Analysis of the Usage of Humor in Television Advertising

Advised by Dr. David Schumann

May 1

2011
Global Leadership Scholars Honor Thesis

Table of Contents
Introduction .. Phenomenon of Interest ... Importance of Topic .. Value of Research on Topic .. Research Questions .. Literature Review .. Background Information . Coding the Variables .. Explaining the Results . Method Stage One Secondary Data Collection .. Stage Two Primary Data Collection .. Week and Networks Selection .. Coding . Results and Discussion .. Humorous Ads on Television .. Humorous Ads on Humorous Shows .. Statistically Insignificant Findings .. Research Questions Recap .. Conclusion .. Beneficiaries and Application .. Limitations .. Future Research ... References .. Appendix A .. Appendix B .. Appendix C .. Appendix D .. Appendix E .. Appendix F .. Appendix G .. Appendix H .. Appendix I .. Appendix J .. 2 2 2 3 3 4 4 5 8 10 10 11 11 12 13 13 14 14 16 17 17 17 18 19 20 21 22 23 24 25 26 27 28 29

Introduction
Phenomenon of Interest Americans spend 29,000 minutes watching television per year, which includes more than 9,000 minutes of television advertising. In order to stand out in this ocean of commercials, advertisers often utilize humor to communicate with consumers. Humor is far from a lighthearted method of communication. It involves deep psychological processes that trigger a myriad of responses, both conscious and subconscious. Indeed, a successful television commercial can subconsciously inform a viewer of specific information about the product, enhance brand awareness, and reveal an unknown need in just fifteen seconds all while using the mask of humor. Such incredible results from effectively used humor beg to be further examined. This thesis analyzes actual commercial advertisements during primetime television programming and applies prior knowledge on the subject to make conclusions about the nature of humor in television advertising.

Importance of Topic This topic is important because the high costs of television advertising create a serious dilemma for companies wishing to advertise. On the one hand, the undeniable reach of millions of Americans that television advertising offers is virtually impossible to duplicate using other advertising media. On the other hand, advertising is just one of many costs that companies must consider, and primetime advertising is as expensive as it gets. The price of a 30-second ad during primetime on one of the four major networks ABC, NBC, CBS, and Fox averages $250,000 (Steinberg, 2010). That includes shows that fail to meet ratings expectations that are still able to charge high prices for 30-second ad spots, like 2

ABCs Flash Forward which did not earn a second season. One of its 30-second spots cost advertisers over $175,000 (Steinberg, 2010). For each networks top shows, prices can be double that figure. Foxs hit show American Idol commanded a price of $460,000 for 30-second spot in 2010 (Steinberg, 2010). From the results of this study that are explained more later, we know that humor is used in nearly half of all commercials on television. That means advertisers must understand what effect humor has on their messages, consumers, and ultimately, bottom-line.

Value of Research on Topic Much can be learned from researching this topic. If through close examination of the commercials and the humor processes found within them it is concluded that advertisers are indeed effectively getting their messages across and objectives accomplished, then the decisionmakers for said advertisers can be at ease with their strategies in humorous commercial advertisements. This would reduce their buyers remorse and other cognitive dissonances, lowering their mental stress that is so rampant in business today. However, if the study reveals that advertisers are in fact not using humor effectively, they can use it as a tool to make the necessary changes in their television advertising tactics. Either way, what can be learned in this study is beneficial to companies that advertise on television.

Research Questions Why is humor used so often by advertisers? Does humor affect consumers ability to recollect specific information provided by the advertisement? Are advertisers successfully communicating with consumers via television commercials?

Literature Review
The literature review is divided into three sections. The first is an overview on the past and present of humor in television advertising which provides insight into the effectiveness of todays commercials. The second is a detailed description of two variables that were used to code all of the television commercials observed in the study, humor process and humor type. The third is a detailed description of an existing study that concluded certain aspects of what works and what does not work with regard to humor in advertising.

