Building a Study Plan

It’s not about knowing everything

It would be wonderful if being knowledgeable about the field that a certification covers guaranteed that you would pass the related certification exam. Personal experience: I can tell you that’s not necessarily the case (let’s just say I had a bad experience the first time I took the ASP.NET certification exam). Passing a certification exam is a goal related to, but not identical to, being knowledgeable about or experienced in the field.

You get a certification by passing a certification exam, not by excelling on the exam (though there’s nothing wrong with excelling, if that happens: I once got 93% on an SQL certification exam because I was so familiar with SQL). As far as your career goes, all anyone wants to know is “Do you have the certification?” No one really cares how well you did (though they care very much about how well you do on the job). Besides, no one can check on how well you did.

To achieve the goal of passing the certification exam, you’re going to need to decide what you need to study (a study plan). In deciding what to study, you should start by recognizing that you don’t have to know everything. A typical passing mark for a certification exam is in the 70% to 75% range. If the certification vendor (e.g. Microsoft, Amazon, Anthropic) wanted you to know everything then the pass mark would be 100%.

Which leads to some math.

May the Odds Be Ever in Your Favor

The good news: The odds actually are in your favor.

On a typical certification exam, the number of questions ranges from 45 to 70. If the exam’s passing grade is 70%, that means you need to get 32 to 49 questions right to pass. To put it another way: On a 50 question exam with a 70% passing grade, you can get 15 questions wrong and still pass.

Since, to support automatic marking, most certification exams consist of multiple-choice questions (or cleverly disguised variations on multiple choice questions), you can count on getting some questions right just by chance. Specifically, your chances of getting a question right are one-in-four to one-in-five, depending on the number of answers offered for any question. Even with questions that ask you to pick “two-of-five” (and assuming you won’t get any marks unless you pick both right answers), you’ll still have a one-in-ten chance of getting the question right.

That means that if there are 50 questions on the exam with four answers on each question, you’re going to get 12 questions right, just by chance. On that 50 question exam with a 70% pass mark, if you get 30 questions right because you knew the answer and one-quarter of the remaining 20 questions right just by chance, you’ve passed.

But those odds actually overstate the difficulty because they assume that you’re picking answers among those “remaining 20 questions” completely blind. Because you will have studied and prepared for the exam – because you know stuff – that won’t be true. For most questions whose answer you’re not sure of, you’ll probably be able to eliminate some “obviously wrong” answers. That will narrow your choice down to “It’s either B or C” or (for “pick two” questions), “It’s definitely A…and either D or E.”

Which means that your odds on the “remaining twenty questions” are actually one-in-two: You’ll get half of the “remaining 20 questions” right, even if you don’t know the answer for sure because you’ll have narrowed the candidate answers down to two options.

Going back to my 50 question example with a 70% passing mark, you need to be

• Absolutely certain of 20 of the 50 questions = 40% of the questions – less than half of the exam
• Able to eliminate obviously wrong answers on the remaining 30 questions = 60% of the questions – over half of the exam

Even that overstates the difficulty because it assumes that you do no better than chance on the “It’s either B or C” questions – you’ll probably do better than that.

As an extreme example, I recently took a certification exam consisting of multiple-choice questions with four potential answers on each question. The first 23 or so questions on that exam came out of an area I wasn’t expecting and had not included in my study plan. But, because I knew something about the field and was able to eliminate obviously wrong answers, I still got about 33% right on the questions in that area – better than the 25% I should have got by pure chance…and that was without any preparation in that area except for general knowledge of the field (and, by the way, I passed).

You should use those numbers drive your study plan. That’s my next post.

*From the e-book Certification Exams: Strategy and Tactics.