Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Peter Kelly • • • • • • Introduction: Who I am and what I did Philosophy of impact evaluations How to do a good evaluation Collecting data Ethical issues Additional resources Philosophy of impact evaluations • Relationship to accountability – Why needed – Clean but fails – Corrupt but succeeds • About the weakest aspect – Data quality – External validity • 250-point checklist How to do a good evaluation • Think of yourself as an investor in a similar project—what do you want to know? • Collect data yourself, know the language • Ask respondents your research question directly—when? • Adjust theory to reality, not vice versa, but beware of data mining Collecting data • No time like the present • Don’t take no for an answer (advisers, bureaucrats, collaborators) • Offer something to collaborators, make it part of a larger project • Opinion on grants and conflicts of interest Ethical issues • Getting the right answer • Research staff usually in much more danger than subjects Additional resources • Validity checklist • Safety for staff members