All 120 subjects completed the full experiment. Hence there were sixty full sets of data for the neutral and the heightened attention conditions.
Memory Test and Manipulation Check
First, in the spontaneous memory test, subjects correctly remembered on average 10 and 12 prime words in the forward and backwards priming tasks. The second memory test, i.e. the manipulation checks was in line with instructions: subjects in the attend condition recognized on average 13 and 15 words in the recognition tests after the forward and retroactive priming tasks, respectively. Subjects in the neutral instructions recognized only about 4 and 6 words in the recognition memory after the proactive and retroactive priming tasks, respectively, a difference which was significant, at p<0.05.
However, the present study failed to replicate the retroactive priming effect. To reiterate, in accordance with Bem’s original reports, we expected to find a backwards priming effect, i.e. a facilitation of processing of probes depending on the subsequently shown primes; and secondly, in accordance with earlier priming literature, we expected that this retroactive priming effect would be more pronounced when subjects were prompted to focus on the prime words. However, these predictions were not fulfilled.
Neither in the neutral instructions condition, nor in the attend condition could we detect any evidence suggestive of a retrocausal effect of prime words on probe evaluations. Our results rather show that there was no significant facilitation effect of subsequent prime exposure in the neutral condition (t(119)=0.89; p=0.207), or in the attend condition (t(119)=1.02; p=0.155). Though results move in the right direction, including a somewhat more pronounced trend in the expected direction in the attend instructions condition, they fail to reach significance in either condition.