CONFIRM
What to Confirm?
To confirm is to verify the accuracy of information. Decide what it is you want to confirm – if you settle on indicators (indirect evidence), instead of looking at measures of the phenomena (direct evidence), you will confirm some epiphenomena rather than “the real thing”. For instance, to confirm an increase in entropy the indicator of chaos is often used, but the real measure of entropy is a decrease in the energy’s work-potential. Entropy is NOT chaos; assuming it is leads to confusion, not confirmation.
Why to Confirm?
The purpose of confirmation is to double-check on results as reported or understood. Why a person would bother doing this is that being sure of something is important or interesting. If neither, don’t bother. There is a certain compulsion prevalent amongst a few people called “the epistemological imperative”, the impulse to get and keep verisimilitude even at extraordinary costs. This is a mild form of psychopathology.
When to Confirm?
When the issue is important or interesting, and the verisimilitude is either unknown or unstable, that is the appropriate time to confirm – this is the qualitative aspect. The length of time it will take to authorize, conduct, interpret, and report on the situation, represents the quantitative aspect. A good tactic is to scan the indicators of important or interesting situations, then switch to measures when the need for confirmation occurs.
Whether to Confirm?
Since confirmation requires time, resources, and effort, it is important to know when NOT to try to confirm: (1) when there is not enough time; (2) when there are not enough resources; (3) when it will take more effort that there is a willingness to expend. All of this refers to “formal” confirmation. There is always a “quick and dirty” version, but the result may only be an estimate – yet sometimes, even a guess may be better than nothing.
How to Confirm?
Like so: clarify focus; identify measures; acquire resources, get commitment of requisite personnel; develop a confirmation plan that includes division of labour, timelines for confirmation tasks, and a deadline for delivery; report on progress to client; stick to plan until required results are achieved or findings necessitate immediate or drastic action. In other words, conduct confirmation as research unless the need arises for re-assessment.
Reference
Howard S. Becker
TRICKS OF THE TRADE: How To Think About Your Research While You’re Doing It
University of Chicago Press, Chicago, 1998