Changed Alibis

Abstract

Despite numerous calls for more research on alibis, there remain many unanswered questions. Three experiments tested the beliefs about and behaviors toward criminal suspects who changed or maintained their alibi statements. In Experiments 1 (N = 350) and 2 (N = 211), participants included current law enforcement officers, criminal justice and psychology undergraduate students who intended to enter law enforcement (future law enforcement officers), and those students who did not intend to do so (laypersons). Only students were used in Experiment 3 (N = 423). Alibi statements were changed (strengthened or weakened) or were maintained in all studies. Results showed that groups had more favorable beliefs about and behaviors toward those alibi statements that were maintained than those that were changed. This research suggests that any lapse in memory for one’s timeline of events could prove detrimental to the suspect.

Keywords

alibi law enforcement interview evaluation Alibi Belief Questionnaire

An alibi, from Latin meaning “in another place,” is a defense that the accused was somewhere other than the scene of the crime at the time the offense occurred. The first use of the word alibi in the English language was in the 18th century (Merriam-Webster, 2012). The offering of an alibi as a defense to criminal charges has a much longer history. In any place in history, it would only be natural to ask a suspect where he or she was at the time of a criminal offense. Offering a story about one’s timeline of events is a natural response to questioning by those investigating crimes. The alibi is perhaps the oldest documented criminal defense and is “frequently resorted to” by defendants (Gooderson, 1977, p. 1). Also, there is a substantial understanding of the alibi concept by laypersons outside of the legal system. It is surprising, therefore, that the alibis of suspects or criminal defendants have received little empirical attention, as very few studies have been published (Culhane & Hosch, 2004; Culhane, Hosch, & Kehn, 2008; Dahl, Brimacombe, & Lindsay, 2009; Olson & Wells, 2004; Sommers & Douglass, 2007).

Kerans, a Canadian Appeals Court Justice, noted instances when the alibi can actually harm a defendant and support the prosecution’s case (1982). Included in the disadvantageous alibis are those offered only after the trial has started and those proven to be false. By that logic, suspects must be accurate in their recollection to avoid having their alibi proven to be false; and for their own benefit, they must present this truthful alibi prior to the start of the trial. Even truthful alibis are viewed with great skepticism in the legal system. For example, Epstein (1964) found that more than 75% of prosecutors surveyed would proceed with a case even though an investigated alibi was perceived by said prosecutors to be true. A large number of these prosecutors intended to discredit the corroborator on the stand because of their general distrust of alibi corroborators. The study is nearly 50 years old, and results may vary with prosecutors of today. However, this reported prosecutorial behavior has been inconsistent with the United States Supreme Court’s (USSC) notion of how the legal system should deal with credible alibi corroborators. In Williams v. Florida (1970), one issue the USSC was asked to evaluate was Florida’s pretrial notification of alibi defense requirements. Chief Justice Burger, concurring with the majority opinion, argued that if the prosecutor has time to investigate alibi corroborators and if he or she finds them reliable and unimpeachable, the prosecutor will “very likely” dismiss the charges. This does not, however, appear to be the case in practice, as evident by Epstein’s results, although Epstein’s work was 6 years before the decision.

Alibi Research

Previous alibi research has centered on the evaluation of alibi statements. It has been argued that there are two types of alibi corroborators: those who are friends or relatives of the defendant (or have another interest in the case) and those who are not (Sullivan, 1971). The former will be referred to as motivated corroborators, because they are motivated to keep the defendant/suspect out of jail for reasons that are personal or financial. The latter or unmotivated corroborators have no such interest and would not be affected if the defendant/suspect were to be convicted or not.

Lindsay (1994) found that people do not report differences in credibility between motivated and unmotivated alibi corroborators. However, mock jurors’ behavior suggests otherwise. Lindsay and colleagues found that compared to having no alibi corroborator, jurors were more likely to convict defendants whose alibi corroborator was motivated (the defendants’ relative) and less likely to convict defendants with an unmotivated corroborator (Lindsay, Lim, Marando, & Cully, 1986). It is important to note that neither of the Lindsay articles utilized alibis as a focal variable.

