Assessing alibi believability: the alibi story, presence of physical evidence and timing of disclosure

Abstract

The content of the alibi story, corroborative physical evidence and timing of the alibi’s disclosure may all affect alibi believability. CloudResearch workers (N = 276) read a mock case. The defendant’s alibi story was scandalous (or neutral), included a receipt (or no mention), and was disclosed one day (or three weeks) later. Participants rated the alibi’s believability, the defendant’s character, and rendered a verdict. As expected, alibi believability ratings were higher when the alibi story was neutral versus scandalous (p < .01) Similarly, the defendant was viewed positively on all eight traits when the alibi story was neutral (ps < .05). The defendant was seen as more trustworthy when he did not have a receipt than when he did (p < .05). Alibi timing had little impact on the dependent measures, but participants were more certain in their verdicts when the alibi was early versus late (p < .05).

The application of DNA testing in criminal cases has been instrumental in the exoneration of innocent people in the United States. There have been 375 DNA exonerations in the United States to date (Innocence Project, 2021). Researchers have studied the factors that contributed to these wrongful convictions. Some of the causes of these miscarriages of justice are eyewitness errors, false confessions, junk science and jailhouse informants (Innocence Project, 2021). A relatively understudied factor that also affects wrongful convictions is a weak alibi. An alibi refers to the claim that during the commission of a crime, a defendant was elsewhere and thus could not have committed the crime (Gooderson, 1977). Alibis are considered weak when they lack corroborating evidence like a receipt or person who confirms the alibi claim (Olson & Wells, 2004).

The general public appears to assume that defendants should be able to recollect their whereabouts and actions from an earlier date when faced with potential incarceration (Burke & Turtle, 2003). When these expectations are not actualized, people tend to be suspicious of a defendant’s failure to remember. Equipped with this mindset, alibi evaluators tend to misinterpret alibi inconsistencies as evidence of deliberate deception (e.g. Crozier et al., 2017; Culhane & Hosch, 2012). The skepticism that police and laypeople hold toward alibis creates a courtroom environment where alibis are assumed to be fabrications until otherwise proven to be true (Burke & Turtle, 2003).

Physical evidence

Not only are defendants expected to easily and accurately remember their whereabouts at the time of the crime, but they are also often expected to have evidence to support the alibi claim. Olson and Wells (2004) proposed a taxonomy of alibi evidence. They postulated that alibi believability varies as a function of the supportive alibi evidence, perceived ease of fabrication of said evidence and perceived credibility of the alibi corroborator or witness. Alibis are more believable when there is corroborating evidence as opposed to none (Olson & Wells, 2004). The relationship between the alibi corroborator and defendant seems to drive alibi believability ratings – the more distant the corroborator is from the defendant, the more believable the alibi (e.g. Eastwood et al., 2020; Hosch et al., 2011; Lindsay et al., 1986; Olson & Wells, 2004).

Physical evidence has been found to be one of the most compelling types of corroborative alibi evidence (Olson & Wells, 2004). Several studies have found that the presence of physical evidence has a strong impact on alibi believability (e.g. Allison et al., 2020; Dysart & Strange, 2012; Jung et al., 2013). What seems to distinguish one form of physical evidence from another is its ease of fabrication (Olson & Wells, 2004). Having strong physical evidence (e.g. videotaped footage) to support one’s whereabouts has not only increased alibi believability ratings, but also lowered guilt ratings (e.g. Jung et al., 2013; Olson & Wells, 2004; Sargent & Bradfield, 2004). Law enforcement personnel have confirmed the findings from experimental alibi studies. In a survey of police officers, they indicated that the most difficult to fabricate forms of physical evidence, and therefore most believable, were incarceration records and surveillance footage even though physical evidence is rarely used in court (Burke & Turtle, 2003), in police investigations (Dysart & Strange, 2012) and in alibi generation studies in the lab where participants are asked to gather evidence to support their alibis (Olson & Charman, 2012; Strange et al., 2014).