Background Information Stephen Foxs The Mirror Makers: A History of American Advertising and Its Creators was an excellent source for the history of television advertising. He describes the commercials that dominated the airwaves in the 1960s when television was new that often ran longer than two mintues. These precursors to infomercials were able to be used because of the extremely small number of channels, especially by todays standard (Fox, 1997). However, three major technological refinements caused a significant shift in television advertising: Cable TV, VCRs, and remote controls (Fox, 1997). Together these three innovations led to zapping (changing the channel whenever a commercial begins, made possible when cable TV brought many more channel options), zipping (fast-forwarding through the commercials of pre-recorded programming, made possible by the recording feature of VCRs), and grazing (flipping through channels without watching any commercials for any significant period of time, made possible by cable TV and the remote control). As early as 1989, 20 percent of consumers responded that they paid no attention to television commercials, a figure that continues to grow (Fox, 1997).

Today, advertisers are beginning to be more strategic with their ad placements. In the magazine Advertising Age, Brian Steinberg wrote about the state of the current television advertising market. In order to eliminate the further development of zipping that digital video recorders have made possible, advertisers have started to seek live-action sports as a program to advertise within (Steinberg, 2010). Consumers rarely wait to see a sporting event after the fact so that they can skip the commercials. The same rule goes for hit shows like American Idol, NCIS, and Glee which are so engrained in popular culture that consumers want to see them in real-time, which inadvertently includes watching advertisements in real-time (Steinberg, 2010). Advertisers must realize the opportunities they have to ensure that their entire ads are not zipped or zapped, and Steinbergs article pointed them out. The prices of the 30-second television advertisements in the Introduction came from Steinbergs article in Advertising Age, as well.

Coding the Variables In 1991, James Leigh and Claude Martin, Jr. outlined the three different humor processes and the five different humor types. These definitions were published in the book, Current Issues & Research in Advertising Vol. 13, No. 1 and 2. Because the definitions that Leigh and Martin gave for humor process and humor types were used to code the primary data, a detailed description of the journal article is required. The three humor processes that humans experience are arousal-safety, incongruityresolution, and humor disparagement. Arousal-safety is felt when a person witnesses a beloved character who narrowly avoids a disaster (Leigh & Martin, 1991). An example of a commercial using arousal-safety would be one that showed a cute puppy tracking mud all over a pristine white living room. The owner comes in mad at the puppy but then uses [insert brand name

cleaning product here] to clean everything up and save the day for both him and the puppy. Arousal-safety causes viewers to worry for a moment but satisfies them when the situation is happily resolved. Incongruity-resolution is felt when a person is confused by a particular situation that is later resolved (Leigh & Martin, 1991). People, places, things, and events that are out of their normal setting will cause a viewer to feel disconnected with the situation, urging them to remain engaged so that the situation can be resolved. Once the scenario is explained fully, resolution is felt, and the process is complete. An example of this humor process in a commercial would be one that shows an in-shape man running in a marathon. His performance is dismal, stopping to catch his breath and drink water frequently, causing viewers to wonder what is wrong. Suddenly, his spirits change as he begins to run faster; he seems to have caught his stride. The screen pans down to his feet which are now inside of [insert brand name running shoes here]. The confusion that was felt at the beginning of the commercial was the disconnect and the reason for the disconnect, that we was wearing the wrong shoes, was resolved in the end. The third and final humor process is humor disparagement. This process is comprised of a joke teller, a victim, and an audience (Leigh & Martin, 1991). An important factor in humor disparagement is that the viewer must understand that whatever is going on is a joke, not intentional. If that is not made clear, the final result is not humorous according to our inner processes but cruel. An example of humor disparagement in a commercial would be one that shows a man who is grilling in huge storm. He is getting soaked but appears to being having a great time. The screen zooms in on the [brand name of the grill] and adds a catchy phrase about how no one can resist grilling out every night no matter what the circumstance are. In this case, the griller is the victim of the situation, which is understood to be a joke because of the use of an

earlier humor process, incongruity-resolution, to explain the unusual situation. It should be noted that humor disparagement can never stand alone as an experienced humor process. The five humor types coincide with the three humor processes. The first humor type, sentimental humor, occurs when arousal-safety is felt. The second humor type, comic wit, is experienced with incongruity-resolution. The third humor type, sentimental comedy, is a combination of the arousal-safety and incongruity-resolution humor processes. The fourth humor type, satire, is a combination of incongruity-resolution and humor disparagement. The fifth and final humor type, full comedy, is just as its name indicates and includes all three humor processes (Leigh & Martin, 1991). Once again, it is important to notice that no humor type is experienced with the humor disparagement process alone. To illustrate the relationships between humor process and humor type, refer to Table 1:

