Running head: MEMORY FOR INFORMATION PAIRED WITH HUMOR

The present study examines the effects of humor and relevance on memory for factual statements. Participants (N=48) read pairs of statements. Each pair included a fact that was always nonhumorous, and a joke that was either humorous or nonhumorous and relevant or irrelevant to its paired fact. In a later memory test, recall of pairs (i.e. both fact and joke) was better when the joke was humorous vs. nonhumorous and relevant vs. irrelevant. Further, the effect of humor on recall was greater when the joke was relevant than when it was not. Results suggest that when instructors present facts with jokes, jokes may cue memory for facts on later tests, and that relevant jokes may provide better and more memorable cues than irrelevant jokes. Running head: MEMORY FOR INFORMATION PAIRED WITH HUMOR 3 Memory for Information Paired with Humorous, Relevant Jokes Many professors and teachers sometimes insert humorous comments or stories into their lectures, which results in various possible effects on students. Importantly, humor may affect memory for the material being taught. The effect could be positive: humor may enhance memory for the material it is paired with in class through heightened attention and cognitive engagement (e.g., Speck, 1991), favorable affect (e.g., Speck, 1991), or increased arousal (e.g., McGhee, 1983). Alternatively, the effect could be negative: humor may be distracting and thus impair memory for accompanying information (e.g., Lammers, Leibowitz, Seymour, & Hennessey, 1983). Finally, humor could have no effect on memory (e.g., Berg & Lippman, 2001). The current study examines humor’s effect on memory and the possible mediating role of relevance (i.e., the degree to which the humor is relevant to the to-be-remembered material); perhaps relevant humor boosts memory while irrelevant humor hinders it or has no effect on it. Studies on humor and memory focus on two distinct potential effects: (1) humorous information itself may be remembered better than non-humorous information, and (2) humor may aid memory for nonhumorous information that is presented together with the humor (concurrently, immediately following, or immediately preceding the humor; to the best of my knowledge no research to date has compared the effect produced by these different options). Results from previous studies provide evidence for the first effect. In a study by Schmidt (1994), each participant read a list of sentences, half of which were humorous and half were nonhumorous. Subsequent tests of free-recall and cued-recall found that participants were significantly more likely to remember the humorous sentences than the nonhumorous sentences. The same effect was found for the Running head: MEMORY FOR INFORMATION PAIRED WITH HUMOR 4 proportion of words recalled from each sentence. Similarly, a study conducted by Schmidt and Williams (2001) presented participants with three types of cartoons: humorous (in which the cartoon and the caption were incongruous in a way intended to be humorous; for example, the cartoon showed a woman calling her dog as it runs toward a door with a nailed-shut dog-door, and the caption reads “Here Fifi! C’mon! ... Faster, Fifi!”), literal (in which the cartoon and the caption were not incongruous, and therefore not humorous; in the literal version of the example above, evidence that the dog-door is nailed shut is removed) and “weird” (in which the caption and the cartoon were incongruous in a way not intended to be humorous; in the “weird” version of the example above, the dog is replaced with a snake). The purpose of the “weird” cartoon type was to test the hypothesis that humor enhances memory only because it contains incongruous information. A recall test (administered after a short filler task) revealed that memory was better for humorous cartoons than for both literal and “weird” cartoons. A later study replicated the results (Schmidt, 2002). Finally, Kintsch and Bates (1977) conducted an experiment investigating the effect of humor in the context of education; participants attended a lecture and were later given a recognition test for various statements from the lecture. Compared to topic statements and details, extraneous remarks – which included jokes and other remarks – were better remembered. Evidence for the second type of effect of humor (humor’s enhancement of memory for other information presented with the humor) is less consistent. Much of the research on the topic has focused on humor in the context of advertising (an effort that has led to several literature reviews: Duncan 1979; Madden & Weinberger 1984; Speck 1987; Sternthal & Craig 1973; Weinberger & Gulas 1992; and two meta analyses: Running head: MEMORY FOR INFORMATION PAIRED WITH HUMOR 5 Berneman, Bellavance, & Jabri, unpublished manuscript; Eisend 2009). The typical experimental design included varying the humor level of advertisements and testing memory for ad information other than the humor (most commonly brand name, selling points, and/or product slogan). Results have been mixed. Many studies have found a positive effect of humor on memory (Chung & Zhao, 2003, Duncan, Nelson, & Frontczak, 1984, Furnham, Gunter, & Walsh, 1998, and Yong & Zinkhan, 1991), while others have failed to find an effect (Berg & Lippman, 2001, Duncan, & Nelson, 1985, Norris & Colman, 1994, Sutherland & Middleton, 1983, Sutherland & Sethu, 1987, and Wu, Crocker, & Rogers, 1989), and still others have found a negative effect (Cantor & Venus, 1980, and Lammers, Leibowitz, Seymour, & Hennessey, 1983). Outside of the advertising domain, two studies found a positive effect of humor on memory for accompanying information. Zillmann and colleagues (1980) presented kindergarten and first grade students with an educational television program that either did or did not have humorous segments (unrelated to the program’s content) interspersed through it. Memory for information from the educational television program with humorous segments was better compared with memory for information from the same educational television program lacking such humorous segments. Researchers concluded that humor increased attentiveness and thus led to better memory. Further, McAninch, and Austin (1993) showed that recall of nonsensical line drawings was better when the drawing was accompanied by a humorous caption than when it was accompanied by captions that were either physically descriptive, abstract, or meaningful, or when there was no caption. Based on these findings, McAninch and Austin conclude that humor aids memory by organizing ambiguous information in a coherent, meaningful way, although it is not clear Running head: MEMORY FOR INFORMATION PAIRED WITH HUMOR 6 why humorous captions would be more effective in organizing information than descriptive or meaningful captions. Relevance may be an important factor in humor’s effect on memory. If humor aids memory when it is relevant, instructors could use humorous, as opposed to nonhumorous, illustrating examples to enhance students’ memory for the material. If humor is distracting when it is irrelevant, instructors may want to avoid telling funny stories that are unrelated to the course’s educational material. The possible role of relevance in humor’s effect on memory has been largely unexamined; only a few studies have focused on the topic. In an experiment by Kaplan and Pascoe (1977), classes of university students viewed one of four lectures: a nonhumorous lecture, a lecture with humorous examples that were related to concepts in the lecture, a lecture with humor unrelated to lecture concepts, or a lecture with a combination of these two types of humor. Kaplan and Pascoe found that in a test given 6 weeks later participants who viewed a lecture with humorous examples illustrating concepts (i.e., relevant humor; either exclusively relevant humor or combined with irrelevant humor) did better on questions about these concepts compared with participates who viewed a lecture with humorous comments that were unrelated to lecture concepts (i.e., irrelevant humor) or a lecture with no humorous comments. The results from this study suggest a positive effect of relevant humor on memory for accompanying information, but they should be interpreted with caution. First, there was no “relevant-nonhumorous” condition in which participants viewed a lecture with relevant examples that were nonhumorous. The absence of such a condition makes it impossible to know what role, if any, humor played in the memory boost. It is possible that simply adding relevant information, humorous or Running head: MEMORY FOR INFORMATION PAIRED WITH HUMOR 7 not, is sufficient to produce an enhancement of memory. Further, the study used at least some humorous examples about sex and blood, making it difficult to disentangle the effect of humor from the effect of high-arousal topics that are not necessarily humorous but might affect memory because they serve as highly memorable retrieval cues or because they increase arousal and therefore improve encoding (Walker & Tarte, 1963). In an advertising study, Krishnan and Chakravarti (2003) studied the effect of humor strength and relevance on memory for brand claims. In one experiment to test only the effect of humor strength, they varied the humor strength of ad’s headlines, so that some participants saw very funny headlines, some moderately funny headlines, and some low-humor headlines. In this experiment humor was irrelevant to the brand claims the ad made. Researchers tested participants’ memory (recall and recognition) for the brand claims of these ads, and found that memory was better when humor was moderate compared to when humor was low or high. They concluded that a low degree of humor was insufficient to aid memory, a moderate degree of humor did aid memory, and a high degree of humor was distracting and thus cancelled out the positive effect of humor. These results suggest that humor at a moderate level can enhance memory even when it is irrelevant, although it is difficult to draw conclusions without a nonhumorous condition as a point of comparison. In a related experiment, Krishnan and Chakravarti (2003) tested if the effect of high strength humor on memory could be made more positive by increasing humor’s relevance. They therefore varied humor’s relevance to the brand claims – either low or high relevance – while keeping humor strength high. Memory for brand claims was better when humor was relevant to brand claims than when it was irrelevant. As is the case with Kaplan and Pascoe’s study, this experiment lacked a Running head: MEMORY FOR INFORMATION PAIRED WITH HUMOR 8 relevant-nonhumorous condition, which leaves open the