The number of competing sensory inputs are increasing at a rapid rate in today’s world. Researchers have long been interested in the underlying mechanisms that enable humans to selectively attend to relevant information while simultaneously disregarding irrelevant information. These mechanisms are critical to safeguard against sensory overload.

Priming has been an avenue through which to study these mechanisms to help discover precisely how the brain is able to direct attention to information that is pertinent to the current task or situation. The size of the contribution of memory in this process has been a long-standing debate.

This dissertation presents a series of experiments that explore the underlying cognitive mechanisms in long-term attended repetition positive priming effects and discusses these findings through an abstractionist perspective, an episodic perspective, and a predictive brain account. The paradigm used presents participants with a pair of items, one being an uppercase distractor word and the other being a lowercase target word, followed by another pair of such items with the lowercase item being either a word or a non-word. Participants were instructed to ignore the uppercase words and respond to the lowercase target items. The prime display required participants to read aloud the lowercase target word. In the following probe display, participants were asked to make a lexical decision (word/non-word judgment) to the lowercase target item. The relationship between the prime and probe targets was manipulated to produce two conditions. In the attended repetition (AR) condition the lowercase target word that was read aloud in the prime display was identical to the lowercase prime target word that appeared in the probe display. In the neutral control condition, the words presented in the prime display had no relationship to the words in the probe display.

Six experiments are presented that were designed to explore AR effects at different lags involving a different number of intervening prime-probe couplets (trials). In all of the experiments, words appeared only once except to satisfy the AR conditions. Only in the AR conditions did words appear exactly twice, once as the prime target and subsequently as the probe target. For example, in a lag 10 AR condition the lowercase probe target word had appeared as the lowercase prime target word exactly 10 trials prior. Notably, one of the six experiments included a longer lag (80) combined with a lag 0 AR condition. The results demonstrated the importance of the inclusion of the lag 0 AR condition. That experiment did not produce long-term AR positive priming, consistent with findings reported by McLennan, Neumann, and Russell (2019, Experiment 2), which included a lag 144 condition combined with a lag 0 condition. However, the remaining experiments with long lags up to lag 144 did produce AR positive priming without a lag 0 condition. Apparently, the presence of a lag 0 AR condition abolishes long lag facilitatory priming effects. This finding cannot be easily accounted for by the two major theories of positive priming; the abstractionist account, and the episodic account. It is proposed that the absence of long-term positive priming in the presence of a lag 0 AR condition is the result of an unconscious inhibitory mechanism. This inhibitory processing mechanism is discussed in the context of a statistical learning process and is supported by past research involving a predictive brain hypothesis. Overall, the results suggest that facilitatory priming effects are mediated by both bottom-up input and top-down processing and, depending on the task context, the brain subliminally forms a perceptual hypothesis that can induce an inhibitory mechanism that modulates future responding.