The original poster is complaining that evaluating a "Community Relations Manager" against their ability to reguritate the first half of Skiena's algorithm book (on a whiteboard, without syntax errors) is not a valid test of their ability to manage community relations.
Google's "scientific hiring process" appears to be broken , not because of any problems with the principles of testing candidates, but because the engineers who designed the tests failed to ensure that their tests were valid measures for the roles they're hiring people into.
You know the phrase "when all you've got is a hammer, everything in the world starts to look like a nail"? I think that's possibly what the engineer-hiring process has managed to create at google.
It isn't all disadvantages. There's every possibility that placing an engineer in a community-manager role could lead to new solutions to "the problem of community management".
So maybe the external evaluation of "invalid and broken" is actually a business decision and done that way by design.
The original poster is complaining that evaluating a "Community Relations Manager" against their ability to reguritate the first half of Skiena's algorithm book (on a whiteboard, without syntax errors) is not a valid test of their ability to manage community relations.
From an engineering point of view (or more accurately Psychological), their tests are failing to stand up against any measures of validity http://en.wikipedia.org/wiki/Validity_(statistics)
Google's "scientific hiring process" appears to be broken , not because of any problems with the principles of testing candidates, but because the engineers who designed the tests failed to ensure that their tests were valid measures for the roles they're hiring people into.
You know the phrase "when all you've got is a hammer, everything in the world starts to look like a nail"? I think that's possibly what the engineer-hiring process has managed to create at google.
It isn't all disadvantages. There's every possibility that placing an engineer in a community-manager role could lead to new solutions to "the problem of community management".
So maybe the external evaluation of "invalid and broken" is actually a business decision and done that way by design.