1. Okay, strictly speaking, your question just asked how organizations and individuals evaluated, it didn't specify training. And my answer to that is: outside of training, almost no-one uses Kirkpatrick. The Finance Dept, the Engineering Dept., Product Research, the Call Center--almost none of them use Kirkpatrick (whether by that name or any other) in doing evaluation.

2. Kirkpatrick (and even Phillips' critique) are still limited. You see, the arguement mostly assumes a REACTIVE approach to evaluation. That is, you do training and than you do the respective levels. And part of the reason levels 1-2-3 become important is if you're being reactive (and doing evaluation AFTER the intervention) you need levels 1-2-3 to suggest the link to any change in level 4 or even 5 (ROI).

But if you do a front-end analysis, you identify what the business goals are up front, what performance accomplishments or outcomes drive those goals, and so on. So when you introduce the solution and get movement to the goals, there is no need to do a level 1-2-3. Because frankly, most senior managers don't care if people liked or hated the program or if they learned or even if they behave differently. They want to know if they scored or moved the chains.

As for critiques of Kirkpatrick, there are hundreds. Practically any performance consultant (or anyone in the performance field) is familiar with them. I remember a session Mary Broad did at the ISPI conference probably 12-15 years ago that explicitly critiqued Kirkpatrick. Ramias has a white paper at the ISPI website on evaluation that puts Kirkpatrick to shame. The only folks that really take the Kirkpatrick model seriously (and use it) are in the training field. And that's because most trainers tend to evaluate reactively (after the fact).