Table 1

Explaining the Results James Kellaris and Thomas Cline wrote a paper on the relationship between consumers need for humor and expectancy of experiencing humor on the ability of said consumers to recall specific brand claims made in an advertisement. The results of this study are central to this thesis conclusions on whether or not a humorous ad is effective. Therefore, a detailed description of the study is required. Those who are viewing advertisements are classified with either a high or low need for humor (Kellaris & Cline, 2007). A simple heuristic that can be used to determine which tag fits an individual consumer is that a younger aged consumer will have a high need for humor while an older aged consumer will have a low need for humor. This heuristic depends on the fact that different generations of television viewers use television programming for different purposes. While older generations certainly watch for enjoyment of humorous shows, they also use television for news, weather, and dramatic shows that do not employ humor. Their need for humor when watching television is low, varying in small degrees between individuals. The same cannot be said for younger generations of television viewers. They have grown up in a society where television is a tool for pure entertainment. When a younger consumer sits to watch a television program, it is with a high need for humor because that has been their environmentally learned expectation. Television viewers are also categorized between those with a high expectancy of humor and those with a low expectancy of humor. This expectancy is attained as soon as the viewer realizes what type of programming that he or she is experiencing. In the experiment, Kellaris and Cline split the test group in these two categories quite easily by creating fictitious names for

two coffee brands. One name triggered humor processes, Jubilee. The other did not, Windsor. The four test groups were exposed to advertisements for both brands. The results of the two aforementioned variables relationship with brand recall were that consumers that had a high need for humor and a high expectancy of humor had negative brand recall with the humorous brand name. Their attention was so focused on the humorous programming around the advertisement that their ability to remember specific claims made in the ad was worse than any other group. Consumers that had a high need for humor and a low expectancy of humor remembered neither better nor worse than the other groups. Consumers with a low need for humor and a high expectancy of humor mirrored the results of the latter test group. Finally, the study revealed that consumers that had both a low need for humor and a low expectancy of humor were able to make specific brand claims recollections at a significantly higher rate than the other groups (Kellaris & Cline, 2007). Refer to Table 2 for a visual on the positive, negative, and neutral effects that humor expectancy and need for humor (NFH) have on brand recall: Table 2

The study later tried to improve the brand recall of the test group that would be experiencing humor by including more specific brand claims in the ad. The amount of humor as well as the number brand claims in the ad proved to be highly detrimental to the recollection of specific brand claims. The mixture of humor and claims was concluded to have adverse results (Kellaris & Cline, 2007).

Methods
The method of the research occurred during two stages. Stage one was secondary data collection, and stage two was primary data collection. During stage two, secondary data was still collected when questions over the early findings in the results of the primary data required additional secondary research. Therefore, the stages were also simultaneous activities.

Stage One Secondary Data Collection Stage one of the methods used was collecting secondary data. First, the research questions were carefully examined, which ensured that the plethora of academic literature on humor in television advertising had boundary limitations. The first step of stage one was to find previous studies that tested what made humor useful in advertising. Diligent research and advisement led to the discovery of the experiment outlined in the Literature Review conducted by James Kellaris and Thomas Cline in 2007. This study proved that a humorous advertisement enhances recall of specific brand claims when geared toward an audience that has both a low need for humor and a low expectation of humor. Once this foundation had been established, the second step was to learn as much as possible about the psychological factors associated with humor. The reasoning for this was that stage two needed several predefined variables in which to code humorous television commercials. Such variables are not common knowledge and thus needed an academic source to support them. Through the thesis advisement of Dr. David Schumann, James Leigh and Claude Martins book was found. It also outlined in the Literature Review. It defined the three humor processes and five humor types which provided precisely the kind of coding material needed for stage two of the method.