The alibi story

Physical alibi evidence aside, alibi stories can vary in terms of their surface traits. Alibi stories can vary according to their morality, their sexual content (i.e. salaciousness) and their legality (e.g. Allison et al., 2012; Keeping et al., 2017). Alibis may have elements of all three or some combination of them. Cheating on an exam may be seen as immoral, and it breaks a university’s honor code, but it is not illegal. Having an affair may be seen as immoral, and it is salacious, but it is not illegal. Admitting to a theft when faced with a homicide charge may be seen as immoral, and it is illegal, but it is not salacious behavior. The impact of the alibi story on alibi evaluators’ ratings of the alibi’s believability is unclear because past studies have yielded mixed results. Sometimes such scandalous alibis lead to positive views of the alibi and the defendant (Allison et al., 2012), sometimes it has no impact on alibi believability (Allison et al., 2014), and sometimes it can lead to negative views of defendants (e.g. Jung et al., 2013).

The defendant in Allison et al. (2012) was accused of armed robbery. He claimed to have been watching pornography (vs. a regular movie) at the time of the crime, and the alibi was supported by a corroborator. Having a salacious alibi led to higher alibi believability ratings, more positive views of the defendant’s character (less suspicious, more honest) and more positive views of the corroborator (higher corroborator believability). In another study, defendants accused of rape claimed to be copying pornographic movies to DVDs with the intent to sell them (salacious), copying a regular movie with the intent to sell (illegal) or simply watching a movie (neutral control, Allison et al., 2014). While alibi salaciousness did not affect alibi believability ratings, having an illegal alibi did lead to more positive views of the defendant and corroborator. Thus, in some circumstances, those evaluating scandalous alibis might be more inclined to believe them because they reason that someone would not fabricate a story about engaging in immoral or scandalous behavior unless it was true.

Other researchers have found that scandalous alibi stories result in negative views of suspects and defendants. Participants in Jung et al. (2013) read about a sexual assault, physical assault or theft. They then read alibis where the defendant was having dinner either with a work colleague (control) or with his mistress (salacious). While the salaciousness manipulation did not affect alibi believability, defendants with neutral alibis were viewed more positively on several character trait measures than defendants with salacious alibis, but only when physical evidence corroborated the alibi. The type of crime affected prison sentences but type of crime and salaciousness did not interact. In another salaciousness study involving a homicide, the researchers included a variety of alibis that varied in terms of sexual salaciousness, legality and consistency (Keeping et al., 2017). For example, some alibi activities involved illegal behavior like insider trading, others involved sexually salacious behaviors like calling a sex hotline, and some were both illegal and sexually salacious (e.g. voyeurism). Illegal alibis were associated with lower believability ratings and higher guilt ratings. Sexually salacious alibis were viewed in the same way (less believable, more likely to be guilty).

One group of researchers found that higher believability ratings occurred when alibis were changed to become salacious. In Nieuwkamp et al. (2016), online community participants evaluated alibis from suspects accused of committing robbery. The alibis either were given right away (immediate) or were changed from the initial telling because of a memory mistake or a lie. The alibis were also either salacious (at a hotel with mistress) or neutral (with cousin). Alibis that were changed because of a lie were seen as more believable than alibis that were simply a result of a memory mistake. Neutral alibis in the immediate condition were more believable than salacious alibis, confirming some past studies (e.g. Jung et al., 2013; Keeping et al., 2017). However, when alibis were presented a second time, salacious alibis were now the more believable ones when compared with neutral alibis. This study sheds some light on the conflicting results of past work: when an alibi is presented only one time, having a scandalous alibi can hurt defendants. However, when alibis are presented twice, and there is a change in the second telling, these new scandalous alibis actually helped defendants (Nieuwkamp et al., 2016).

Timing of disclosure

The timing at which an alibi is brought forth also may affect perceptions of the alibi and defendant. In the United States at the federal level, unless otherwise changed by state statute, defendants have 10 days to notify the prosecution of their intention to offer an alibi defense (Notice of an Alibi Defense, 1975). However, several state jurisdictions have statutes with separate time limits for when the prior notice must occur. For instance, in North Carolina after a case is set for trial, the defense then has 20 working days to inform the prosecution of the intent to employ an alibi defense (North Carolina General Statutes § 15 A-905, 2020). The court sets such a time frame to ensure that the prosecution has the opportunity to investigate the alibi claim.

The effects of the timing of an alibi’s disclosure on jurors have only been minimally investigated. In one study involving a vandalism case, undergraduate participants either read about an alibi in a control condition where there was no mention of context or read about the alibi in the context of a police investigation or trial (Sommers & Douglass, 2007). The alibi was seen as stronger when it was presented in the context of a police investigation and control condition versus at trial. The alibi was also seen as more credible in the investigation context than at trial. The authors reasoned that alibis in the context of a trial were not to be believed because the prosecution would only have brought the case to trial if they thought they would win (and so the defendant was likely guilty).

