Dante E. Mancini, “The “CSI Effect’ in an Actual Juror Sample: Why Crime Show Genre May Matter.”
Highly limited empirical research has supported the existence of the “CSI effect”, defined as the influence of heavy forensic television program viewership on perceptions of scientific evidence and juror decision making. Recently, cultivation theory has been proposed as a potential explanation of any television viewing influence, but virtually no prior studies have distinguished crime show sub-genres in viewership measures. Participants were 79 actual jurors who viewed a video recorded summary of a real murder trial, then completed measures of forensic television viewership and perceived realism, verdict preferences, verdict reasons, and perceptions of the presented evidence. Viewership predicted verdict preference as hypothesized, with heavier fiction viewers rendering more acquittals compared to lighter viewers. However, heavier fiction viewers were no more likely to mention DNA evidence as a verdict reason or report a greater preference for either the prosecution’s or defense’s evidence. These findings suggest that, relative sub-genre viewership may be important in establishing empirical support of the CSI effect.
Highly limited empirical research has supported the existence of the “CSI effect,” defined as the influence of heavy forensic television program viewership on perceptions of scientific evidence and juror decision- making. Recently, cultivation theory has been proposed as a potential explanation of any television viewing influence, but virtually no prior studies have distinguished crime show sub-genres (fiction and documentary-style) in viewership measures. The current study investigated whether a difference score measure of fiction-to-documentary forensic television program viewership was related to juror decision-making. Participants were 79 actual jurors who viewed a video recorded summary of a real murder trial, then completed measures of forensic television viewership and perceived realism, verdict preferences, verdict reasons, and perceptions of the presented evidence. Viewership predicted verdict preference as hypothesized, with heavier fiction viewers rendering more acquittals compared to lighter viewers. However, heavier fiction viewers were no more likely to mention DNA evidence as a verdict reason or report a greater preference for either the prosecution’s or defense’s evidence. These findings suggest that relative sub-genre viewership may be important in establishing empirical support of the CSI effect.
There is little doubt that the American public has taken a keen interest in crime investigation and forensic science over the past decade, particularly as reflected in television programming. In a recent Frontline program episode entitled “The Real CSI,” well-known forensic pathologist Dr. Cyril Wecht remarked, “There’s this great, great hunger, this incredible fascination with forensic science. I often quip that we are up there now with sex, motherhood, apple pie, and baseball” (Cediel & Bergman, 2012). This interest has likely been sparked by the enormous popularity of television programs such as CBS network’s CSI: Crime Scene Investigation and a plethora of other crime and forensic science- based programs. In June 2012, at the end of its twelfth season, CSI was named the most watched television program in the world for the fifth time in seven years, with an estimated 63 million viewers worldwide (Bibel, 2012), and clone program NCIS: Naval Criminal Investigative Service was named the most popular television program in America during the 2012-13 season, even garnering more viewers than Sunday Night Football (Patten, 2013). Furthermore, as of this writing, CSI, NCIS, and NCIS: Los Angeles can all be found among the ten most watched programs in a given week in the United States (The Nielsen Company, 2013). Although less popular than fictional programs, documentary-style crime programs based on actual criminal cases have also proliferated in the past decade, such as A&E’s The First 48, Crime 360, and American Justice, and truTV’s Forensic Files and The Investigators. The popularity of these programs appears to have led to even more outlets for exposure to forensic science and crime investigation via other means as well, as reflected by the rather recent development of specialized exhibits (CSI: The Experience in Las Vegas, Orlando, and Macao through the Museum of Science and History) and even an entire museum (The National Museum of Crime and Punishment, Washington, D.C., which features specialized exhibits and workshops on forensic science investigation).
Only two years after the premiere of CSI in 2000, various media outlets began suggesting the existence of a “CSI effect” in which frequent viewers of CSI and other forensic-themed programs gained an unrealistic perception of the availability and utility of forensic scientific evidence as well as a false sense of expertise in forensic investigative methods (e.g., Roane, 2007; Toobin, 2005). A majority of these media reports have suggested that programs such as CSI have led to unrealistic juror expectations for forensic evidence (Harvey & Derksen, 2009). As of this writing, at least four books have been published on the subject (Byers & Johnson, 2009; McDonald, 2009; Ramsland, 2006; Stevens, 2011), and some scholars have proposed more wide-ranging effects. Cole and Dioso-Villa (2009, 2011), for instance, outlined eight subtypes of the CSI effect to clarify its extensive influence on jurors, police and forensic investigators, attorneys and judges, educational institutions, and the general public. Of these subtypes, perhaps the most threatening to the criminal justice system is the “strong prosecutor’s effect” in which jurors who are heavy viewers of forensic science television programs develop unrealistically high standards for the availability and utility of scientific evidence, and when they are undoubtedly disappointed with actual forensic evidence presented in court, they are more likely to acquit defendants compared to their non-vie wing cohorts. The potential impact of the strong prosecutor’s effect remains a hotly debated topic, with some speculation that it may have played a part in some of the more media sensationalized acquittals, such as in the criminal trials of Robert Blake or Casey Anthony (e.g., Blankstein & Guccione, 2005; Heinrick, 2006; Picht, 2011; Ryan, 2011).
