the nuts and bolts of this approach, in order to have a good background for myself and to build a good rationale for my own research. Thanks. Tamkeen21st August 2009 at 6:19 am #5846DebbieParticipant
They are not on methodspace. Dr. Paul Bolton is at the Johns Hopkins school of public health. Here is the link to his faculty profile:
and Dr. Kenneth Miller is at Pomona College,
here is his faculty page http://www.psychology.pomona.edu/kenmiller/
You can look up some of their articles and also contact them via their info posted on the sites.
Debbie21st August 2009 at 12:41 pm #5845
Thanks Debbie for providing such useful kind of links. Hope its gonna be helpful for my ongoing work.
Lets stay in touch.
Tamkeen1st September 2009 at 2:21 pm #5844Mark FreestoneMember
Ah-ha! I think we might be at cross-purposes here: it sounds like you are describing a study which uses linear mixed models (i.e. applied regression with fixed and random effects) when your question was about mixed methods, which is usually understood in the UK to mean the combination of qualitative and quantitative research methods!
From what you have described above, a mixed-models approach could be fruitful given the high dropout rate but I might have concerns about bias if your sample is not large. Any more information you can give about the study would be useful.
As far as background goes, David Garson’s excellent Statnotes website might be helpful for you here. The web address for the section on Mixed Models is
–Mark2nd September 2009 at 9:10 am #5843
Thanks Mark. Infact i am working on two projects simultaneously. In first one we worked with quatitative data solely and now we are intending to use Mixed model for that(and not mixed methods). The second study is involving program evaluation and cross cultural perspectives, which is actually involving mixed methods and a blend of qualitative and quantitative approaches.
You are very right, that it got some degree of confusion and sorry about that.
So at present we are only talking about Mixed Model(as i mentioned) and not the mixed methods.
Well this study is for training parents for behavios management techniques, through group sessions. Sample size of our control group is 30, and of experimental group it is 55. This sample is asymmetric also becuase of difference of settings. We started our work from school settings and ended up at hospital settings, because of less co operation from schools for experimental group. Becuase of some other challenges which we encountered, during this study we couldn’t go for randomized control group and also experienced a high drop out rate. In order to deal with these issues i have been suggested to use Maximum likelihood approach, and i need second opnion about that.
Thanks for providing me this weblink, but the document you mentioned is not available to me on this server(right now i am in Pakistan, and webaccess is quite limited). May i request you to send me some SPSS point and click pathways and if any resource available to interpret the outputs of Linear Mixed Model. My access is quite limited from here i am not getting the clear instructions for the method from here.
Thanks and looking froward.
- The forum ‘Default Forum’ is closed to new topics and replies.