When development evaluation gets too honest?

For anyone who’s ever written a methodology section, treading the thin line between acknowledging limitations and completely undermining the findings….

There should be a development/RCT/M&E version of this:

Dye? Or survey sample?

…or because access was limited by the logistically challenging environment (read: muddy roads).
Too easy: donor funding, government cooperation, when the interns came in the summer? Take your pick.

More here.