The CSI effect is presently among the most highly conjectured media effects (Sarapin & Sparks, 2009). However, while it has certainly received a great deal of attention in the mass media, and while survey studies have suggested that judges, attorneys, law enforcement officers, and criminal investigators hold strong beliefs in its existence (e.g., Huey, 2010; Patty, Smith, & Stinson, 2008; Robbers, 2008), it has received little empirical support. As of this writing, a total of nine published studies have investigated the strong prosecutor’s effect subtype (Baskin & Sommers, 2010; Hayes-Smith & Levett, 2011; Holmgren & Fordham, 2011; Kim, Barak, & Shelton, 2009; Mancini, 2011; Podías, 2006, 2009; Schweitzer & Saks, 2007; Shelton, Kim, & Barak, 2007), all of which used measures of frequency of forensic television viewing and defendant guilt in response to criminal trial scenarios and/or descriptions. Out of these, only two (Baskin & Sommers, 2010; Hayes-Smith & Levett, 2011) provided any evidence of a relationship between amount of forensic- themed television viewing and judgments of defendant guilt. However, a majority of the studies provided evidence that frequency of forensic- themed television viewing was associated with increased demands for scientific evidence and perceptions of scientific evidence quality (Baskin & Sommers, 2010; Hayes-Smith & Levett, 2011; Holmgren & Fordham, 2011; Kim, Barak, & Shelton, 2009; Schweitzer & Saks, 2007; Shelton, Kim, & Barak, 2007) and familiarity with forensic investigators’ tasks (Mancini, 2011; Schweitzer & Saks, 2007). Furthermore, an experimental study which did not investigate the strong prosecutor’s effect (Smith, Patty, & Stinson, 2007) found that compared to participants who watched no or few (one to three) CSI episodes over a two-week period, participants who watched more (four to eight) episodes of CSI rated scientific evidence (e.g., DNA, fingerprint) more favorably than non-scientific evidence (e.g., eyewitness testimony, confession). Some limitations common in these studies included use of homogeneous undergraduate mock juror samples (Mancini, 2011; Podías, 2006, 2009; Schweitzer & Saks, 2007) and lack of ecological validity due to use of extremely briefly summarized criminal trial scenarios (Baskin & Sommers, 2010; Kim, Barak, & Shelton, 2009; Shelton, Kim, & Barak, 2007).
Recently, some authors (Bilandzic, Busselle, Spitzner, Kalch, & Reich, 2009; Hayes-Smith & Levett, 2011) have suggested that a well- established theory of media effects, known as cultivation theory (Gerbner, 1972; Morgan, Shanahan, & Signorielli, 2009), may be important in understanding the CSI effect, if it exists. In fact, Bilandzic et al. (2009) remarked that CSI effect research has indeed corresponded closely with a cultivation perspective. According to cultivation theory, commonly repeated images and messages presented on television cultivate a distorted perception of social reality such that-within similar demographic subgroups-frequent television viewers are more likely than infrequent viewers to perceive the real world through the lens of the television world. Simply put, the more television a person watches, the more that person is likely to perceive reality as similar to what is viewed on television. Literally hundreds of cultivation studies have been conducted over the past three decades; two meta-analyses of some of this research (Dossche & Van den Bulck, 2010; Shanahan & Morgan, 1999) indicated that television consumption exerts a small but stable contribution to viewers’ beliefs and perspectives, particularly in those of crime and the criminal justice system. As some examples of cultivation study findings that are potentially relevant to the CSI effect, frequent television viewing has been associated with increased fear of crime and criminal victimization (e.g., Eschholz, Chiricos, & Gertz, 2003; Römer, Jamieson, & Aday, 2003), increased perceived prevalence of violence (e.g., Gerbner & Gross, 1976; Nabi & Sullivan, 2001; Signorielli, 1990) and of law enforcement occupations within society (Gerbner & Gross, 1976), and science knowledge and belief in the promise of science (e.g., Nisbet, Scheufele, Shanahan, Moy, Brossard, & Lewenstein, 2002).
The original cultivation hypothesis (Gerbner, 1972; Gerbner & Gross, 1976) contended that general television exposure was responsible for any cultivation effects due to common, consistent messages conveyed across all television programming that could not be attributed to any specific genre or program. However, a common criticism is that grouping all television programming together and regarding its content as homogenous is inaccurate and outdated (Morgan & Shanahan, 2010), particularly due to dramatic changes in recent decades in the way in which the public consumes television programming. Given the massive expansion of cable systems-with some networks specializing in specific genres of programming (e.g., Animal Planet, truTV, SyFy, Travel Channel, History Channel, etc.)-and modern technological developments (e.g., cable and satellite networks, VCRs, DVD players, DVRs, and the Internet), viewers now can exercise far more control than ever before over the selection and scheduling of television programming (Morgan, Shanahan, & Signorielli, 2009). Along with this greater control, it has been suggested that some viewers may selectively watch specific genres of programming (Hammond, Farrar, & Jalette, 2009). As a result, more recent research investigating genre-specific cultivation processes has been conducted and thus far has yielded some promising results, finding relationships between viewing and genre-specific perceptions and attitudes in such genres as talk shows (e.g., Glynn, Huge, Reineke, Hardy, & Shanahan, 2007; Woo & Dominick, 2003), “makeover” programs (e.g., Kubic & Chory, 2007), and even in specific programs such as the popular medical drama Grey’s Anatomy (Quick, 2009). In fact, in their recent meta-analysis, Dossche and Van den Bulck (2010) found that studies measuring genre-specific viewing actually yielded greater effect sizes in predicting cultivation variables than those measuring general television viewing.
With regard to studying how cultivation theory may be used to investigate the CSI effect, it may be important to distinguish the crime show sub-genres of fictional and “reality-based” or documentary-style programs because of the considerable difference in the ways in which the availability, efficiency, and utility of forensic scientific investigative procedures and evidence are depicted in each. In fictional programs (e.g., CSI, NCIS, Bones), highly astute crime scene investigators use extremely sophisticated, efficient techniques and procedures to recover and analyze evidence from crime scenes, which inevitably leads to the identity of the perpetrator, who often confesses to the crime by the end of the one-hour episode. However, in reality, many of these depicted techniques are often wildly exaggerated or even non-existent (Heinrick, 2006; Podías, 2009; Schweitzer & Saks, 2007). Documentary-style programs (e.g., The First 48, Crime 360, Cold Case Files), on the other hand, follow real law enforcement officers and agents investigating real crimes, and thus naturally provide a more realistic depiction of crime investigation, with investigators at times making human errors and using fruitless procedures. In fact, sometimes the perpetrator is never identified by the end of the episode. Thus, perhaps each sub-genre involves distinctively different cultivation processes; fictional programs may involve cultivation processes based on the highly distorted images and messages about the infallibility of forensic science, while documentary- style programs may involve cultivation processes that are less distorted and more grounded in reality.