10

Stage Two Primary Data Collection Armed with the knowledge of what makes humor effective in advertising, the process of collecting and coding the primary data began. The objective was to record primetime television, 8:00 PM eastern time to 10:00 PM eastern time, for a full week beginning Monday and ending Friday. In order to be able to make legitimate conclusions at the end of the study, however, it was necessary to find a week in which there were no unusual programming schedules, such as the Christmas season, Valentines Day, and the college basketball tournament in March. All three of these were actual examples of events that had to be avoided because of the types of commercials that were specifically designed just for them and not for a general audience.

Week and Networks Selection A week was found in the early part of March and taping began at precisely 7:55 PM eastern time on Monday, March 14. The planned and executed order of taping was ABC on Monday (The Bachelor), CBS on Tuesday (NCIS), Fox on Wednesday (American Idol), NBC on Thursday (The Office), and WGN on Friday (How I Met Your Mother). Each of those nights corresponded with a major show for the respective network, which meant the kind of commercials that would be included in the shows breaks were top of the line, perfect for the study. One tape was used for each day so as not to accidently record over previous data. Once the taping was finished on Friday, March 18, a company that converts VHS tapes to DVDs was contacted and hired for conversion. The purpose for the conversion was to make analyzing rewinding, fast-forwarding, pausing the commercials faster and easier.

11

Coding Of the entire thesis, coding, a subsection of stage two of the method, was by far the most time consuming. It was, though, exciting and vital to the conclusions that were later reached. The commercials were watched and analyzed on a laptop computer, allowing for actual coding to take place on Microsoft Excel at the same time that the commercials were being watched. An effort to watch and code the commercials without stopping or slowing down the speed was important because that is the way the television viewers observe commercials on the fly. At the beginning of each commercial, the company and product that the advertisement was promoting were noted. During the ad, basic observations on whether or not humor and/or specific brand claims were made were indicated on a yes or no basis. If any level of humor was detected, the coding recognized the ad as humorous. If more than one unrelated claim was made in the ad, or the ads sole purpose was to make one specific claim, the coding recognized the ad as one that made specific brand claims. Once the commercial had ended, a quick glance at the time elapsed was needed to note how long the commercial was. Also at the end of the commercial, the humor process(es) and humor type were coded. When the commercial breaks ended, the data was looked over to make sure that everything made sense. The length of the show was noted, as was the time of the night using thirty minute blocks of 8:00 PM, 8:30 PM, 9:00 PM, and 9:30 PM. An example of Mondays spreadsheet can be found in Appendix A. A keen eye and strong knowledge about humorous ads, brand claims, humor processes, and humor types was necessary, but there was some level of subjectivity involved. While it could be argued that the allowance for any ambiguity is a weakness, the thesis benefitted greatly from it. Commercials are based on perception and because a human was coding the ads and not a computer filled with formulae, a realistic collection of primary data was obtained.

12

Results and Discussion

Humorous Ads on Television It was found that 44.9 percent of all television commercials have humor in them. Specific brand claims are in 59.6 of commercials, but only 10.7 percent of television commercials have both humor and specific brand claims in them. This relationship was highly statistically significant with a p-value of .000 (Appendix D). Furthermore, bivariate correlation analysis revealed that there is a statistically significant negative correlation between commercials that have humor and commercials that have specific brand claims. The p-value of the correlation was .000, and the correlation coefficient was -0.657 (Appendix E). The statistical analysis above is encouraging in answering whether or not advertisers are successfully using humor in advertising, one of the research questions. It has been previously stated that combining humor with specific brand claims has adverse effects of the ability of the consumer to recall the brand that are being advertised (Kellaris & Cline, 2007). Therefore, that a mere 10.7 percent of television commercials employ both humor and claims concludes that advertisers are indeed aware of this negative reaction and do not overload consumers too with much information.

