Mixed Methods

Mixed methods combines qualitative and quantitative analysis. In this combination the advantages of one method help counter the limitations of the other method. 

Example

Multiple analysis methods for a mixed method study:

Click on diagram for a larger PDF view.

Example - Chad McAllister’s [Prospectus Version]

By using exploratory mixed-methodology, the first phase of research will create two lists of factors that influence misunderstanding requirements.  The lists will be an aggregate of the work produced by three to four pairs of small groups using NGT.  To produce the aggregated lists, the definitions of each factor will be compared and similarly defined factors will be consolidated.

The second phase of research will result in weighted lists of factors, indicating the importance of each factor as perceived by users versus developers.  Each participant will individually weight the factors.  AHP or another appropriate technique will be used to create a combined weight across all participants.

Differences between users' and developers' perceptions of factors influencing the misunderstanding of requirements will be analyzed in five ways:

1. Identifying factors identified by users but omitted by developers.

2. Identifying factors identified by developers but omitted by users.

3. Consistency of weightings assigned by users and those by developers using Kendall's Coefficient of Concordance (Havelka et al., 1998a).

4. Consistency in weighting critical factors between users and developers using the Wilks' lambda test (Havelka et al., 1998a).

5. For the critically ranked factors, a thematic analysis will be performed of the definitions to identify similarities and differences between users and developers.

Tables and bar charts will be used to convey the findings in 1 through 4 above.  Results of the thematic analysis will be described textually and supported with figures highlighting relationships. 

john latham (c) 2000 - 2012  all rights reserved