Few published cultivation studies have investigated genre-specific cultivation processes within crime shows (e.g., Escholz, Chiricos, & Gertz, 2003; Grabe & Drew, 2007; Oliver & Armstrong, 1995), and even fewer have made any distinctions among the sub-genres. Furthermore, no previous CSI effect research has made this distinction, but instead has tended to use viewership measures which group all crime shows together as if they were comparable. However, some research has suggested that there may be meaningful distinctions in viewer attitude and personality variables among those who tend to view one sub-genre more often than the other. For example, in an early survey study comparing these sub- genres-but prior to the existence of CSI and other similar forensic science-based procedurals-Oliver and Armstrong (1995) found that reality-based viewing (e.g., older programs such as Cops, America’s Most Wanted, FBI: The Untold Story) was associated with punitive attitudes toward crime, higher levels of racial prejudice, and higher levels of authoritarianism, while fictional viewing (e.g., Law and Order, Matlock, Heat of the Night) was not associated with these variables. In addition, viewers rated the reality-based programs as more realistic than the fictional programs.
In summary, the strong prosecutor’s subtype of the CSI effect (Cole & Dioso-Villa, 2009, 2011) contends that heavy crime show fiction- viewing jurors develop a distorted view of the availability and utility of forensic scientific evidence and are more likely to acquit a defendant when presented with seemingly substandard pro-prosecution evidence in real life. In other words, these viewers are more likely to favor the defense when pro-prosecution scientific evidence is perceived as below the standards of that depicted in forensic fiction crime show programs and will identify this perceived substandard evidence as the reason for their acquittals. Previous CSI effect and cultivation research has suggested that increased television viewership is associated with differences in beliefs and perceptions about crime, law enforcement occupations and tasks, and even scientific evidence, but there is little to no evidence that viewership is associated with jury decision-making processes. However, given the recent research indicating genre-specific cultivation processes and that virtually no previous CSI effect studies have distinguished sub-genres in measures of crime show viewing, it appears that making this distinction may be important as a next step in this line of research, a step that several researchers have already recommended (Grabe & Drew, 2007; Hayes-Smith & Levett, 2011). As Morgan and Shanahan (2010) contend, heavy television viewers are unlikely to watch only isolated genres of programs, and thus any influence of a specific program or genre ought to be considered in the context of other viewed programming. Therefore, the current study aimed to investigate whether making the distinction between two sub-genres of forensic science television programming (i.e., fictional vs. “reality- based” or documentary-style) and using a difference score measure of their relative viewing to divide participants into lighter and heavier fiction viewership groups might be a successful method of determining any differences in juror perceptions and decision-making. To maximize ecological validity, an actual juror sample viewed a video recorded summary of footage from an actual murder trial and was administered a series of questionnaires measuring viewership habits of the two sub- genres of forensic science programs as well as their verdicts, verdict reasons, and perceptions of the presented evidence and familiarity with forensic investigators’ and scientists’ tasks. The following hypotheses were proposed according to the strong prosecutor’s subtype of the CSI effect: (1) relative to their documentary-style viewership, heavier fiction viewers would perceive a defendant in a criminal trial as less guilty compared to lighter fiction viewers; (2) heavier fiction viewers would more often cite scientific evidence as the main reason for their verdicts compared to lighter fiction viewers; (3) heavier fiction viewers would report greater familiarity with forensic investigators’ and scientists’ tasks than lighter fiction viewers, similar to previous findings (Mancini, 2011; Schweitzer & Saks, 2007); and (4) heavier fiction viewers would report less satisfaction with the prosecution’s evidence and greater satisfaction with the defense’s evidence compared to lighter fiction viewers.
Eighty jurors (68% female, 96% Caucasian, Mage = 49.7, SI) = 12.4) actively serving jury duty in a Western Pennsylvania courthouse were offered $20 each for their participation. Regarding education levels, 1.3% had less than a high school diploma or GED, 25.3% had a high school diploma or GED, 34.2% had some college or a two-year degree, 29.1% had a four-year college degree, 7.6% had a master’s degree, and 2.5% had a doctoral degree. Annual family income levels were comprised of 3.7% with $10,000 or less, 11.0% between $10,001 and $25,000, 18.3% between $25,001 and $50,000, 39.0% between $50,001 and $100,000, 12.2% between $100,001 and $150,000, 6.1% between $150,001 and $200,000, and 3.7% over $200,000. Three participants did not report annual family income.
Criminal trial stimulus video. A 48-minute edited video recorded summary of the criminal trial Colorado v. Galek (2009) included actual courtroom footage as presented on the program In Session on the truTV television channel in November 2009. As shown in the footage, 17-year- old defendant Andrew Galek was charged with the first-degree stabbing murder of 63-year-old Hans Winter, the father of the defendant’s friend, Kevin Winter, during a sleepover in their home. According to the prosecution, after Galek and Kevin Winter had consumed a large amount of tequila over a three-hour period, Galek then stabbed Mr. Winter in the abdomen with a kitchen knife while he lay in bed, awakening both the victim and his wife, Beverly Winter, who reportedly witnessed Galek flee the room. Mrs. Winter testified that Mr. Winter cried out, “Why did you hit me, Andy [Galek]?” when he was stabbed, and while she did not realize that Galek had stabbed the victim initially as she talked with Galek minutes later in the basement of their home, she later noticed the wound while she was on the phone with a 911 emergency operator. Mr. Winter lost consciousness and was unable to provide useful information about the stabbing, and he died in the ambulance as a result of internal bleeding from a severed iliac artery. The prosecution emphasized that Galek eventually fled the home after speaking with Mrs. Winter and hid under a truck, only to be quickly apprehended by police as they threatened to use a Taser. The defense contended that Galek was too intoxicated to remember anything, that Mrs. Winter did not have a clear view of the perpetrator as he left the darkened room, and emphasized that the surface of the bed sheet near the cut in which the knife penetrated contained DNA obtained from skin cells which excluded Galek as a possible perpetrator. The footage included expert testimony from the laboratory technician who conducted the DNA testing. It should be noted that while Galek later confessed to the crime in several letters to a girlfriend while incarcerated and awaiting trial, these details were excluded from the edited trial footage.
