Exploring the effect of subliminal single-word and multiple-word primes on working memory performance

This PhD thesis focused on subliminal priming, that is, the presentation of information below conscious awareness (Vernon, 2009), which has been shown to influence both cognitive and affective behaviours. Information can be presented subliminally using both ‘Single-Word’ and ‘Multiple-Word’ written primes, although the two prime types have not yet been compared. Currently there is no reported optimal procedure for the presentation of subliminal stimuli, thus such a comparison could guide future research concerning prime choice. Hence, this thesis empirically compared the effects produced by Single-Word and Multiple-Word primes in a series of experiments. In Experiment 1 96 participants were subliminally stimulated with one of six alternative primes, three Single-Word primes (cognitive: intelligent; affective: one; neutral-control: walking), and three Multiple-Word primes (cognitive: I am intelligent; affective: mommy and I are one; neutral-control: people are walking), and their performance measured on a range of cognitive (e.g., working memory, intelligence, selective attention) andaffective (e.g., mood and state anxiety) tasks. Results of Experiment 1 showed no clear change in participants’ intelligence, selective attention, mood, or state anxiety. However, post hoc analysis found participants’ significantly improved their working memory performance following exposure to all positive (e.g., cognitive and affective) subliminal primes, regardless of prime type (i.e., Single-Word and Multiple-Word). Experiment 2 followed this up by exploring the effect of subliminal priming on working memory performance. Sixty participants were primed with one of the six subliminal stimuli to assess whether the non-differential effect between prime types found in Experiment 1 was due to the varied length of time between the end of subliminal exposure and the onset of the task. Results found all participants’ performance improved regardless of prime type and prime content and thus was concluded to reflect a practice effect. Experiment 3 considered that the absence of any subliminal priming may have been due to participants’ potential lack of motivation to attain the goal of improving working memory. Hence, 106 participants were primed with one of the six subliminal stimuli and their motivation to achieve this goal was enhanced using falsepositive feedback on performance and reading a false article extract highlighting the benefit of a good working memory. Results found, despite increased motivation to improve working memory, that subliminal priming did not have any effect on performance. Experiment 4 considered whether the specific content of the subliminal stimuli, the short prime-target stimulus onset asynchrony (SOA), or the type of task could be accountable for the null results. Thus, in addition to enhancing participant motivation, the content of the stimuli were refined to become more task-relevant, the prime-target SOA was extended from 14ms to 514ms to allow more time for unconscious processing. Eighty-three participants were primed with one of four subliminal stimuli; two Single-Word primes (memory-specific: remember; neutralcontrol: walking) and two Multiple-Word primes (memory-specific: I can remember well; neutral-control: people are walking), and performance was measured using two working memory tasks. Results found all participants’ performance improved on both working memory tasks regardless of prime type and prime content and was concluded to reflect a practice effect. Finally, a meta-analysis conducted on the data from the Conceptual Span Task from all four experiments confirmed an improvement on performance over time but no evidence of any subliminal priming effects. Overall, this thesis found it was not possible to establish a difference between the two prime types, although findings indicate that subliminal priming may not be able to improve performance of the phonological loop component of working memory