More recently, Fawcett (2016) examined the timing of alibi disclosure in an assault case. She varied the timeliness of the alibi (‘timely’ vs. ‘ambush’) along with the motivation of the alibi witness (motivated vs. unmotivated; Fawcett, 2016). Alibis considered timely were those revealed during a police interview, while ambush alibis were those only disclosed during the trial. There were no significant effects of timing on the mock jurors’ ratings of defendant or witness reliability, nor on verdicts. However, Fawcett (2016) found that mock jurors gave higher reliability ratings to unmotivated witnesses in the timely alibi condition.

In Allison et al. (2020) a defendant was accused of breaking into an apartment, assaulting the resident (either felony or misdemeanor assault) and committing theft. The defendant claimed to have been shopping at the time and had no corroborative physical evidence, had a store receipt or had security camera footage from the store. After he was formally charged, he presented his alibi to the prosecution either 1 day (early) after the case was set for trial or 19 days later (later). While physical evidence affected alibi believability (with the video footage being the most believable), timing and crime severity had no impact on believability. Physical evidence also affected views of the defendant’s character, with strong evidence leading to more positive views on all measures. Interestingly, timing did affect two character trait measures: defendants who disclosed early were viewed as more trustworthy than those who disclosed later. Further, in the presence of corroborative physical evidence, the participants’ ratings of defendant confidence were increased by early alibi disclosure (Allison et al., 2020). A deeper understanding of the influence of alibi timing on alibi believability is needed because few studies have included it as a variable, but these results seem to suggest that early disclosure can lead to more positive views of the suspect or defendant.

The current study

The purpose of the present study was to investigate how physical evidence, the scandalous nature of the alibi story and the timing of disclosure influence alibi believability. Workers from CloudResearch crowdsourcing platform read a mock crime scenario in which a defendant was on trial for a serious drug offence. He offered an alibi that varied between participants in terms of the presence of corroborative physical alibi evidence (receipt present or not), the scandalous nature of the alibi story (scandalous or neutral) and timing of presentation of the alibi (earlier or later). To date, no other studies have examined how the content of the alibi story affects views of alibis that are disclosed earlier or later. We hypothesized that alibi believability would be highest when corroborative physical alibi evidence was present (Olson & Wells, 2004) and when the alibi was disclosed early (Allison et al., 2020; Fawcett, 2016). The impact of the scandalous nature of the alibi story was more exploratory, given the conflicting results from past research.

Method

This study was approved by the authors’ Institutional Review Board.

Participants

Participants (N = 276) were recruited via CloudResearch (formerly TurkPrime; Litman et al., 2017, and https://www.cloudresearch.com/) crowdsourcing platform that uses Amazon’s Mechanical Turk online workers. Only users with U.S. IP addresses could participate in the study. Further, all participants were labeled as ‘master workers’ in CloudResearch, had completed at least 50 hits (i.e. MTurk jobs) and had received approval ratings of 70/100 or higher on that work. This means that the workers were considered high-quality workers by CloudResearch and were not likely to rush through the study. In CloudResearch, participants select which studies they wish to participate in and are informed of the compensation amount and approximate time length before signing up. This project was described as a study about how people make decisions in the criminal justice system. They would be asked to read a crime report and answer several items on a questionnaire and would take about 30 minutes. Participants could stop taking part at any time and still be compensated. Participants were compensated with $1.50 for their participation in completing the survey. We collected the data between 31 May 2020 and 5 August 2020.

Participants selected their age from a list of eight categories. The majority of participants were between 26 and 35 years old (35.1%). The next most frequent age categories were 36–45 years (28.3%), 46–55 years (17.4%) and 56–65 years (9.4%). Fewer participants selected 18–25 years (5.4%) and 66–75 years (4.3%). No one selected 76–85 years or 85+ years. Participants were fairly evenly split between male and female (52.9% male, 44.6% female), while some participants identified as gender nonconforming (1.1%), and a few participants preferred not to answer (1.1%). The participants were from 46 different states. The most represented states were California (12.3%), New York and Texas (each 8.3%), and Florida (7.2%). The majority of the participants identified as White (77.2%), while others identified as African American or Black (12.3%), Hispanic, Latino, or Spanish Origin (6.9%), or Asian American (5.15%).