Demographic questionnaire. This was a five-item questionnaire assessing participants’ gender, age, race/ethnicity, highest level of education, and approximate family income.
Verdict questionnaire. This was a 10-item questionnaire measuring participants’ prior familiarity with the Colorado v. Galek case (yes or no), dichotomous verdict (guilty or not guilty), confidence in verdict on a 10-point Likert-type scale (1 = not at all confident to 10 = completely confident), an open-ended item requesting the participants’ two most important reasons for their verdicts, an open-ended estimate of the percentage of likelihood of crime commission generally required for a defendant to be found guilty, and an estimate of the probability that the defendant committed the crime (0% to 100% in 10% increments). Participants were asked to render a guilty verdict based upon the “beyond a reasonable doubt” standard. The open-ended verdict reason item was dichotomously dummy-coded (0 = absent, 1 = present) for the presence or absence of the mention of DNA because this was the main piece of scientific evidence highlighted in the trial footage. In addition, the mention of any scientific evidence at all (e.g., DNA, blood alcohol content of the defendant, fingerprints, blood spatter analysis, etc.) was also dummy-coded. The last four items measured participants’ level of agreement with statements of perceived familiarity with forensic investigators’ and scientists’ tasks (similar to items used successfully by Schweitzer & Saks, 2007 & Mancini, 2011) as well as satisfaction with the prosecution’s and defense’s presented scientific evidence, all rated on 7-point Likert-type scales (1 = not at all agree to 7 = completely agree).
Forensic television viewing habits questionnaire. This questionnaire measured participants’ estimates of viewing frequency and realism over the prior three-month period of a total of 25 currently aired forensic science television programs, either first-run or in rerun syndication. These were divided into 11 forensic science fiction (e.g., CSI: Crime Scene Investigation, CSI: NY, NCIS: Naval Criminal Investigative Service, Body of Proof, Bones) and 14 forensic science documentary (e.g., Forensic Files, Cold Case Files, The First 48, Crime 360, American Justice) programs. Viewing frequency was rated using a scale from 0 = never/less than once per month to 10 = ten or more times per month. For each program that was rated at least a 1 (indicating viewership of once per month), the participant was instructed to rate that program’s realism using a 5-point Likert-type scale (1 = not at all realistic to 5 = extremely realistic).
Prior to each of eight data collection sessions, potential participant jurors were solicited while seated in an approximately 200-seat capacity jury waiting room during a period in which it was known that they would not be selected for voir dire for the day’s docket, usually right before their 90-minute lunch break. They were asked to participate in a study of jury decision-making and were told that the study would involve the viewing of a 48-minute video of footage from a real criminal trial and the completion of a set of questionnaires, all of which would require approximately 60-75 minutes. Finally, they were told that they would receive $20 compensation for their participation, and that those who were interested were to take a seat towards the front of the room where the video equipment-including a DVD player and two 42-inch LCD televisions-was located. Those who were not interested were dismissed. Once the participant group was assembled, they were provided with informed consent forms and a packet of questionnaires sealed in envelopes, which they were instructed not to open until further notice. After completing and submitting the informed consent, the participants were provided with a set of instructions-also read aloud to them- asking them to imagine they were jurors in the criminal trial of Colorado v. Galek in which defendant Andrew Galek was charged with the first- degree murder of Hans Winter. Participants were instructed to watch the trial video carefully and to use the standard of beyond a reasonable doubt in their judgments of guilt. The video was then shown to the participants in its entirety, at which time the participants were instructed to open and complete their questionnaire packets. The participants were then debriefed upon completion and submission of their questionnaires and were notified that each would receive a $20 check by mail in the following two to three weeks.
As Dossche & Van den Bulck (2010) recommend, summation scales were used for both the fiction and documentary-style program viewership scores by summing the 11 forensic fiction program ratings and the 14 forensic documentary-style program ratings. A fiction-documentary difference score was then calculated by subtracting the documentary- style viewership scores from the fiction viewership scores; thus, positive scores indicated greater relative fiction viewership, negative scores indicated greater relative documentary-style viewership, and zero or near zero indicated a relatively equal amount of viewership of each sub-genre. Mean realism scores for each of the fiction and documentary-style programs were calculated. A verdict scale score was also calculated as a scalar variable, multiplying the confidence in verdict rating by 1 for guilty verdicts and by -1 for not guilty verdicts, creating a scale range of + 10 (complete confidence in guilt) to -10 (complete confidence in non- guilt).
Of the sample, none indicated familiarity with the case prior to the study, 23 (28.8%) reported viewing no forensic fiction programs, 33 (41.3%) reported viewing no forensic documentary-style programs, and 17 (21.3%) reported viewing no forensic fiction or documentary-style programs during the previous three months. Using only participants (n = 29) who rated the realism of at least the median number of fiction (Md = 2) and documentary-style (Md = 1) programs, a paired samples t test revealed that documentary-style programs (M = 3.48, SI) = 0.85) were rated significantly more realistically than fiction programs (M = 2.89, SI) = 0.87), t (28) = 3.13, p = .004, d = .58. In addition, among the entire sample, mean fiction program realism ratings were positively correlated with frequency of fiction program viewership, r (52) = .28, p = .048, and with mean documentary-style program realism ratings, r (36) = .38, p = .024. However, mean documentary-style program realism ratings were unrelated to frequency of viewership of documentary-style, r (44) = -.08, p = .625, or fiction programs, r (44) = -.04, p = .816. Thus, forensic fiction and documentary-style programs were perceived differently in their levels of realism, realism ratings for both sub-genres were positively associated, and while increased viewing frequency of forensic fiction programs was related to increased perception of the realism of both fiction and documentary-style programs, increased viewing of documentary-style programs was unrelated to perception of realism of either sub-genre.