13

Humorous Ads in Humorous Shows Humorous television shows have 57.2 percent of all humorous commercials during their programming. Humorous shows also have 64.4 percent of all commercials that make specific brand claims. Both of these values remain statistically suggestive, rather than significant, due to their respective p-values of .126 and .123. Already proven is the fact that very few ads have both humor and specific brand claims, which supports the notion that there are a lot of commercials during humorous shows. Indeed, analysis indicates that 63.1 percent of all commercials on television occur during humorous programming (Appendix F and Appendix G). Logically, television audiences that tune into humorous programming have a high expectancy of experiencing humor. Consumers that have a high expectancy of humor are neutral to humorous advertisements at best (Kellaris & Cline, 2007). At worst, their ability to recall anything in the ad itself decreases because they also have a high need for humor (Kellaris & Cline). It is safe to assume that a significant portion of an audience that is watching a humorous show either tunes in with a high need for humor or develops it over the course of the program. In the latter case, advertisers should plan to stack humorous ads near the beginning of humorous shows before consumers become oversaturated with the humor emotion and the advertisers precious dollars go to waste (Kellaris & Cline, 2007 and Steinberg, 2010). As it stands, they are not doing this, and that is a mistake that is costing them money, which counters the previous sections conclusion that advertisers effectively use humor to communicate with consumers.

Statistically Insignificant Findings The first commercial of the commercial break is not statistically proven to include humor. With a p-value of .720, there is evidence that whether or not humor is used in the first

14

commercial of the commercial break is random. The time of day when the programming takes place has no relation on the usage of humor with a p-value of .682. The length of the show had no bearing on the presence of humor in the commercials, a p-value equal to .689 (Appendix H, Appendix I, and Appendix J). The statistically insignificant findings are still able to answer many questions about the humor phenomenon. The first commercial of a commercial break is the most expensive of the entire break (Steinberg, 2010). Because of this, advertisers must be certain that they meet their objectives for the advertisements and use their money wisely. An effectively constructed television commercial that links brand claims to relevant humor is the best way to do this (Kellaris & Cline, 2007). However, advertisers are not using humor on at a statistically significant level, which means that they are looking at advertising from an internal standpoint of controlling expenses and maximizes the specific brand claims that reach consumers. Humor is a complicated process that involves many different areas of the brain (Leigh & Martin, 1991). An unexplored opportunity exists for advertisers to formulate mini-campaigns inside of one, long program. Humor builds on itself through the five humor processes naturally, allowing advertisers to progress through the course of an hour or two hours, depending on the program (Leigh & Martin, 1991). The statistical analysis proves that advertisers are not considering the length of the show when determining whether or not to use humor, which is a missed opportunity.

15

Research Questions Recap Why is humor used so often by advertisers? Humor breaks the monotony of advertising that consumers experience in todays world. It discourages zipping, zapping, and grazing because it involves viewers in an active process when correctly used. Humor helps consumers remember specific claims in the ads that advertisers want to make sure that consumers know, like the price of a discounted car or a special sale going on at the local grocery store.

Does humor affect consumers ability to recollect specific brand claims provided by the advertisement? Yes. Humor taps into subconscious parts of the brain that aid in memory retrieval. After all, a memory means nothing if your brain is unable to retrieve it. Humor helps specific brand claims stand out because of their unusual usage of one or more humor processes.

Are advertisers successfully communicating with consumers via television commercials? The most important of all the research questions is still up for debate. In a general overview of television advertising, advertisers are following basic rules. This thesis proves that they avoid including too many specific brand claims with humor, but they miss out on the finer details of humor. A high percentage of humorous ads, 57.7 percent, are found embedded in humorous programming, which offers evidence that many advertisers are not successfully communicating with consumers. Advertisers have the opportunity to be much more successful in where, how, and when they use humorous television commercials.

16

Conclusion
Beneficiaries and Applications The stakeholders that benefit from this thesis are companies looking to advertise on television. Advertisers directly benefit from the discoveries made by the coding and interpretation of their commercials because they can use the results to support why they should cut humorous advertising campaigns that are scheduled to go on during humorous programs that are targeted toward a young consumer audience, increase spending on humorous commercials during the first half of non-humorous shows geared toward an older consumer audience, and eliminate all campaigns that mix specific brand claims with humor, especially irrelevant humor. The value of this thesis lies in its pioneering nature. Much has been written and studied about the effects of humor from an academic standpoint, just as much has been written and studied about the cost of advertising from a corporate standpoint. This study brings the two sides together by taking literature from both parties, combining it with primary data, and determining what makes humor work best for all.

