Page 23 - Recommendations-for-biostatisticians-in-managing-and-conducting-medical-research-consultations
P. 23

4.2 Conceptual framework


               Conceptual framework (CF) is a visualization technique to portray the association of the
               independent variables and moderator variables with the dependent variable. It should not
               be  confused  with  a  theoretical  framework  (TF).  Although  TF  may  look  similar  to  CF,
               however the key variables of TF are based on theory where the relationship or association
               has already been proven valid. Unlike TF, CF is a hypothesized relationship or association
               that has yet to be tested by research (Rocco and Plakhotnik 2009). This explains why CF is
               a  visualization  technique  commonly  used  by  students  when  they  are  writing  up  their
               dissertations.  Occasionally,  this  technique  is  also  being  applied  in  the  preparation  of
               scientific manuscripts.

                   Previously, TEV provides a broader picture of a disease event among patients with a
               similar disease while CF describes a specific hypothesis intended to be tested within a TEV
               model. TEV has previously been used to describe the sequence of events  that commonly
               occur in the T2DM patients from an early stage until all the clinical outcomes have been
               developed.  An  example  of  CF  is  shown  in  Figure  2,  which  shows  that  a  researcher  is
               interested  to  identify  the  risk  factors  for  poor  control  of  HbA1c  level;  and  this  diagram
               shows how TEV and CF are inter-related and also depicts the hypothesized relationship or
               association between possible risk factor(s) and disease outcome(s).





























               Figure  2:  The  connection  between  Time  Event  Visualization  (TEV)  and  Conceptual
               Framework  (CF)  which  delineates  the  association  between  possible  risk  factor(s)  and
               disease outcome(s)

                   Based on a list of all the independent variables, four of them are being investigated in
               this TEV model; for example: the subjects’ demographic profiles, their clinical parameters,
   18   19   20   21   22   23   24   25   26   27   28