Culhane and Hosch (2004) demonstrated that mock jurors were skeptical of an alibi corroborator who was not related to, but had a relationship with the defendant (a girlfriend). The conviction rates in conditions where the defendant’s girlfriend testified as an alibi corroborator were not different from the conviction rates of a defendant with no alibi corroborator. The finding that motivated corroborators are not effective alibis was confirmed by Olson and Wells (2004) and Hosch, Culhane, Jolly, Chavez, and Shaw (2010). Hosch and those colleagues found that only alibi corroborators with a distant relationship (regardless of whether the relationship category was biological, affinal, or social) to a defendant/suspect were believable. For example, a sister was viewed with greater skepticism than a second cousin. A wife was also viewed quite skeptically, whereas a second cousin by marriage was not. Finally, people who were not kin, but who had a social relationship with the defendant, elicited increased skepticism as the degree of closeness increased. A best friend was a poor alibi corroborator in the raters’ opinions, whereas an alibi from a total stranger was viewed positively.

Sullivan’s categorizations of corroborators have some overlap with a typology laid out by Olson and Wells (2004). Olson and Wells argue that there are three types of possible corroborators: (1) motivated familiar, (2) nonmotivated unfamiliar, and (3) nonmotivated familiar. The latter two differ in their knowledge of the suspect or defendant. Theoretically, the three types should be most credible in the reverse order above, because familiarity with the suspect by an unmotivated individual should increase accuracy, and thus increase the reliability of the alibi. However, it appears that familiarity between the unmotivated alibi corroborator and the suspected individual actually increases skepticism. Olson and Wells (2004) found that a store clerk who recognized the suspect (one accused of a crime having nothing to do with the clerk) and was familiar with the individual was less believable as an alibi corroborator than a total stranger. Both individuals testified that the suspect could not have committed the crime. It was expected that familiarity with the accused would be interpreted as increased accuracy in the ability to identify and recall the suspect. The addition of accuracy should lead the clerk to be the most believable alibi corroborator. A store clerk was clearly unmotivated, but he or she was not more believable even with his or her increased accuracy.

We can equate this to the similar difference in perceived lying by a “familiar stranger” (Milgram, 1977; Pearce, 1980) and a total stranger found by Hosch et al. (2011). A familiar stranger is clearly more knowledgeable about who the suspect is than a total stranger and should be seen as more accurate when asked to recall seeing the suspect; but the total stranger was viewed less skeptically by mock jurors. These findings lend further support to the notion that increased interpersonal distance equals increased believability.

Other research has examined context and its relationship to alibi evaluations. It has been noted that police investigators would be the most likely to initiate the recall of and evaluate the accuracy of an alibi (Burke, Turtle, & Olson, 2006). This can often occur in the beginning of an investigation and force suspects to produce statements that are not very well thought out or developed. Other participants in the criminal justice system (e.g., judges, lawyers, jurors) may be exposed to a story that has been retold from the initial questioning. Sommers and Douglass (2007) conducted two studies on the effect that context (police investigation vs. criminal trial) has on alibi believability. In both studies, participants in the police investigation condition rated the alibi as significantly stronger and more credible, but the likelihood of guilt judgments did not differ for either condition.

Changes in Alibi Statements

The repeated use of an alibi at different stages of the criminal justice process leads to questions about alibi consistency. Research has examined the frequency of changes in an alibi statement. Olson and Wells (2003) questioned individuals about their whereabouts for particular times and dates. By recording a participant’s alibis for two dates, one recent (2 days) and one distant (30 days), at an initial session and conducting follow-up interviews after a delay, Olson and Wells were able to record change in an alibi statement. Their results showed that between 30% and 45% of individuals make some change to an initial alibi statement with higher change percentages coming from alibis that were more distant in time (Olson & Wells, 2003). These changes were not necessarily complete shifts in the participant’s stories, but they were modifications to some aspect of the story. This work was furthered by Culhane, Kehn, Horgan, Meissner, and Hosch (2008), who found a significant difference between changes made to true and false alibi statements. Specifically, when participants were asked to fabricate an alibi for a particular time, they made more changes in the details provided in their story than those who were asked to tell the truth. However, the difference between the two had a small effect size (η² = .02). Like Olson and Wells (2003), very few individuals in either group made fundamental changes to their story, and the number of cases with detail changes was very high (>50%).

