Academic references to outcomes theory: Duignan (2009d; 2009a; 2008d; 2008b; 2004a)*
Outcomes theory rules about how the building-blocks relate to each other
Building-blocks two and three (indicators) are complementary - you need both typesA comprehensive outcomes system will measure both not-necessarily controllable and controllable indicators.
Building-block one (outcomes model) is essential for all the others because it identifies your prioritiesThis means that you can ensure that all the other building-blocks are tightly aligned to these priorities.
Often controllable indicators (two) do not reach to the top of the outcomes model (one)
In this instance, you cannot make any impact attribution statement attributing changes in high-level outcomes to a program only on the basis that you've measured improvements in a not-necessarily controllable indicator (in two).
Five (impact evaluation) should only be done if you have already optimized your program with four (non-impact evaluation)Usually it only makes sense to do impact evaluation on a program which has the greatest possible chance of success. There is little point in finding out that a program does not work when this is likely to have been caused by the program not being implemented properly.
In the previous situation, the only way you can make an impact attribution statement is using building-block five (impact evaluation)Impact evaluation uses more one-off evaluation designs rather than routine indicator data collection which is used in two and three.
For evidence-based cost-effectiveness & cost-benefit evaluation (six) you need effect sizes from five (impact evaluation)
Only some of the impact evaluation designs in four are able to provide these effect-size estimates. Economic analyses should clearly identify the basis of the effect-size estimates which are being used.
Building-block seven (accountability) should focus on all of the other building-blocks rather than just two or three (indicators)Some accountability, performance management and contracting systems focus exclusively on bulding-blocks two or three. Focusing on three exclusively is usually a mistake (because these indicators are not-necessarily controllable by the program) and just focusing on two usually restricts accountability to a low output-type level.
* These particular references can be cited to refer to the material on this page.