possibility that the effect is independent of humor. In other words, the results do not allow us to conclude that humor’s effect on memory is mediated or altered by its relevance; the same results could potentially have been obtained without the humor component if relevance alone is driving the memory enhancement. Finally, Cline and Kellaris (2007) conducted a study looking at the effects of humor strength and relevance on memory for ad claims. Participants viewed an advertisement for a coffee product with either a highor low-strength one liner joke as a headline. Additionally, the ad contained a tag line (i.e., a brand or product slogan) and product claims, which either referred to the one-liner (high-relevance) or did not (lowrelevance). On a recall test for ad claims, there was a main effect of relevance; ad claims with a high-relevance one-liner were better remembered than ad claims with a lowrelevance one-liner. There was no significant main effect of humor strength. The study did not include a relevant-nonhumorous condition; it is possible that humor’s presence can boost memory for relevant information compared with humor’s absence, but that increasing humor’s strength from low to high does not enhance memory further. There was also an interaction of relevance and humor strength: relevance’s positive impact on recall for ad claims was stronger when humor’s strength was high than when it was low. If humor enhances relevance’s effect, as suggested by these results, then a hypothetical nonhumorous condition should show a weaker relevance effect than the low-strength condition. The study’s results support the enhancing or mediating role of relevance in humor’s effect on memory for ad claims. The two studies described above (Krishnan and Chakravarti, 2003; Cline and Running head: MEMORY FOR INFORMATION PAIRED WITH HUMOR 9 Kellaris, 2007) looked at humor and relevance in the advertising domain. In advertising, the to-be-remembered information is a claim made by a company whose goal is to convince consumers to purchase a product or service. Consumers are not necessarily motivated to remember the ad claims or accept them as true. The current study focuses on humor and relevance in education, where the to-be-remembered information is factual and is communicated by instructors whose goal is to transmit knowledge. Students are presumably motivated to remember the factual information. Moreover, consumers may attend to humor because it grabs their attention, but be unmotivated to attend to product claims, which are less attention grabbing. Students, on the other hand, presumably attend mostly to the educational information because their main goal in attending lecture is to comprehend and retain educational information. Thus, humor’s effect on memory for facts in an educational context, in which there is motivation to attend to remember facts, might be different from its effect in the context of advertising, where there is little or no motivation to attend to and remember product claims. For example, individuals viewing advertisements may focus on the humor and make no effort to attend to and remember product claims. Thus, humor may have no effect or an attenuated effect on memory for product claims, as they were weakly encoded to begin with. In contrast, in an educational setting students are motivated to attend to and remember the educational material, and will likely make an effort to encode it. In this case, relevant humor may have a significant effect on memory for educational material. The current study explores the effect of humor and the effect of relevance on memory for factual information, as well as the interaction between them. If adding relevant information helps elaborate memory for the fact by providing more association Running head: MEMORY FOR INFORMATION PAIRED WITH HUMOR 10 links to the fact than irrelevant information (see Craik & Tulving, 1975 for support of the positive effect of elaboration on memory), then there should be a main effect of relevance; it was hypothesized that relevance would enhance memory for accompanying factual information across humor conditions. There are three possible effects humor could have on memory for accompanying information. If humor aids memory independently of its relevance to the to-beremembered information (for example, if humor increases arousal and thereby enhances retention of any information encountered immediately after), there should be a positive main effect of humor. In accordance with results of previous studies described above (Cline & Kellaris, 2007; Kaplan & Pascoe, 1977; Krishnan & Chakravarti, 2003), there might be an interaction between humor’s and relevance’s effects on memory for accompanying factual information, so that humor would enhance memory more when it is relevant than when it is irrelevant to the factual information. If humor aids memory only through its relevance to the accompanying information, then humor’s main effect would be qualified by this interaction. Alternatively, if humor in fact distracts from the information it is paired with regardless of its relevance, as several researchers have suggested (e.g. Berneman, Bellavance, & Jabri, unpublished manuscript; Kaplan & Pascoe, 1977; Krishnan & Chakravarti, 2003; Osterhouse & Brock, 1970; Zillmann et al., 1980), then there should be a negative main effect of humor.