To test the hypotheses, the most logical approach based on recent cultivation theory research-which has suggested that amount of sub- genre television viewing influences one’s perception of social reality- appeared to require taking the relative amount of both forensic science fiction and documentary-style viewership into account; therefore, the fiction-documentary difference score was used as a basis of dividing participants into two groups. “Lighter fiction” viewers (n = 41 or 51% of the sample) were designated as those with difference scores at or below zero (M = -3.95, SI) = 8.48), while “heavier fiction” viewers (n = 39 or 49% of the sample) were designated as those with difference scores greater than zero (M = 11.69, SD = 10.43). Thus, lighter fiction viewers reported watching either more documentary-style than fiction programs-or at least the same amount of each sub-genre-while heavier fiction viewers reported watching more fiction than documentary-style programs.
Since the first hypothesis was the most critical, descriptive statistics for the verdict scale were compared between the lighter fiction (M = – 2.98, SD = 7.00) and heavier fiction (M = -6.05, SD = 5.17) viewer groups, revealing one outlier case in the heavier fiction group, identified by having rendered a verdict scale score (x = 10) that was greater than 3 standard deviations beyond the group mean (x > 9.46). This case was removed from the dataset, leaving a sample size of n = 79. The overall sample data descriptive statistics are provided in Table 1. A chi-square test of independence revealed that heavier fiction viewers (n = 3 or 7.9%) rendered significantly fewer guilty verdicts than lighter fiction viewers (n = 11 or 26.8%), x2 (1, N = 19) = 4.85, p = .028, cp = .25. Furthermore, following up with the more sensitive scalar verdict scale score, while the continuous difference score viewership measure was only marginally correlated with the verdict scale score in the hypothesized direction, r (79) = -.20, p = .078, heavier fiction viewers endorsed significantly lower verdict scale scores as well as lower defendant guilt percentage scores than lighter fiction viewers (see Table 2), supporting the first hypothesis. Next, a chi-square test of independence revealed that heavier fiction viewers (n = 22 or 57.9%) were no more likely to mention DNA evidence as a verdict reason than lighter fiction viewers (n = 21 or 51.2%), x2 (1, N = 19) = 0.35, p = .552, cp = .07, and a second chi-square test of independence revealed that heavier fiction viewers (n = 26 or 68.4%) were no more likely to mention any scientific evidence at all as a verdict reason compared to lighter fiction viewers (n = 24 or 58.5%), x2 (1, N = 79) = 0.83, p = .362, cp = . 10. These results did not support the second hypothesis. The third hypothesis was also unsupported; as indicated in Table 2, heavier fiction viewers did not endorse significantly greater familiarity with forensic investigators’ or scientists’ tasks compared to lighter fiction viewers. Finally, the fourth hypothesis was unsupported; as indicated in Table 2, compared to lighter fiction viewers, heavier fiction viewers did not report less satisfaction with the prosecution’s evidence or more satisfaction with the defense’s evidence.
Because the fiction and documentary-style programs appeared to be perceived differently in their realism accordingly to the preliminary analyses, and since previous research has relied solely on summed ratings of all forensic television programming as the measure of viewership, a summed score of the fiction and documentary-style viewership ratings was calculated (M = 15.24, SI) = 22.24) to compare it with the difference score measure. Indeed, this summed score did not correlate with the difference score, r (79) = .03, p = .784, suggesting that these scores measured different constructs. To compare them, a series of Pearson product-moment correlation coefficients were calculated between these two scores and the continuous outcome variables used in hypothesis testing and two point-biserial correlations were calculated between these two scores and the dichotomous verdict reason variables (see Table 3). Furthermore, both viewership scores were entered as predictor variables in a series of multiple regression analyses with the same continuous outcome variables (see Table 4).
Considering that the current study used an actual juror sample in a courthouse setting with an ecologically-valid trial stimulus, and in light of the highly limited previous empirical evidence for a relationship between forensic television program viewership and juror judgments of defendant guilt, these findings are important because they lend at least partial support for the strong prosecutor’s subtype of the CSI effect (Cole & Dioso-Villa, 2007, 2009) in real jurors. Using a forensic fiction- documentary viewership difference score to divide participants into two groups, heavier fiction viewers were indeed more likely to acquit the defendant in a murder trial and perceived him as less guilty compared to lighter viewers. However, the proposed route by which heavy fiction- viewing jurors make their decisions according to the strong prosecutor’s effect-in response to dissatisfaction with perceived substandard scientific evidence-was not supported here; there was only a marginally significant positive correlation between the fiction-documentary viewership difference score and mention of any scientific evidence at all as a verdict reason, and neither viewership group more frequently cited DNA or any scientific evidence as a verdict reason, or indicated greater satisfaction with either the prosecution’s or defense’s evidence.
One key methodological difference between the current study and prior CSI effect research is the distinction between fiction and documentary-style sub-genre viewership, using them in relation to one another, which had not been done previously. This was fruitful; while the majority of prior research using single crime show genre viewership measures has failed to detect any relationships between viewership and guilt decisions, the difference score viewership measure in the present study was a comparatively more sensitive predictor of guilt decisions; this is consistent with predictions based upon the strong prosecutor’s subtype of the CSI effect in which heavy fiction viewership will predict acquittals.
