Linkedin

Evaluation

There is a multitude of approaches to evaluation.

Similarly, the dimensions of evaluation tasks vary radically, from small local tasks to multi-million dollar national (or trans-national) investigations.

Our specialisation is on smaller evaluations (or components of larger ones) with a focus on quantitative analysis.

Some quantitative techniques for evaluation

Data mining, tracking and presentation: Often enough, presentation of a complex set of data in the right way, without explicit analysis, is sufficient to be of great help to decision-makers.

Cost-benefit analysis: A simple concept of comparing costs with benefits is made problematical because of the difficulty in defining what is included in costs and how to measure the benefits.

Statistical inference: The discipline of statistics can extract patterns and inferences from data. The danger is that incorrect statistical tests are used or that a user continues to deploy statistical tests until one gives the desired result. What is needed is not so much statistical significance as causal significance.

Sensitivity analysis: Sensitivity analysis can test whether variations in possibly uncertain input data would affect the final result. This adds to the robustness of the analysis.

Previous clients (of the director)

  • Department of the Environment
  • Department of Defence
  • Department of Immigration

Performance audits and work on performance measures (see relevant tabs) are disciplines that are closely allied to evaluation.