Limitations Though 401 commercials is no small amount, the law of averages always suggests a higher sample size. If commercials were taped over the course of two weeks, six months, or even a year an even clearer and more accurate representation of the humorous television commercial phenomenon could have been reached. The goal of the taping procedure was to avoid major events that were unusual in programming, but one cannot deny that such events exist. Atypical television programming certainly uses humor that could have contributed to this thesis had it not been intentionally left out. The taping of the commercials used to conduct this study was over a

17

five-day period in mid-March. The amount, type, and frequency of humor used during this period changes on a week-to-week basis, limiting the widespread acceptance of all the results.

Future Research To tie up the loose ends discussed in the Limitations section, a longer schedule of taping is recommended so we can come closer to the true statistical results. Also, a study that consisted only of television commercials embedded within major television events would be highly fascinating. The results from such a study would be interesting to test against the results from this study to try to find any correlations. An in-house experiment using the results found here would be an excellent way to test the validity of this thesis. A similar study as this one could be conducted every five years to find detect any society shifts that have occurred over time in how we perceive and respond to humor. Finally, an identical study as this one that used a different advertising media radio, internet, mobile phones would be an excellent addition to this thesis as an all-encompassing study on humor in all advertising media.

18

References
Fox, Stephen. The Mirror Makers: A History of American Advertising and Its Creators. Champaign, IL: University of Illinois Press, 1997. vii-xxvii. Print.

Kellaris, James, and Thomas Cline. "Humor and Ad Memorability: On the Contributions of Humor Expectancy, Relevancy, and Need for Humor." Psychology and Marketing 24.6 (2007): 497-507. Web. 1 Oct 2010.

Leigh, James, and Claude Martin. Current Issues & Research in Advertising. 13. 1991. 1-35. Print.

Steinberg, Brian. "Simon Who? 'Idol' Spots Still Priciest in Prime Time: Ad Age's Exclusive Annual Survey of Broadcast TV Ad Costs Has Fox Sitting Pretty With Five of Top 10 Shows." Advertising Age 18 Oct 2010: n. pag. Web. 11 Nov 2010.

19

Appendix A
ABC The Bachelor First Break Time 8:00 PM 8:19 PM 1 2 3 4 5 6 7 8 9 10 Second Break 8:23 PM 8:32 PM 1 2 3 4 5 6 7 8 9 10 11 Length 0:30 0:15 0:15 0:15 0:15 0:15 0:15 0:30 0:30 0:30 4:00 Length 0:30 0:15 0:30 0:15 0:15 0:30 0:30 0:15 0:30 0:15 4:00 Product Humor? KIA Optima1 Y Limitless N Subway1 Y Clean&Clear Y Alaway N Sprite N How Do You N Know Claritin Y McDonald's Y Comcast xFinity1 Y Claims? N Y N N Y Y Y N N N Process IR IR AS Type CW CW SH

AS IR/HD IR Process IR/HD AS AS AS/IR/HD AS IR/HD

SH S CW Type S SH SH FC SH S

Product Humor? Claims? Avon N Y Target1 Y N Chevy Y N Zyrtec Y N Angel SoftY N Pure Clean Yshampoo Y Source Code N Y Target2 Y N Shark TankN Y Body of Proof N Y

8:36 PM

20

Appendix B
Product As is Length 10 15 30 45 60 75 90 Humor 0 = No 1 = Yes Time of ad 8:00 8:30 9:00 9:30 Length of show Humor (in show? min) 30 0 = No 60 1 = Yes 120

First ad of commercial break? 0 = No 1 = Yes

Claims? 0 = No 1 = Yes

Process Type 0 = None 0 = None 1 = AS 1 = SH 2 = IR 2 = CW 3 = AS/IR 3 = SC 4 = IR/HD 4 = S 5 = AS/IR/HD 5 = FC

