Psychic ability test scores — what does 68% on Zener cards actually indicate
I took one of the standardized psychic ability assessments online last week and scored 68% on the ESP card recognition portion — Zener cards, 25 trials. From what I've read, chance performance is 20% with 5 symbol options, so 68% seems significantly above chance. But I've done 6 sessions now and my scores range from 44% to 72%, which feels inconsistent.
The tests I'm using have 25-card trials, which I know from reading about parapsychology research isn't enough to draw statistically significant conclusions. J.B. Rhine's original research used hundreds of trials to establish significance. I'm not sure if shorter modern tests are just entertainment or if there's an actual scoring interpretation that means something.
I'm approaching this from scientific curiosity rather than belief. If anyone has experience with more rigorous testing protocols or knows what score thresholds researchers have historically used to flag results worth investigating, I'd find that useful.
Rhine's early card experiments had major methodological issues identified later — sensory leakage, recording errors, file drawer effect. Modern parapsychology research uses much stricter protocols with automated card selection and pre-registered hypotheses. The 25-card online tests are basically entertainment.
Your range of 44-72% across 6 sessions is exactly what you'd expect from chance with that sample size. The mean across all your sessions is probably very close to 20% if you calculated it out. The 68-72% sessions feel meaningful but they're within normal random fluctuation for 25 trials.
With only 25 trials, a 68% score isn't statistically significant enough to mean much. The variance on small-n tests is huge — you'd need 100+ trials at that hit rate for the result to cross standard significance thresholds. Your range of 44-72% across sessions is actually typical of random variance.
If you're genuinely curious about rigorous testing, the Rhine Research Center at Duke still runs controlled experiments and publishes results. They've been running ganzfeld experiments since the 70s and have a database of results — that's the closest thing to a legitimate protocol you'll find.