Brain dumps are the biggest threat to the certification industry these days, significantly devaluing certifications that rely primarily on multiple-choice answers. Similarly to the threat-prevention measures adopted by airport security (read the insightful analysis of their behavior from Bruce Schneier, a renowned security guru), IT vendors are responding with high-tech measures.
Some of these measures are simply brilliant; for example, attaching a photograph of the candidate to the exam printout. Others, such as exam data forensics, will probably stop some cheaters, but also generate a lot of false positives. However, brain dumps could be stopped in a very low-tech way: human brains have limited capacity; if you increase the size of the question pool, the brain dumps become useless.
Can you increase the size of the question pool? Absolutely. Terry Slattery has proposed a method by which even a security breach by one of the participants would not compromise the majority of the question pool.
Wouldn’t that cost a lot of money? That’s a relative question. Cisco recently announced that it has issued over a million certifications. The cheapest certification you can get is the CCNA composite exam at $250, and individual exam tests cost upward of $100. If only a small percentage of the certification-related revenue would be reinvested in writing more questions, brain dumps wouldn’t be a problem anymore.
Can you even write that many questions? Once you can’t write new questions, you’ve obviously exhausted the contents of the certification materials, and those materials (training classes, books etc.) become functionally equivalent to the brain dumps. Anyone who can correctly answer all the questions covered by the certification material is by definition certifiable, regardless of how those correct answers were acquired. Obviously, it helps if the questions are written in a way that tests the understanding of the subject matter, not the memorization of useless facts.