Example of how to do extensive classification scheme from Literature Review

This is from Dibbern et al, 2004

Pacis_4
Pacis_5

3.3.2 Identification and Classification Process


Once an article was selected, it was classified according to research focus, theoretical foundation, and methodology. To do so, we followed the general approach used by Swanson and Ramiller (1993) in their analysis of submissions to Information Systems Review (ISR). That is, we read each paper’s abstract, introduction, discussion section, and conclusion to determine the paper’s research questions or objectives. Some papers clearly stated the research questions; others did not. Consequently, this step involved some degree of interpretation on our part.


Similarly, we searched each paper for indications of its theoretical foundation. Once more we found that, while some authors were quite explicit in drawing from a particular theoretical base, others were less clear.

Identifying the theoretical foundations of these papers necessitated additional judgment and interpretation. Finally, we determined each paper's research methodology. Our intent was to place each paper in
the single category to which it most strongly, if not exclusively, belonged. Again, this required us to make some judgment calls. For instance, the McLellan et al. (1995) and Marcolin and McLellan (1998) papers are
combinations of positivist investigation and interpretive, emergent findings.8 However, since the purpose was to categorize papers according to their primary research method, we placed these papers in
the positivist category9.

Once the research objectives, theoretical foundations, and methodologies were identified, we then reviewed their reported results looking specifically for emergent patterns, trends, and/or groupings. (It should be noted that some papers reported results that covered more than one outsourcing stage. In such cases, the paper is listed more than once.)