Practice Scores Tracked the Margin, Not the Pass: 36 Reported EPPP Results

Three members passed the EPPP this month without ever reaching 80% on one of our full-length practice exams.

That matters because we used to describe 83% as the number every passer we could measure had reached. We retracted that on September 11 when the first exception arrived. There are now three, and their real exam scores point to a better way to read a practice score.

The limits come first. These are self-reported results from a small group. They are not a pass rate, and nothing here shows that a study habit caused an outcome.

The count as of September 19

Twenty-three members have told us they passed. Thirteen have told us they did not. That is 36 reports, including one anonymous, count-only pass from a member who deleted their account data.

Available information	Passed	Did not pass
All reported outcomes	23	13
Exam date on the report and a full-length practice exam before it	12	1
Exam date on the report, no full-length practice exam before it	1	4
No exam date on the report	9	8
Anonymous report after account deletion	1	0

We count a practice attempt as full-length when it has at least 150 questions. We only use attempts dated before the real exam. Same-day attempts are excluded because we do not know the exam time.

One change from the last review. Four passers left the exam date blank on their report but had one saved in their profile, dated on or before the day they reported. We now include them, and we label them wherever they change a number. That brings the measured passers from 12 to 16. We could not do the same for members who did not pass. Every profile date in that group falls after the report itself, so it is the retake date, not the exam that was failed. Practice after a failed attempt says nothing about preparation before it.

What the three passers under 80% scored

Their best full-length practice scores before test day were 72%, 77%, and 78%. All three passed. Their real scaled scores fell between 510 and 555. Most licensing boards set the pass mark at 500.

Now the other thirteen. All reached 83% or higher before test day, and eleven shared a real score.

Best full-length practice score before the exam	Passers with a real score	Real scaled scores
Under 80%	3	510 to 555
83% to 86%	5	570 to 630
88% to 89%	3	538 to 600
92% to 93%	3	661 to 711

Nine of those eleven scored 594 or higher. Two did not: one in the 530s and one in the 570s. Both of those come from the profile-date group. Without them, every measured passer at 83% or above scored 594 or higher. With them, the pattern is still there, but it is looser. We are showing you the looser version because it is the honest one.

So here is the reading we can defend. Among passers, the practice score did not separate who passed from who did not. It leaned toward how much room they had. The low 70s went with a pass by a small margin. The low 90s went with the highest scores anyone has reported to us. The middle is noisy.

What this cannot tell you

It cannot tell you that 72% is enough. We only see the people at 72% who passed and then wrote to us. Twelve of the thirteen members who did not pass have no full-length practice exam that we can place before their exam. Most of them joined after it. So we have no idea how many people sit at 75% and then fail, and we will not guess.

It also cannot rescue the one hard exception. One member reached about 91% and did not pass. We are not going to explain that away, and it was not a lack of effort: this member logged lessons, quizzes, and ten full-length practice exams with us.

Two details are worth writing down. First, nine of those ten exams were in "Learn as you go" mode, where you see each explanation right after you answer. The 91% was one of them. The one exam in "Measure where you stand" mode, where nothing is shown until the end, came in at 83%. We do not know yet whether learn-as-you-go scores run high. Among the seven passers who used both modes, the learn-as-you-go best was lower than the other in four. Second, the 91% came about ten weeks before the exam, every full-length score after it was lower, and the last one was more than two weeks before test day.

Neither detail is a rule. Four passers also used only the learn-as-you-go mode. Ten of the sixteen measured passers took a full-length practice exam in their final week, but one took their last one seven weeks out and scored well above the line. These are questions to keep watching.

A late drop in practice scores is a different story again, and it has its own post.

How to use your next practice score

Read it as a margin check, not a verdict.

If you are in the low 70s close to your exam, passing is possible, but your cushion is thin. One rough section can decide it. Spend your remaining days on your two weakest domains rather than spreading across all eight. The last-week study plan covers what to do with those days.

If you are in the mid 80s, protect what you have. Open your last practice exam and choose three questions you missed or guessed correctly. For each one, explain why your answer seemed plausible, what you overlooked, and why the correct answer fits. Then sleep.

In either case, try to take one full-length practice exam in your final week, in "Measure where you stand" mode. Ask yourself which answers came from understanding and which came from recognizing a question you had already seen. A score built on recognition will not travel to test day.

Keep the reports coming, either way

Several members had exam dates this week. If you have taken the real exam, you can tell us how it went without agreeing to have your name, score, or comments published. Permission to use a testimonial is a separate choice.

The reports that break our earlier advice are the ones that fix it. This post exists because three people told us they passed with numbers we had said were too low.