21

Appendix C
Product Length Time Humor show? First? Humor Claims? Process Type KIA Optima1 0:30 8:00 0 1 1 0 2 Limitless 0:15 8:00 0 0 0 1 0 Subway1 0:15 8:00 0 0 1 0 2 Clean&Clear 0:15 8:00 0 0 1 0 1 Alaway 0:15 8:00 0 0 0 1 0 Sprite 0:15 8:00 0 0 0 1 0 How Do You Know 0:15 8:00 0 0 0 1 0 Claritin 0:30 8:00 0 0 1 0 1 McDonald's 0:30 8:00 0 0 1 0 4 Comcast xFinity1 0:30 8:00 0 0 1 0 2 Avon 0:30 8:30 0 1 0 1 0 Target1 0:15 8:30 0 0 1 0 4 Chevy 0:30 8:30 0 0 1 0 1 Zyrtec 0:15 8:30 0 0 1 0 1 Angel Soft 0:15 8:30 0 0 1 0 5 Pure Clean shampoo 0:30 8:30 0 0 1 1 1 Source Code 0:30 8:30 0 0 0 1 0 Target2 0:15 8:30 0 0 1 0 4 Shark Tank 0:30 8:30 0 0 0 1 0 Body of Proof 0:15 8:30 0 0 0 1 0

2 0 2 1 0 0 0 1 4 2 0 4 1 1 5 1 0 4 0 0

22

Appendix D

Humor Cumulative Frequency Valid 0 1 Total 221 180 401 Percent 55.1 44.9 100.0 Valid Percent 55.1 44.9 100.0 Percent 55.1 100.0

Claims Cumulative Frequency Valid 0 1 Total 162 239 401 Percent 40.4 59.6 100.0 Valid Percent 40.4 59.6 100.0 Percent 40.4 100.0

23

Appendix E

Correlations Humor Humor Pearson Correlation Sig. (2-tailed) N Claims Pearson Correlation Sig. (2-tailed) N 401 -.657 .000 401 401
**

Claims -.657 .000 401 1

**. Correlation is significant at the 0.01 level (2-tailed).

24

Appendix F

HumorShow * Humor Crosstabulation Humor 0 HumorShow 0 Count Expected Count % within HumorShow % within Humor % of Total 1 Count Expected Count % within HumorShow % within Humor % of Total Total Count Expected Count % within HumorShow % within Humor % of Total 78 85.4 50.3% 35.3% 19.5% 143 135.6 58.1% 64.7% 35.7% 221 221.0 55.1% 100.0% 55.1% 1 77 69.6 49.7% 42.8% 19.2% 103 110.4 41.9% 57.2% 25.7% 180 180.0 44.9% 100.0% 44.9% Total 155 155.0 100.0% 38.7% 38.7% 246 246.0 100.0% 61.3% 61.3% 401 401.0 100.0% 100.0% 100.0%

Chi-Square Tests Asymp. Sig. (2Value Pearson Chi-Square Continuity Correction Likelihood Ratio Fisher's Exact Test Linear-by-Linear Association N of Valid Cases 2.337 401 1 .126
b

Exact Sig. (2sided)

Exact Sig. (1sided)

df
a

sided) 1 1 1 .126 .153 .126

2.343

2.038 2.340

.149

.077

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 69.58. b. Computed only for a 2x2 table

25

Appendix G

HumorShow * Claims Crosstabulation Claims 0 HumorShow 0 Count Expected Count % within HumorShow % within Claims % of Total 1 Count Expected Count % within HumorShow % within Claims % of Total Total Count Expected Count % within HumorShow % within Claims % of Total 70 62.6 45.2% 43.2% 17.5% 92 99.4 37.4% 56.8% 22.9% 162 162.0 40.4% 100.0% 40.4% 1 85 92.4 54.8% 35.6% 21.2% 154 146.6 62.6% 64.4% 38.4% 239 239.0 59.6% 100.0% 59.6% Total 155 155.0 100.0% 38.7% 38.7% 246 246.0 100.0% 61.3% 61.3% 401 401.0 100.0% 100.0% 100.0%

Chi-Square Tests Asymp. Sig. (2Value Pearson Chi-Square Continuity Correction Likelihood Ratio Fisher's Exact Test Linear-by-Linear Association N of Valid Cases 2.374 401 1 .123
b

