The Effects of Priming on Spread of Activation

This experiment looked at the relationship between word priming and spread of activation. It was hypothesized that primed words would be easier to recall in a later task than would non-primed words. This was tested by presenting participants with crossword-like clues and blank spaces indicating the number of letters in the word. There were three experimental conditions for priming: no priming, studying a list for one minute, and rating words on their importance to the subject. Data showed that both priming conditions slightly increased reaction time; this trend was opposite of our hypothesis. Further investigation is suggested with more subjects, another priming condition, and a varied intertrial break.

   Spreading activation is a process by which some aspect of memory is activated and this activation spreads through other areas in the mind to ready them for processing. Anderson (1976) and Collins and Loftus (1975) present a memory model composed of an "interlinked network of nodes" (Ratcliff & McKoon, 1981, p. 454) which stand for concepts. Triggering of the memory at one point spreads through this network to retrieve related information that is connected by links. This theory proposes that those memories that have been activated more or more recently are easier to access (Anderson, 1995).
   Priming is a method that can bring about this activation. According to the theory, prior presentation of a word or concept could trigger that memory and make it more accessible. Loftus (1973) found that subjects could more easily name an item in a category if it was the second such item they were prompted to recall; however, the more extraneous items that came between these prompts, the slower the reaction times became. Research by Ratcliff and McKoon (1981) showed that reaction times to target words primed with closely associated words were faster than target words primed with distantly associated words.
   These studies provide the basis for our current research. We were interested in spreading activation as connected to the doing of crossword puzzles. These require searching one's memory for the answers to clues that are not necessarily directly connected to the desired word's memory nodes. Often puzzles are easier to solve if one has more experience in doing them. Clues are often "recycled," so familiarization can give a person a certain advantage. This leads to this experiment's hypothesis that words primed in the first part of the experiment will facilitate reaction times for those same words in the third part of the experiment. The priming should activate these words and make them more accessible to further processing.

Methods
   Subjects were 21 students from Alma College. 14 were female, and 7 were male. The experiment consisted of three basic parts: priming, an intertrial break, and presentation of the test stimuli. The answers for the test stimuli were generated from two different sources. Crossword puzzles from local papers were given to five people who did not do crossword puzzles regularly, and the clues that the most people answered correctly were considered for use. The remaining words and clues were generated by the researchers. This list was tested on five volunteers. Words not guessed correctly by a significant number of the subjects were replaced with more accessible clue-word pairings.
   In the experiment, clues to the desired words were presented using the SuperLab computer program. All subjects were shown the same test stimuli in the same order. To the right of the clues were spaces indicating how many letters were in the word. The intertrial break consisted of the subjects putting together a computerized jigsaw puzzle for two minutes.
   There were three different priming conditions. Each of the conditions had seven randomly assigned subjects. In the non-primed condition, subjects were not primed with any words prior to the experiment. In the list-reading condition, subjects were shown a list of words for one minute. Half of the words were used in the third part of the experiment and the other half were not associated at all with the rest of the experiment. Participants were instructed to read through the list and familiarize themselves with it, but they were not told to memorize anything. In the rating condition, subjects were primed with the same list of words, but this time they were asked to rate each word on a scale from 1 to 7 as to its importance to them, 1 being the least and 7 being the most.
   Subjects were instructed to respond to the test stimuli in part three as quickly as possible. They were told that the first answer they gave would be the response accepted by the experimenter. The experimenter controlled the progression of the clues with the keyboard. At the frst response vocalization of the participant, the experimenter pressed a key which stopped the timing device in the computer