Dear Committee,
Upon request of Teunis, please find below a short and clear description of the regression or ‘drivers’ analysis that will be used in the process of the survey. Globescan has also attached the drivers section from the 2006 survey.
Best regards,
Daniela
Drivers of Partnership – Regression Analysis
The survey includes a number of dimensions and attributes related to perceptions of CGIAR on various areas related to partnerships. GlobeScan will analyze the data against an overall ‘outcome’ or dependent variable (i.e., quality and successful partnerships), to determine which dimensions are most conducive to effective partnerships.
We will conduct PLS regression with the attributes and dimensions in Q8 as independent (X) variables and a combination of Q6 (satisfaction), Q7 (quality) as dependent (Y) variables. The statistics in the output called “variable importance in the projection” will be used to determine how important an attribute is in driving perceptions of successful and quality partnerships. The importance score for an attribute indicates how much of the variance in the dependent variables is explained by the individual attribute. The higher the score, the more variance it explains, and hence the more impact it has on the dependent variables; thus driving perceptions of partnerships.
We then plot the data on a matrix – with the x axis being ‘perceptions’ (i.e., the mean rating on the individual attributes), and the y axis being ‘importance (i.e., the importance score from the regression analysis). The output then identifies those areas that are strengths (higher perceptions + higher importance) and risks (relatively lower perceptions + higher importance).
Drivers Analysis – Strategic Matrix
• The matrix below provides an example from the CGIAR’ stakeholder perception survey in 2006.
• In this example, efficacy of research, responsiveness, and serving regional and global needs are those drivers which are key strengths for the CGIAR. Efficiency is noted as an area that organization should focus on.

Michele Cunningham | Associate Director, Corporate Affairs
evidence and ideas. applied
Many thanks Daniela! I look fw to receiving the revised questionnaire.
Kind greetings, Teunis.
--