Exact Sig. (2sided)

Exact Sig. (1sided)

df
a

sided) 1 1 1 .123 .150 .124

2.380

2.068 2.372

.143

.075

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 62.62. b. Computed only for a 2x2 table

26

Appendix H

Humor * FirstAd Crosstabulation FirstAd 0 Humor 0 Count Expected Count % within Humor % within FirstAd % of Total 1 Count Expected Count % within Humor % within FirstAd % of Total Total Count Expected Count % within Humor % within FirstAd % of Total 193 191.8 87.3% 55.5% 48.1% 155 156.2 86.1% 44.5% 38.7% 348 348.0 86.8% 100.0% 86.8% 1 28 29.2 12.7% 52.8% 7.0% 25 23.8 13.9% 47.2% 6.2% 53 53.0 13.2% 100.0% 13.2% Total 221 221.0 100.0% 55.1% 55.1% 180 180.0 100.0% 44.9% 44.9% 401 401.0 100.0% 100.0% 100.0%

Chi-Square Tests Asymp. Sig. (2Value Pearson Chi-Square Continuity Correction Likelihood Ratio Fisher's Exact Test Linear-by-Linear Association N of Valid Cases .128 401 1 .720
b

Exact Sig. (2sided)

Exact Sig. (1sided)

df
a

sided) 1 1 1 .720 .833 .720

.129

.044 .128

.768

.415

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 23.79. b. Computed only for a 2x2 table

27

Appendix I

Humor * TimeofDay Crosstabulation TimeofDay 800 Humor 0 Count Expected Count % within Humor % within TimeofDay % of Total 1 Count Expected Count % within Humor % within TimeofDay % of Total Total Count Expected Count % within Humor % within TimeofDay % of Total 49 47.4 22.2% 57.0% 12.2% 37 38.6 20.6% 43.0% 9.2% 86 86.0 21.4% 100.0% 21.4% 830 58 62.3 26.2% 51.3% 14.5% 55 50.7 30.6% 48.7% 13.7% 113 113.0 28.2% 100.0% 28.2% 900 56 57.3 25.3% 53.8% 14.0% 48 46.7 26.7% 46.2% 12.0% 104 104.0 25.9% 100.0% 25.9% 930 58 54.0 26.2% 59.2% 14.5% 40 44.0 22.2% 40.8% 10.0% 98 98.0 24.4% 100.0% 24.4% Total 221 221.0 100.0% 55.1% 55.1% 180 180.0 100.0% 44.9% 44.9% 401 401.0 100.0% 100.0% 100.0%

Chi-Square Tests Asymp. Sig. (2Value Pearson Chi-Square Likelihood Ratio Linear-by-Linear Association N of Valid Cases 1.499
a

df 3 3 1

sided) .682 .682 .611

1.501 .259 401

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 38.60.

28

Appendix J

Humor * LengthofShow Crosstabulation LengthofShow 30 Humor 0 Count Expected Count % within Humor % within LengthofShow % of Total 1 Count Expected Count % within Humor % within LengthofShow % of Total Total Count Expected Count % within Humor % within LengthofShow % of Total 95 93.7 43.0% 55.9% 23.7% 75 76.3 41.7% 44.1% 18.7% 170 170.0 42.4% 100.0% 42.4% 60 38 41.3 17.2% 50.7% 9.5% 37 33.7 20.6% 49.3% 9.2% 75 75.0 18.7% 100.0% 18.7% 120 88 86.0 39.8% 56.4% 21.9% 68 70.0 37.8% 43.6% 17.0% 156 156.0 38.9% 100.0% 38.9% Total 221 221.0 100.0% 55.1% 55.1% 180 180.0 100.0% 44.9% 44.9% 401 401.0 100.0% 100.0% 100.0%

Chi-Square Tests Asymp. Sig. (2Value Pearson Chi-Square Likelihood Ratio Linear-by-Linear Association N of Valid Cases .746
a

df 2 2 1

sided) .689 .690 .840

.744 .041 401

a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 33.67.

29