Back in 1927 the Western Electric executives responsible for managing their "Hawthorne" plant near Chicago agreed to participate in a research study conducted by a Harvard Business School professor. At the time, very little was known about what factors had the greatest effect on industrial productivity, so it seemed like a good idea to set the consultants loose.
Interviews were conducted, environmental factors were changed and a host of labor, management and psychological issues were discussed and examined. Virtually everyone at the plant was involved in some way or another. It must have been quite exciting to have this brilliant group from a hot-shot northeastern university showing such interest in the making of telephone cable!
Anyhow, one of the recommendations was to brighten the place up by turning on more lights. Plant management complied and sure enough production increased. Now here's where it gets interesting... After a few months and more scholarly attention, the lights were turned back down. What happened to production? It went up again!
The plant guys couldn't figure out why; the executives couldn't figure out why. Even the research team was stumped, so they packed up all their data and went home to sort it out and come up with an answer. To make a long story short, they never did successfully uncover any statistical connection between changes made and productivity. (Even after all involved decided to extend the one-year project again and again until they finally wrapped it up five years later.)
So what's the moral to the story? With 20-20 hindsight and 75 more years study, it's clear:
Bring in credible "out-of-town experts" to...
...work closely with those people doing the actual work
Set a clear expectation of great success
Thoroughly document & analyze everything, both qualitatively and quantitatively
Keep at it conscientiously for an extended period of time