Materials and procedure

Design

We used a 2 (timing of disclosure: early vs. later) × 2 (alibi story: scandalous vs. neutral) × 2 (presence of physical evidence: present vs. absent) between-participants design. Participants were randomly assigned by the Qualtrics software to one of eight experimental conditions. The number of participants in each cell was either 34 or 35.

Procedure

Participants signed up for the research study through CloudResearch. Once participants signed up, they were prompted with a screen containing an informed consent form, and if they consented then they moved on to the crime description. Here, participants read one of eight versions of a crime scenario involving the illegal sale of methamphetamines (see Appendix). The crime scenario included a description of a crime in which a drug deal was reported outside of a coffee shop on a college campus. The investigation focused on a student who was seen in the coffee shop behaving suspiciously. The police obtained a search warrant of the student’s apartment where they found unlabeled methamphetamines in a common room. The suspect denied that the drugs were his and stated that he believed he was taking an exam at the time of the crime. As the case moved to trial, the defendant indicated he was pursuing an alibi defense. It was at this point in the description that the crime scenarios varied according to the independent variables: the alibi was disclosed either one working day after the case was set for trial or 20 working days after the case was set for trial (timing); the defendant claimed to be taking an exam or was being paid to take an exam for a friend (scandalous alibi story); and either the defendant had a time stamped vending machine receipt from outside the exam room or there was no mention of a receipt (presence of physical alibi evidence). The implication here is that there was physical alibi evidence but it could have been fabricated.

Participants were asked to act as mock jurors as they completed the survey items. The survey items measured participant perceptions of alibi believability using an 11-point rating sale (0 = ‘I do not believe the alibi at all,’ and 10 = ‘I believe the alibi completely’) and eight defendant character traits (e.g. credibility, honesty, trustworthiness) based on Olson and Wells (2004) also using an 11-point rating scale (0 = ‘does not describe at all,’ and 10 = ‘describes perfectly’). Finally, participants rendered a dichotomous verdict of either ‘guilty’ or ‘not guilty,’ and indicated the extent to which they were certain in their verdict decision using an 11-point scale (0 = ‘very uncertain,’ and 10 = ‘very certain’). Those who found the defendant guilty were then asked to give a sentencing recommendation from a set of 10 fixed alternative responses (‘less than 3 months’ to ‘9+ years’) that were based on the sentencing guidelines for the state in which the authors live (North Carolina General Statutes. § 90-95, 2020).

Participants were asked six manipulation check questions to make sure they had read the crime scenario carefully. These multiple-choice questions asked about the name of the defendant, the crime he was accused of and the date of the crime, and then there were three questions about the independent variables (timing of when the alibi was disclosed, the activities of the alibi and the physical evidence that supported his whereabouts). Each question contained four possible responses. We examined the responses to the manipulation check questions from the 276 participants and removed and replaced seven participants from five experimental conditions who scored less than 50% on a summed total score of the manipulation check questions. Thus, 283 participants took part in total (the original 276, then 7 were removed and replaced with 7 new participants). The results presented below include only those participants who scored 50% or higher on the summed total score (M = 78.3%, SD = 16.7%). Participants scored well on all manipulation check questions (80–94%) with the exception of alibi timing (M = 36%, SD = 48%). Participants in the early condition performed well on the timing manipulation check (M = 70%, SD = 46%) while those in the later condition performed very poorly on this question (M = 2%, SD = 14.6%).

After finishing the survey, participants clicked on a separate link that gave them a special code to enter into Mechanical Turk as proof of their participation. They were paid the same day that they completed their work. Participants then took 9.95 min on average to complete the study (SD = 6.43).

Results

Alibi believability

An analysis of variance (ANOVA) with the three independent variables as between-subjects factors on alibi believability showed that neutral alibis (M = 6.94, SD = 2.35) were more believable than scandalous alibis (M = 5.75, SD = 2.63), F(1, 268) = 15.68, p = .00, ${\eta }_p^2$ = .06, confirming several past studies (e.g. Keeping et al., 2017). Neither of the other independent variables affected alibi believability (ps > .05).

