It's been pointed out to me that some preliminary results from this study are available:
"The following information reflects live study sessions. The study currently has 6582 erotic trials gathered from a total of 366 participants. There has been 6 (0.09%) missing data points due to incomplete sessions. We observed a total of 49% successful guesses within 6582 erotic trials (99.5% CI = 47.3%, 50.7%; posterior mode = 49.8%, posterior 90% HDI = 48.8%, 50.6%). Observing this success rate is 57 times more likely if humans are guessing randomly than if they can guess future randomly determined events. Taken at face value, the data provide strong evidence that the probability of successfully guessing later computer-generated random events is not higher than chance level as previously reported by Bem (2011) and others (Bem, Tressoldi, Rabeyron, & Duggan, 2015). The results proved to be robust to different statistical approaches, increasing our confidence in our inference. None of the stopping rules have been triggered yet, so data collection is still in progress. The next crucial test will be at reaching 37836 erotic trials."
http://178.128.174.226:3838/TPP_follow_results/
This is a replication of Bem's Experiment 1 - "Precognitive Detection of Erotic Stimuli" - in which the subjects had to guess which of two curtains on a computer screen had a picture behind it. The position of the picture was randomly determined after the subject had guessed. Bem found a hit rate of 53.1% for erotic pictures, where 50% would have been expected by chance.
Given the number of participants, a hit rate of 49.0% does provide very strong evidence against the presence of an effect of the size seen by Bem.
"The following information reflects live study sessions. The study currently has 6582 erotic trials gathered from a total of 366 participants. There has been 6 (0.09%) missing data points due to incomplete sessions. We observed a total of 49% successful guesses within 6582 erotic trials (99.5% CI = 47.3%, 50.7%; posterior mode = 49.8%, posterior 90% HDI = 48.8%, 50.6%). Observing this success rate is 57 times more likely if humans are guessing randomly than if they can guess future randomly determined events. Taken at face value, the data provide strong evidence that the probability of successfully guessing later computer-generated random events is not higher than chance level as previously reported by Bem (2011) and others (Bem, Tressoldi, Rabeyron, & Duggan, 2015). The results proved to be robust to different statistical approaches, increasing our confidence in our inference. None of the stopping rules have been triggered yet, so data collection is still in progress. The next crucial test will be at reaching 37836 erotic trials."
http://178.128.174.226:3838/TPP_follow_results/
This is a replication of Bem's Experiment 1 - "Precognitive Detection of Erotic Stimuli" - in which the subjects had to guess which of two curtains on a computer screen had a picture behind it. The position of the picture was randomly determined after the subject had guessed. Bem found a hit rate of 53.1% for erotic pictures, where 50% would have been expected by chance.
Given the number of participants, a hit rate of 49.0% does provide very strong evidence against the presence of an effect of the size seen by Bem.
![[-]](https://psiencequest.net/forums/images/collapse.png)