The perception of and reaction to these “new” alibis is still unknown. It would seem that there are pros and cons to changing one’s alibi statement. It may be beneficial to a suspect if an alibi were to be strengthened (e.g., an additional person was included as a corroborator, a “new” corroborator was not as vested in the suspect). Conversely, if the alibi statement was changed from an unmotivated to a motivated corroborator or the statement initially included an alibi corroborator, but the new statement had no alibi corroborator, the suspect/defendant would be at greater risk to be disbelieved. An additional possibility is that any changing of an alibi could be viewed as evidence of lying and that any change to one’s statement has a negative impact toward the believability of the individual’s alibi. It may be then, that the best alibi is one that is unwavering.

Alibi Evaluators

As discussed above, the officer may be the first to encounter an alibi, whereas laypersons may encounter one as an evaluator in a jury trial. In between are those who intend to pursue careers in law enforcement, yet have not received any formal training. Previous research by Kassin, Meissner, and Norwick (2005) compared the officers and students and their ability to detect deception in statements made by inmates. Their findings suggested minimal differences between the two groups in their abilities to do so, but those differences actually favored students. Culhane, Hosch, and Heck (2008) compared the endorsement of interrogation techniques by law enforcement officers, future officers (students who intended to enter law enforcement), and laypersons. Their findings indicated multiple differences between the three groups. For example, officers had increased willingness to use passive or nonthreatening methods to garner cooperation compared to other groups. Here, the evaluation of alibis by different participants in the justice system is of interest. Limited evaluation research has been conducted with vested parties in the criminal justice system. Specifically, police officers, lawyers, and judges also evaluate alibi corroborators, yet most research has focused on undergraduates serving in “mock” roles. It is important that data be gathered in a more applied fashion, using other interested individuals. However, no one to date has examined this issue.

Overview of this Research

This article presents three experiments to evaluate the impact of changed alibi statements. In Experiment 1, we presented three groups of individuals with a form describing various types of alibi statements to be evaluated. In Experiment 2, we expanded those statements into the context of nine police interrogations and asked the same three groups to make determinations of guilt or innocence. In Experiment 3, we utilized a single interrogation case with nine variations tested on one of the three groups. The design of Experiments 1 and 2 were a 3 (alibi: Strengthened, Maintained, Weakened) × 3 (evaluator: Current Officer, Future Officer, Layperson) mixed design. Experiment 3 was a between-subjects design using the three alibi statement possibilities from Experiments 1 and 2 (Strengthened, Maintained, Weakened).

Experiment 1

Questionnaires were administered to current law enforcement officers, future law enforcement officers, and laypersons. To directly compare beliefs and behaviors toward changes in an alibi, an Alibi Statement Belief Form was developed and submitted to the respondents. This form was designed to gauge the believability of different alibi statements that either had been changed (strengthened or weakened) or had been maintained. There are numerous changes that could be made with respect to the alibi statement. Issues such as time, place, and activities could all be altered in the course of time. We chose to focus on a single aspect of change, changes to the alibi corroborator. As previous research has explored differences in alibi corroborator motivation, this article focuses on the impact of change only. Therefore, alibi corroborators, both motivated and unmotivated, were chosen because they received skepticism ratings by independent evaluators that were the closest to 50%, as found by Hosch et al. (2011). These skepticism ratings were participants’ perceptions of how likely the corroborator was to lie on a defendant’s behalf. It was felt that choosing a corroborator with an extremely high or low skepticism rating would have too much effect on the manipulations because motivated and unmotivated alibi corroborators have different patterns of skepticism ratings. The participants also responded to the Alibi Belief Questionnaire (ABQ; Culhane, Hosch & Kehn 2008), which is composed of three factors and is discussed later. The ABQ is independent of the aforementioned Alibi Statement Belief Form.