To illustrate further the difference between the summed and difference viewership scores, consider the relationships between these scores with the outcome variables in Tables 3 and 4. Perhaps most importantly, the difference score measure appeared to relate to the outcome variables in the hypothesized directions according to the strong prosecutor’s effect more consistently than the summed score. More specifically, the beta weight of the difference score was marginally significant in predicting the verdict scale in the hypothesized direction while taking the summed score’s contribution into account. In addition, the correlation of the difference score with mention of any scientific evidence as a verdict reason was marginally significant in the hypothesized direction. One counterintuitive finding, however, was that the correlation of the difference score and satisfaction with the prosecution’s evidence was marginally significant opposite the direction hypothesized. Comparatively, not only was the summed score uncorrelated with the verdict scale, but some of the relationships of the summed score with other outcome variables were puzzling and counterintuitive. For example, while the correlations and beta weights between the summed score and perceived familiarity with forensic investigators’ and scientists’ tasks were marginally significant in the hypothesized direction, the summed score was negatively associated with percentage certainty required to find defendant guilt in general (the reasonable doubt standard), meaning that heavier viewership of both sub- genres was associated with a lower general threshold required to render a guilty verdict. Furthermore, the correlation between the summed score and mention of DNA evidence as a verdict reason was marginally significant in the opposite direction as hypothesized. In light of these differences and considering that prior CSI effect research using the summed viewership measure has generally failed to predict guilt decisions, it appears that using the summed viewership measure makes little sense. Thus, separate consideration of crime program sub-genres in context may be important in determining whether viewership is related to juror guilt decisions.
Cultivation theory may be helpful as a potential explanation for these findings. Again, as Morgan and Shanahan (2010) have recently suggested, cultivation processes involved in one or more genres of viewership should be considered in the context of overall program viewership. Perhaps the images and messages about forensic science and criminal investigative procedures conveyed in these two main sub-genres of crime programs are distinctive enough from each other to promote different cultivation processes. Furthermore, perhaps those processes in documentary-style programs moderate those in fictional programs, as observed in the differences in acquittal rates between the viewership groups. It would seem that-consistent with the strong prosecutor’s effect subtype-dissatisfaction with the prosecution’s evidence would lead to heavier fiction viewers’ acquittals, and thus cultivation processes in fiction programs would involve more distorted perceptions of the apparent infallibility of forensic scientific evidence, while those involved in documentary-style programs would be more realistic. Nonetheless, if separate sub-genre cultivation processes are occurring, the nature of each is yet unclear; while prior CSI effect research has indicated that crime program viewership is related to beliefs and expectations about scientific evidence (Baskin & Sommers, 2010; Brewer & Ley, 2010; Hayes-Smith & Levett, 2011; Schweitzer & Saks, 2007; Shelton, Kim, & Barak, 2007; Smith, Patty, & Stinson, 2007), the results of the current study did not support this contention. On the other hand, clearly the current study made use of a single criminal trial video stimulus which presented a single type of scientific evidence (i.e., DNA) in a specific way. While recent research has suggested that juries regard DNA evidence as among the most reliable forms of scientific evidence (e.g., Wheate, 2010), juries respond differently to various types of criminal trials and evidence, and thus the results may have varied if another trial stimulus using a less reliable type of evidence was presented.
The current study is not without its limitations. First, the use of a homogenous convenience sample of mainly Caucasian jurors may limit the applicability of the results to other, more diverse jury populations. Nonetheless, there is some convincing evidence that minorities have been significantly and consistently underrepresented on jury panels in both the state and federal courts for many years despite efforts to diversify them (e.g., Fukurai, Butler, & Krooth, 1993; Fukurai & Krooth, 2003). Second, the television viewing measures were based on self-report, which could not be avoided but naturally pose a host of reliability and validity issues. Third, the current study used only initial verdict preference measures, which jury deliberation processes indeed affect (Bothwell, 1999). Lastly, this study may have been underpowered; considering that the difference score viewership measure was only marginally correlated with the verdict scale score and with mention of scientific evidence as a verdict reason in the hypothesized directions, perhaps an increased sample size may have revealed significant relationships among these variables.
Several recommendations for future research are raised here. First, no prior published study has utilized a fiction-documentary viewership difference score to test the strong prosecutor’s subtype of the CSI effect, and therefore replication of these results is essential. It appears that summed fiction and documentary and fiction-documentary difference score viewership measures differ in their relationships with other important outcome variables, and therefore determining which viewership measure or measures are most strongly related to measures of guilt decisions is critical for future research. Furthermore, if fictional programs do involve cultivation processes including unrealistically high expectations of the availability and utility of scientific evidence, and if these distorted perceptions do lead to differences in guilt decisions, it should also be noted that heavier relative fiction-to-documentary viewership was associated with acquittals. This occurred despite the fact that viewers rated fiction programs as less realistic than documentary- style programs, similar to Oliver and Armstrong’s (1995) findings for these two sub-genres in pre-CSV era crime show programming. In other words, awareness of the decreased realism of fictional programs relative to documentary-style programs did not appear to influence the process which led to more acquittals among the heavier viewers. Thus, future research might explore potential cognitive biases that may be involved in the relationship between forensic fiction television viewership and juror decision-making.
In addition, future studies should continue to investigate potential moderators of any CSI effect. To date, only three prior studies have made such attempts with limited success, including various demographic variables (Kim, Barak, & Shelton, 2009), amount of forensic evidence (Hayes-Smith & Levett, 2011), and need for cognition (Mancini, 2011) as potential moderators of forensic television viewership measures. The first two of these studies used single genre, summed viewership measures and the third used a ratio measure. In light of the current study’s findings-with the fiction-documentary viewership difference score measure predicting guilt decisions despite lower realism ratings for fictional programs-it may be worthwhile to investigate how these or other moderator variables interact with difference score viewership measures to help explain this potential influence of fictional program viewing. Finally, future studies may include a deliberation component to determine any potential relationship between viewership and verdict following the deliberation process.
To conclude, the current study’s findings suggest that there may be some important differences in viewer variables and/or in cultivation processes among those who tend to watch one crime show sub-genre more than the other, and these differences may influence juror decision- making, including those of defendant guilt. If this is true, then the strong prosecutor’s subtype of the CSI effect may indeed exist in actual jurors. The potential impact of media on juror decision-making is complex and as yet only poorly understood, but it may have serious implications for the criminal justice system. Clearly, replication of the current study’s results is key, and further investigation into identifying specific variables that may moderate the CSI effect is highly recommended.