Defendant character traits

A multivariate analysis of variance (MANOVA) with Bonferroni adjustment was run on all eight defendant character traits with the three independent variables as between-subjects factors. This test showed that there was a multivariate main effect of alibi story using Wilks’s lambda, F(8, 261) = 3.03, p = .00, ${\eta }_p^2$ = .09. There was also a multivariate main effect of the presence of physical evidence using Wilks’s lambda, F(8, 261) = 2.00, p = .04, ${\eta }_p^2$ = .06.

Consistent with the alibi believability findings and Jung et al. (2013), the univariate analyses found that the alibi story affected ratings on all eight character traits – neutral alibis led to higher character trait ratings than scandalous alibis (see Table 1). The presence of physical evidence univariately affected trustworthiness ratings. In contrast to expectations, defendants who did not have a receipt (M = 5.26, SD = 2.71) were seen as more trustworthy than defendants who had a receipt (M = 4.60, SD = 2.73), F(1, 268) = 4.46, p = .04, ${\eta }_p^2$ = .02. The timing of the alibi’s disclosure did not affect character trait ratings (p > .05).

Table 1.

Univariate effects for alibi story on eight character trait measures.

	Alibi story
Dependent measure	Scandalous M (SD)	Neutral M (SD)	F test
Credible	4.63 (2.62)	5.87 (2.30)	F(1, 268) = 17.40, p = .00, ${\eta }_p^2$ = .06
Honest	4.32 (2.70)	5.80 (2.53)	F(1, 268) = 19.94, p = .00, ${\eta }_p^2$ = .07
Persuasive	4.85 (2.61)	5.94 (2.46)	F(1, 268) = 12.80, p = .00, ${\eta }_p^2$ = .05
Knowledgeable	5.38 (2.61)	6.17 (2.20)	F(1, 268) = 7.30, p = .01, ${\eta }_p^2$ = .03
Intelligent	5.78 (2.66)	6.51 (1.94)	F(1, 268) = 6.54, p = .01, ${\eta }_p^2$ = .02
Likeable	4.57 (2.76)	5.42 (2.29)	F(1, 268) = 7.82, p = .01, ${\eta }_p^2$ = .03
Confident	6.25 (2.28)	6.82 (2.05)	F(1, 268) = 4.74, p = .03, ${\eta }_p^2$ = .02
Trustworthy	4.30 (2.90)	5.56 (2.42)	F(1, 268) = 15.82, p = .00, ${\eta }_p^2$ = .06

However, there was an interaction of presence of physical evidence and timing on honesty, F(1, 268) = 6.48, p = .01, ${\eta }_p^2$ = .02. A follow-up ANOVA showed that when the alibi was disclosed early, defendants with a receipt were in fact viewed as less honest (M = 4.70, SD = 2.72) than those without a receipt (M = 5.70, SD = 2.57), F(1, 135) = 4.84, p = .03, ${\eta }_p^2$ = .04, perhaps because participants noted that the receipt could have been easily fabricated. There was an interaction of the presence of physical evidence and timing on trustworthiness ratings as well, F(1, 268) = 4.17, p = .04, ${\eta }_p^2$ = .02. The ANOVA once again showed that early alibi disclosure led to lower trustworthiness ratings when there was a receipt (M = 4.41, SD = 2.74) than when there was no receipt (M = 5.72, SD = 2.59), F(1, 135) = 8.32, p = .01, ${\eta }_p^2$ = .06. There was an alibi story and timing interaction on confidence, but follow-up analyses failed to reach significance (p > .05).

Verdicts

The majority of participants found the defendant not guilty of selling or distributing methamphetamines (67.4%). Of the 32.6% who chose a guilty verdict, the majority recommended a sentence of 1–2 years (20%), 6 months–1 year (16.7%) or 3–6 months (16.7%). A logistic regression with all three independent variables and their two-way interactions and the three-way interaction with verdict as the dependent measure resulted in no significant findings (all ps > .05). The pattern of verdicts (majority not guilty) was not affected by any of the independent variables.

