1. Method flaw -- The survey cited here - Manta Q2SB - provided options of Extremely important - Important - Indifferent - Not important - No value. Lumina's options were very important - somewhat important - Not very important - Not at all important. There is a bushel of research indicating that you get vastly different responses when you have a neutral category (indifferent) vs. no neutral category. In short, an apples and oranges comparison
2. Sampling - Manta's survey was a general survey of their members. Lumina/Gallup a randomized national sample. Convenience sample vs. scientific sampling approach used to provide a good representation of the population of interest. Again, apples and oranges.
3. Sampling - even better. Of the millions of companies registered with Manta, they managed to snag under 1000 responses to their questionnaire. Brutally low response rate.
4. "Success in today's economy" was not the question -- it was about "success in the working world' - more importantly, the other questions in the survey had to do with who they had hired and what sort of degrees they held. The question was clear in the context of whether they felt there staff needed a college degree. Seeing that many positions hired for at small businesses are clerical or customer service related, it seems reasonable to suggest you don't need a college degree in those roles. That doesn't say much about whether you need a degree for higher levels of accomplishment.
5. Tons of small business people are contractors. Although there are many complicated aspects to these positions and many skills involved, these certainly are not the sort of things learned on a college campus - so I expect whatever proportion of the sample reflected these businesses likely go an answer of "not important" - again, duh.