Baskin, D. R. & Sommers, I. B. (2010). Crime-show-viewing habits and public attitudes toward forensic science: The “CSI effect” revisited. The Justice System Journal, 31, 97-113.
Bibel, S. (2012, June 14). ‘CSI: Crime Scene Investigation’ is the most-watched show in the world-again!” [Web log message.] Retrieved from http://tvbythenumbers.zap2it.com/2012/06/14/csi-crime-scene-investigation- is-the-most-watched-show-in-the-world-2/138212/
Bilandzic, H., Busselle, R., Spitzner, F., Kalch, A., & Reich, S. (2009, May). The CSI cultivation effect: The influences of need for closure and narrative engageability. Paper presented at the 59th Annual Conference of the International Communication Association, Chicago, IL.
Blankstein, A. & Guccione, J. (2005, March 22). The Blake verdict and the “CSI effect.” Chicago Tribune. Retrieved from http://articles. Chicago tnbune.eom/2005-03-22/features/0503220030_l_blake-case-csi-effect-bonny -lee-bakley
Bothwell, R. K. (1999). Social cognition in the courtroom. In W. F. Abbott & J. Batt (Eds.), A handbook of jury research (pp. 17:1-17:45). Philadelphia: American Law Institute.
Brewer, P. R. & Ley, B. L. (2010). Media use and public perceptions of DNA evidence. Science Communication, 32, 93-117.
Byers, M. & Johnson, V. M. (2009). The CSI effect: Television, crime, and governance. Lanham, MD: Lexington Books.
Cacioppo, J. T. & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 42, 116-131.
Cediel, A. & Bergman, L. (Writers and producers). (2012). The real “CSI” [Television series episode]. In D. Fanning (Executive producer), Frontline. Boston, MA: WGBH Educational Foundation.
Cole, S.A. & Dioso-Villa, R. (2009). Investigating the “CSI effect” effect: Media and litigation crisis in criminal law. Stanford Law Review, 61, 1335-1374.
Cole, S. A. & Dioso-Villa, R. (2011). Should judges worry about the “CSI effect”? Court Review, 47, 20-31.
Dossche, D. & Van den Bulck, J. (2010, June). 30+ years of cultivation surveys: A meta-analysis. Paper presented at the 60th Annual Conference of the International Communication Association, Singapore.
Eschholz, S., Chiricos, T., & Gertz, M. (2003). Television and fear of crime: Program types, audience traits, and the mediating effect of perceived neighborhood racial composition. Social Problems, 50, 395-415.
Fukurai, H., Butler, E. W., & Krooth, R. (1993). Race and the jury: Racial disenfranchisement and the search for justice. New York: Plenum Press.
Fukurai, H. & Krooth, R. (2003). Race in the jury box: Affirmative action in jury selection. Albany, NY : State University of New York Press.
Gerbner, G. (1972). Communication and social environment. Scientific American, 227, 152-160.
Gerbner, G. & Gross, L. (1976). Living with television: The violence profile. Journal of Communication, 26, 172-199.
Gerbner, G., Gross, L., Morgan, M., & Signorielli, N. (1980). The “mainstreaming” of America: Violence profile no. 11. Journal of Communication, 30, 10-29.
Glynn, C. J., Huge, M., Reineke, J., Hardy, B., & Shanahan, J. (2007). When Oprah intervenes: Political correlates of daytime talk show viewing. Journal of Broadcasting & Electronic Media, 51, 228-244.
Grabe, M. E., & Drew, D. G. (2007). Crime cultivation: Comparisons across media genres and channels. Journal of Broadcasting & Electronic Media, 51, 147-171.
Hammond, C., Farrar, K., & Jalette, G. (2009, November). The impact of genre and perceived realism on the cultivation process: A re-examination of cultivation theory. Paper presented at the 95th Annual Meeting of the National Communication Association, Chicago, IL.
Harvey, E. & Derksen, L. (2009). Science fiction or social fact?: An exploratory content analysis of popular press reports on the CSI effect. In M. Byers & V. M. Johnson (Eds.), The CSI effect television, crime, and governance (pp. 3- 28). Lanham, MD: Lexington Books.
Hayes-Smith, R. M. & Levett, L. M. (2011). Jury’s still out: How television and crime show viewing influences jurors’ evaluations of evidence. Applied Psychology in Criminal Justice, 7, 29-46.
Heinrick, J. (2006). Everyone’s an expert: The CSI effect’s negative impact on juries. The Triple Helix: The International Journal of Science, Society, and Law, 3, 59-61.
Holmgren, J. A. & Fordham, J. (2011). The CSI effect and the Canadian and Australian jury. Journal of Forensic Sciences, 56, S63-S71.
Huey, L. (2010). “I’ve seen this on CSI”: Criminal investigators’ perceptions about the management of public expectations in the field. Crime, Media, Culture, 6, 49-68.
Kim, Y. S., Barak, G., & Shelton, D. E. (2009). Examining the “CSI-effect” in the cases of circumstantial evidence and eyewitness testimony: Multivariate and path analyses. Journal of Criminal Justice, 37, 452-460.
Kubic, K. N., & Chory, R. M. (2007). Exposure to television makeover programs and perceptions of self. Communication Research Reports, 24, 283-291.
Mancini, D. E. (2011). The CSI effect reconsidered: Is it moderated by need for cognition. North American Journal of Psychology, 13, 155-174.
McDonald, A. F. (2009). Media, crime & culture: An investigation into the CSI effect. Saarbrücken, Germany: VDM Publishing Group.
Morgan, M., & Shanahan, J. (2010). The state of cultivation. Journal of Broadcasting & Electronic Media, 54, 337-355.
Morgan, M., Shanahan, J., & Signorielli, N. (2009). Growing up with television: Cultivation processes. In J. Bryant & M. B. Oliver (Eds.), Media effects: Advances in theory and research (pp. 34-49). New York: Routledge.