Turning to certainty in verdict, an ANOVA on this measure with the three independent variables as between-subjects factors found a main effect of alibi story, F(1, 268) = 7.17, p = .01, ${\eta }_p^2$ = .03. Confirming some past studies (e.g. Keeping et al., 2017), participants were more certain in their verdicts when they read the neutral alibi story (M = 6.93, SD = 2.55) versus the scandalous alibi (M = 6.08, SD = 2.80). There was also a main effect for alibi timing, F(1, 268) = 5.46, p = .02, ${\eta }_p^2$ = .02, showing some support for the benefit of early disclosure (Fawcett, 2016). Participants were more certain in their verdicts when the alibi was disclosed early (M = 6.87, SD = 2.58) than when it was disclosed later (M = 6.13, SD = 2.78).

Discussion

The results showed that the content of the alibi story matters when it comes to alibi believability and perceptions of the defendant’s character. The results lend support to some past studies that found that alibi stories that are not scandalous (in terms of morality, legality, sexual behavior) are regarded as more believable than scandalous alibis (e.g. Keeping et al., 2017). Additionally, our study showed that neutral alibi stories yielded higher character trait ratings: the defendant with a neutral alibi was seen as more credible, honest and trustworthy, and so on, than the defendant whose alibi was scandalous.

Hearing that the defendant engaged in academic dishonesty and took an exam for payment made participants view him poorly on several measures. However, these negative views of the defendant did not lead to more guilty verdicts. Jurors’ negative views seemed to be limited to assessments of the alibi and of the defendant’s character, as jurors were perhaps more cautious when it came to verdicts and may have felt that negative behaviors, like cheating, did not necessarily imply guilt. The alibi story, however, was found to influence the mock jurors’ certainty in their verdict such that they were more certain in their verdict decision when the alibi was neutral than when it was scandalous. Thus, having an alibi story that was neutral helped defendants, but the ultimate evaluation, the verdict, was unaffected by moral behavior.

In past studies, any form of supportive physical evidence has boosted alibi believability and defendant character trait ratings (e.g. Allison et al., 2020; Jung et al., 2013; Olson & Wells, 2004). Surprisingly, we found no impact of the presence of physical evidence on believability ratings. The presence of physical evidence in fact lowered ratings on two character traits: having a receipt led to lower ratings of the defendant’s trustworthiness compared to not having a receipt. An interaction found similar results: defendants with a receipt were viewed as less honest than those with no receipt when they presented their alibi early. One possible explanation for these unanticipated findings is that the participants were not reading the crime scenario closely. The crime scenario may have been too detailed and took a long time for the participants to read. However, the participants’ answers to the physical evidence manipulation check question showed that they did pay attention to this manipulation. It is also possible that participants felt that the receipt was fabricated, although some past studies have found that any form of physical evidence boosts alibi believability (Olson & Wells, 2004). Future studies could examine the quality of receipts and compare how alibis are viewed when the receipt is for cash versus for credit card. Presumably credit card receipts would be viewed more positively because the purchase could be verified with the credit card company.

Also inconsistent with our hypothesis, the timing of the alibi’s disclosure had no bearing on the believability of alibis or defendant character trait ratings but early disclosure did lead to participants being more certain in their verdicts. It is possible that the timing of disclosure was deemed unimportant, but we suspect that in fact jurors did not pay attention to it in the later condition. Indeed, the timing manipulation check question results showed that participants in the later timing condition performed exceptionally poorly but the early disclosure participants did notice timing. Had the later disclosure participants paid closer attention to alibi timing, we believe that there may have been differences in the ratings of the alibi’s believability across the different timing of disclosure conditions, perhaps with the higher alibi believability ratings in the early disclosure condition.

However, it is also plausible that even if the participants had paid more attention to the timing of alibi disclosure manipulation the results still would have been the same. Information surrounding the timing of the alibi’s disclosure might not be as salient to jurors when there is alibi story information present. Thus, other factors in the case, such as the presence of physical evidence or the scandalous nature of the defendant’s alibi story, might draw jurors’ attention away from seemingly minor details about the timing of the alibi’s disclosure. The timing of the alibi’s disclosure may matter more to jurors when these other contextual factors are absent from the case. Another possibility is that the late timing manipulation was not late enough. In our study, the defendant’s disclosure was within the legal limits. Future researchers could examine how disclosures that occur outside of the legal time limits are viewed. Perhaps these very late disclosures would be noticed more and would lead to more adverse inferences.

