> that coders who sat at 12-person lunch tables tended to outperform those who regularly sat at four-person tables... >...driving more than a 10% difference in

> that coders who sat at 12-person lunch tables tended to outperform those who regularly sat at four-person tables... >...driving more than a 10% difference in performances Yeah, so this is exactly the sloppy analysis that makes these policies so counter productive: there's a pile of data, it is interpreted so poorly as to be random yet instills an absolutely unreasonable level of confidence. It certainly does not compensate for the loss in productivity that such demoralizing practices create.