Nabi, R. L., & Sullivan, J. L. (2001). Does television viewing relate to engagement in a protective action against crime?: A cultivation analysis from a theory of reasoned action perspective. Communication Research, 28, 802- 825.
Nisbet, M. C., Scheufeie, D. A., Shanahan, J., Moy, P., Brassard, D., & Lewenstein, B. V. (2002). Knowledge, reservations, or promise? A media effects model for public perceptions of science and technology. Communication Research, 29, 584-608.
Oliver, M. B., & Armstrong, G. B. (1995). Predictors of viewing and enjoyment of reality-based and fictional crime shows. Journalism & Mass Communication Quarterly, 72, 559-570.
Patry, M., Smith, S. M., & Stinson, V. (2008). The reality of the CSI effect. In J. Greenberg & C. Elliott (Eds.), Communication in Question: Competing Perspectives on Controversial Issues in Communication Studies (pp. 291- 298). Scarborough, Ontario: Thompson-Nelson.
Patten, D. (2013, May 23). Full 2012-2013 TV season series rankings. [Web log message.] Retrieved from http://www.deadline.com/2013/05/tv-season- series-rankings-2013-full-list/
Picht, J. (2011, July 7). Casey Anthony and the CSI effect. [Web log message.] Retrieved from http://communities.washingtontimes.com/neighborhood /stimulus/2011/jul/7/casey-anthony-and-csi-effect/
Podías, K. (2006). “The CSI effect”: Exposing the media myth. Fordham Intellectual Property, Media and Entertainment Law Journal, 16, 429-465.
Podías, K. (2009). The “CSI effect” and other forensic fictions. Loyola of Los Angeles Entertainment Law Review, 27, 87-125.
Potter, W. J. (1993). Cultivation theory and research: A conceptual critique. Human Communication Research, 19, 564-601.
Quick, B. (2009). The effects of viewing Grey’s Anatomy on perceptions of doctors and patient satisfaction. Journal of Broadcasting & Electronic Media, 53, 38-55.
Ramsland, K. (2006). The C.S.I. effect. New York: Berkley Publishing.
Roane, K. T. (2005, April 25). The CSI effect. U.S. News & World Report, 738(15), 48-54.
Robbers, M. L. P. (2008). Blinded by science: The social construction of reality in forensic television shows and its effect on criminal jury trials. Criminal Justice Policy Review, 19, 84-102.
Römer, D., Jamieson, K. H., & Aday, S. (2003). Television news and the cultivation of fear of crime. Journal of Communication, 53, 88-104.
Ryan, K. J. (2011, July 15). The Casey Anthony acquittal: Mismanaging the “CSI effect.” [Web log message.] Retrieved from http://blogs. discovery, com/criminal report/2011/07/the-casev-anthonv-acquittal-mismanaging-the- csi-effect.html
Sarapin, S., & Sparks, G. (2009, October). The CSI effect: Initial evidence and implications for the jury box. Paper presented at the 1st Annual Meeting of the International Crime, Media, and Popular Culture Studies Conference, Terre Haute, IN.
Schweitzer, N. J., & Saks, M. J. (2007). The CSI effect: Popular fiction about forensic science affects the public’s expectations about real forensic science. Jurimetrics, 47, 357-364.
Shanahan, J., & Morgan, M. (1999). Television and its viewers: Cultivation theory and research. London: Cambridge University Press.
Shelton, D. E., Barak, G., & Kim, Y. S. (2011). Studying juror expectations for scientific evidence: A new model for looking at the CSI myth. Court Review, 47, 8-18.
Shelton, D. E., Kim, Y. S., & Barak, G. (2007). A study of juror expectations and demands concerning scientific evidence: Does the “CSI effect” exist? Vanderbilt Journal of Entertainment and Technology Law, 9, 331-368.
Signorielli, N. (1990). Television’s mean and dangerous world: A continuation of the cultural indicators perspective. In N. Signorielli & M. Morgan (Eds.), Cultivation analysis: New directions in media effects research (pp. 85-106). Newbury Park, CA: Sage.
Smith, S. M., Patry, M. W., & Stinson, V. (2007). But what is the CSI effect? How crime dramas influence people’s beliefs about forensic evidence. The Canadian Journal of Police & Security Services, 5, 187-195.
Smith, S. M., Stinson, V., & Patry, M. W. (2011). Fact or fiction? The myth and reality of the CSI effect. Court Review, 47, 4-7.
Stevens, D. J. (2011). Media and criminal justice: The CSI effect. Sudbury, MA: Jones and Bartlett Publishers.
Toobin, J. (2007, May 7). The CSI effect. New Yorker, 85,30-35.
Tyler, T. R. (2006). Viewing CSI and the threshold of guilt: Managing truth and justice in reality and fiction. The Yale Law Journal, 115, 1050-1085.
Wheate, R. (2010). The importance of DNA evidence to juries in criminal trials. The International Journal of Evidence & Proof, 14, 129-145.
Woo, H., & Dominick, J. R. (2003). Acculturation, cultivation, and daytime TV talk shows. Journalism & Mass Communication Quarterly, 80, 109-127.
Note: Special thanks to Dr. Peter Hutchinson of the Faculty Development Committee and Dr. John Smetanka, Vice President of Academic Affairs, at Saint Vincent College in Latrobe, PA, for providing the funding for this study, and to the staff of the Jury Commissioners’ Office of the Court of Common Pleas of Westmoreland County in Greensburg, PA, for their hospitality and willingness to provide administrative help in data collection.
Dante E. Mancini
Indiana University of Pennsylvania
Author info: Correspondence should be sent to: Dr. Dante E. Mancini, Indiana University of Pennsylvania, Department of Psychology – Uhler Hall 302, 1020 Oakland Avenue, Indiana, PA 15705, E-mail: email@example.com.
Word count: 8041
Copyright North American Journal of Psychology Dec 2013