The majority of participants in this study found the defendant not guilty. This may be because there was not enough evidence against the defendant for participants to convict, regardless of experimental condition. In future studies, researchers should include more evidence against the defendant so that participants feel that there it is enough evidence to warrant a conviction. For example, multiple eyewitnesses who identify the defendant would likely lead to more guilty verdicts. Similarly, DNA evidence that places the defendant at the crime scene also would likely increase the number of guilty votes. It would be interesting to compare strong physical alibi evidence (videotaped footage) against strong incriminating evidence like multiple witnesses or DNA.

Additionally, future research could include measures of the participants’ political views. Participant political leanings might explain why the majority of our participants rendered not guilty verdicts; however, we did not measure whether our participants were more or less politically liberal. Past studies have found that juror verdict decisions are correlated with juror political orientation (e.g. Sivasubramaniam et al., 2020) and that jurors who are more authoritarian or politically conservative are more likely to convict the defendant and give harsher sentences (Bray & Noble, 1978; Devine & Caughlin, 2014; Werner et al., 1982).

Our study was limited by our methodological approach whereby the data were collected entirely online. While the use of the online survey enabled us to obtain a sample that was older than typical undergraduate samples, our study would be strengthened with more racial and ethnic diversity in the sample. It is a good thing that we were able to include participants from many geographic regions within the United States, having 46 states represented. However, in real trials jurors are from the same state. Further, instead of participants witnessing a live mock trial in court, the participants in this study read about a drug case investigation and trial. In this respect, our method again lacked ecological validity because real jurors hear and see witnesses testifying and hear the defense’s and prosecution’s arguments instead of merely reading about them. Our conclusions could be strengthened in future work by using trial videos to simulate the courtroom and deliberation experience; however, past research has shown that the trial modality (written or live or video) does not typically affect dependent measures (e.g. Bornstein, 1999; Pezdek et al., 2009). Further, Bornstein and McCabe (2005) suggested that individual and group differences in verdicts are not always found, indicating that deliberation may not always be needed in jury simulation research.

The nature of the alibi story should continue to be studied, as providing neutral alibi stories may help suspects and defendants. The current study used cheating on a test, a form of academic dishonesty, as its scandalous alibi story. Participants seemed to view the defendant ‘coming clean’ about this transgression as a sign of dishonesty and low credibility. Had the stakes been higher and the alibi activities involved something perhaps more serious and illegal, like engaging in a physical fight, theft or vandalism, then participants may have viewed the alibi and defendant even more negatively. Future studies should continue to investigate which alibi stories elicit positive and negative views of defendants.

In sum, this study indicates that the alibi story impacts mock juror perceptions because defendants whose alibi stories were neutral were viewed more positively than defendants whose alibis were scandalous. These findings suggest that defense lawyers should promote the neutral elements of a defendant’s alibi (and downplay the scandalous elements) because these yield more favorable impressions. There were few effects of alibi timing, perhaps because the later disclosure condition was within the legal limits. More research is needed on how alibis are viewed when defendants provide alibis beyond the legal limits. Typically, having any form of physical evidence helps alibi believability (Olson & Wells, 2004), but we did not find that to be the case in this study. It is possible that mock jurors were skeptical of the vending machine receipt, so further research is needed to determine how sensitive jurors are to the qualities of different kinds of physical evidence (i.e. cash receipt vs. credit card receipt). Alibis that are disclosed later than the legal limits and are supported with cash receipts would presumably be viewed more negatively than alibis disclosed within the legal limits that are supported with a credit card receipt, but this question remains to be studied. Continued research that further investigates the relationship between the alibi story, timing and physical evidence is needed and will help us to discover how these factors impact alibi believability.

Appendix.

Crime and investigation description

Overview of the incident

On Thursday, 21 February 2019, at approximately 3:20 pm, Brakebills Police Department received an anonymous tip that a drug transaction occurred in an alleyway behind the Brakebills University Starbucks. The call was made specifically to the Brakebills Drug Task Force, a response team created to address the rising drug use and sale in the town. The city had received numerous reports of drug deals occurring during daylight hours, and residents and parents became increasingly alarmed because several of these transactions took place near an elementary school. Upon receiving the tip, police questioned the Starbucks employees about whether they had noticed any suspicious behaviors from any customers during that shift. One of the employees mentioned that she thought customer Matthew Brown had exited through the backdoor of the Starbucks. Brown is a sophomore student enrolled in the nearby university, however the worker was not certain whether or not it was Mr. Brown but she stated that she was quite familiar with him because he regularly frequented the coffee shop.

