Possible Effects of Originality 1 Running Head: EFFECTS OF ORIGINALITY Possible Effects of Originality and Number of Valid Responses on Cryptomnesia XXXXX

The purpose of this study is to investigate the effects of the originality and number of valid responses on the level of cryptomnesia (unintentional plagiarism). Participants will be asked to offer examples of real words from categories of varying size, and to create fictional words. To adjust the number of possible responses for the fictional word categories, the participants will be asked to arrange anagrams using a varying number of letters. Participants will be asked to generate words, remember their responses, and then generate new words. Possible Effects of Originality 3 Possible Effects of Originality and Number of Valid Responses on Cryptomnesia Cryptomnesia (the act of inadvertent plagiarism) has been investigated through many different experimental procedures. Brown and Murphy (1989) created a procedure in which participants took turns listing words from specific categories (ex. four legged animals, sports). Cryptomnesia was measured across three phases, the initial generation phase, a phase where participants were asked to recall the words they had given, and a phase where participants took turns generating new words, by how many times participants claimed words as their own that had either been said by another or (for the generate new stage) had been said by themselves in a previous stage. Brown and Murphy (1989) not only found that cryptomnesia was common to most participants, but they also discovered several factors that affected the level of cryptomnesia, including oral versus written tasks (with written tasks producing more cryptomnesia) and orthographic versus semantic tasks (with orthographic tasks producing more cryptomnesia). This suggests that the nature of the task can influence the level of cryptomnesia, but there are other factors that can also have an effect. Brown and Halliday (1991) (using a similar word generation task) found that participants were significantly more likely to unintentionally plagiarize the participants that gave words directly before them than the participants that gave words directly after them (interestingly, Linna and G ̧lgˆz (1994) found that randomizing the sequence of the participants' responses did not affect the level of cryptomnesia). Brown and Halliday (1991) theorized that the effect they found was due to participants thinking of a word in anticipation of their next turn only to have the person before them say that word. They therefore had more difficulty in remembering who had generated the words because the words had already been in their minds before the other Possible Effects of Originality 4 participants said them. Because Brown and Halliday's (1991) theory revolved around source monitoring difficulties, many other experimenters hypothesized about (Johnson, Hashtroudi, & Lindsay, 1993) and investigated the possible connection between source monitoring and cryptomnesia, but their results did not support a source monitoring model for cryptomnesia (Macrae, Bodenhausen, & Calvini, 1999; Marsh and Landau, 1995). But if source monitoring is not the key to cryptomnesia, what is? Roig (1999) found that when students were asked to paraphrase paragraphs, they plagiarized the more difficult and unfamiliar material more than the easier and more familiar material. Marsh and Bower (1993), using a puzzle task in which participants looked for words in an array of letters, found that strengthening the participant's memories (ie. giving them the recall own phase immediately instead of after a delay) decreased cryptomnesia. Furthermore, Marsh and Bower (1993) found that hard puzzles (32 or 43 valid responses) produced more cryptomnesia than easy puzzles (112 or 113 valid responses). Tenpenny, Keriazakos, Lew, and Phelan (1998) took Brown and Murphy's (1989) word generation task and added fictional word categories to investigate the effect of originality of responses on cryptomnesia. They found significantly less cryptomnesia for the fictional words, but considering Marsh and Bower's (1993) results for number of valid responses, there is a large confound in Tenpenny et al,'s (1998) study: the fictional word categories had an infinite number of valid responses while the real word categories were limited. The purpose of the present study is to investigate Tenpenny et al.'s (1998) results more thoroughly by determining to what extent the amount of originality involved in a task and the number of possible answers affect the amount of cryptomnesia. Furthermore, this study seeks to discover whether there is any interaction effect between these two variables. As with other cryptomnesia research, this study has larger implications; for example, professors and Possible Effects of Originality 5 universities that wish to tailor their assignments avoid plagiarism could use this data to help them know what to change. Method