Questioning of Matthew Brown

Matthew Brown arrived at the police station on Friday, 1 March 2019, to follow up with the detectives. When asked about his whereabouts on February 21, Brown stated that he was not involved in a drug deal and that he was pretty sure he was in class at that time.

Development of the investigation

On Tuesday, 5 March 2019, Michael Smith, Matthew Brown’s former roommate, came forward to the police in response to the task force’s public appeal for information regarding the case. Smith had been Brown’s roommate for a semester and he claimed to have information about the drug deal. He said that Brown had asked to buy his prescription Desoxyn, a methamphetamine, a few months earlier. Smith further claimed to have seen Brown abuse drugs repeatedly in their shared residence. Smith explained that the two had a falling out allegedly due to Brown’s drug use, and that this ultimately contributed to Smith moving out.

Upon receiving this statement from Smith, the police were able to obtain a search warrant for Brown’s off-campus house, which he shares with three other roommates. On Thursday, 7 March 2019, the police found a prescription of Desoxyn in the living room, a common area of the house, partially hidden behind a picture frame on a bookshelf. The medication’s label had been peeled off. When questioned about the bottle, Brown denied having taken any of the drug and claimed that it belonged to a friend of one of his housemates, who frequently ‘crashed there’ and often ‘left his stuff behind.’ Brown’s roommates denied knowing who the bottle belonged to.

After exhausting all other possible leads regarding the drug deal, the city prosecutor considered the evidence against Matthew Brown. The prosecution decided to formally charge Brown with distribution/selling methamphetamines. Brown was informed of the charges and he hired a legal team to aid him in challenging them. Brown stated that he may pursue an alibi defense as he was in class at the time of the alleged crime. On Wednesday, 3 April 2019, an initial court proceeding was conducted in order to establish a start date for the trial against Matthew Brown. The trial was set to begin on Thursday, June 13, 2019.

Defense case

The state statutes require that the defense team inform the prosecutor and district court of their intent to offer an alibi defense at trial. This notice must be provided within 20 working days after the case is set for trial. Matthew Brown, through his legal team, informed the prosecution and the court 1 working day later, on 4 April 2019 [earlier condition]/20 working days later, on 30 April 2019 [later condition] that they would be presenting an alibi defense. The defendant, Matthew Brown, and his legal team maintained that Matthew Brown was not at the scene of the crime. They asserted that he was taking an exam for his calculus course [neutral alibi story condition]/he was taking a calculus exam for his friend in exchange for payment [scandalous alibi story condition], a required exam that all students were expected to take. Brown was able to produce a drink receipt from the vending machine located in the building of the exam with a timestamp five minutes before the scheduled start time of the exam [physical evidence present condition]/[physical evidence absent condition – no mention of underlined portion].

Trial

The prosecution claimed that the Starbucks worker was able to make a tentative identification of the defendant, and that Brown had a history of illegal methamphetamine use, as corroborated by his former roommate’s statement. The prosecution also claimed that Brown had access to methamphetamines to sell as indicated by the Desoxyn found in Brown’s current residence.

The defense claimed that there was little evidence linking Brown to the crime, other than the allegations of a resentful former roommate. Further, the Starbucks worker was mistaken that Brown was at the Starbucks that day. They claimed that she likely mixed up the days given that Brown is a regular customer. The defense further claimed that there was no evidence that linked Brown with the medications seized from the house. They were found in a common area and could have been put there by any of the roommates or visitors. Matthew Brown took the stand and repeated the alibi that he gave 1 day [earlier condition]/20 days [later condition] after the case was set for trial. He stated that he was not at the Starbucks at the time of the drug deal, that he was across campus, taking a calculus exam [neutral alibi story condition]/taking a calculus exam for his friend [scandalous alibi story condition]. He provided the vending machine receipt time stamped 5 minutes before the start of the exam [physical evidence present condition only/no mention of underlined portion in absent physical evidence condition]. Brown further testified that he did not leave the exam early, leaving the exam room at 4:00 pm.

Ethical standards

Declaration of conflicts of interest

Meredith Allison has declared no conflicts of interest

Ashlynn Hawes has declared no conflicts of interest

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee (Institutional Review Board) and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent

Informed consent was obtained from all individual participants included in the study