Hypotheses

There are competing hypotheses with respect to alibi beliefs. Previous alibi research suggests that mock jurors do not self-report biases toward various aspects of alibis (Lindsay, 1994). This would suggest that there should be no difference in the evaluations of the changed or maintained alibi statements. However, other social psychological research, which may be more relevant to the current focal variables, has examined the impact of consistency and persuasion. Consistency is defined by the communication of “the same position in all situations and to different recipients regardless of the specific circumstances/recipients” (Ziegler, Diehl, Zigon, & Fett, 2004. p. 353). Legal research has been mixed. As Brewer and Burke (2002) noted, previous studies have found inconsistency (a) influences defendant culpability judgments, (b) undermines the credibility of certain witnesses, and (c) is a cue in truth/lie judgments. Inconsistency can also influence the strength of the prosecution’s case and the likelihood that a defendant committed a crime, but not overall verdicts (Semmler & Brewer, 2002). However, when dealing with crime eyewitnesses, evaluators seem to overlook discrepancies in testimony if the witness is highly confident in his or her identification (Brewer & Burke, 2002; Lindsay et al., 1986).

If the suspect is trying to persuade an interrogator that he or she is innocent, consistency may be the best approach and the most believable. We believe this to be the case. It is important to note that previous evaluation research on inconsistencies manipulated testimony by eyewitnesses or by participants in the trial other than the defendant (Brewer & Burke, 2002; Brewer & Hupfeld, 2004; Semmler & Brewer, 2002). Thus, we still do not know how a defendant’s or suspect’s inconsistency would be evaluated. Participants should view changes to alibi statements with greater skepticism and thus should rate maintained alibi statements more favorably. Also, there should be no difference for alibi statements that were strengthened or weakened, because all change in consistency would be viewed with skepticism.

As for the between-group comparisons, hypotheses were drawn circuitously using relationships with authoritarianism and culpability of suspect/defendants. Specifically, officers have been shown to be more authoritarian (Genz & Lester, 1976), and authoritarians have been shown to find defendants more culpable in criminal proceedings (Narby, Cutler, & Moran, 1993). Therefore, it was expected that laypersons would differ from current and future officers. Laypersons should evaluate the suspects’ alibi statements more positively. Also, since previous data show that officers increase in authoritarianism over time (Evans, Coman, & Stanley, 1992), it was also expected that current officers would demonstrate a greater skepticism of alibi statements than the future officers. It is also important to recognize that the general skepticism (i.e., a floor effect of believability) of most alibis may in fact preclude significant differences. However, alibi evaluations are stronger in a police context than a courtroom context (Sommers & Douglass, 2007), and therefore we expected to see differences.

Finally, the three factors of the ABQ were expected to intercorrelate positively for all groups. Current officers and future officers should demonstrate greater suspicion of alibi corroborators, increased negativity toward alibi changes, and reduced propensity to provide false alibis. However, these are anecdotal expectations, and the tests of the ABQ’s three factors are purely exploratory, as no other work has examined these issues with samples other than undergraduates.

Method

Participants

A total of 350 individuals completed the survey. Of those, 75 were current or previous law enforcement officers (39 police officers, 26 federal agents, 10 others). The remaining participants were criminal justice and introductory psychology students. One hundred and twenty nine students reported intentions to enter law enforcement (future law enforcement) and 146 reported no such intention (laypersons). The majority (67.4%) of future law enforcement participants reported intentions of working in federal positions (i.e., FBI, DEA, U.S. Marshals). Twenty students (15.5%) indicated they were interested in being a police officer as a future profession and 22 (17.1%) were either undecided or interested in other fields. Participants recruited from psychology and criminal justice classes received credit toward their class research requirement or extra credit toward their course grades. The current officers were each paid ten dollars ($10) for their participation.

The mean age of all participants was 26.39 (SD = 9.63). The self-reported ethnicity of the groups was: 74.3% Hispanic, 18.6% non-Hispanic White, 4.9% non-Hispanic Black, and 2.3% other ethnicities. The gender composition of the sample was 41.1% female and 58.9% male.

The demographics of the three groups were somewhat different. The current law enforcement officers had a mean age of 38.65 (SD = 10.40) and the majority of the sample was male (92%). The ethnicity of the officers was mostly non-Hispanic White (50.7%), followed by Hispanic (38.7%), and other ethnicities (10.6%). The officers had an average of 12.59 years (SD = 8.97) on the job. The majority of the future law enforcement officers were male (56.6%) and were mostly Hispanics (88.4%), followed by non-Hispanic Whites (7.0%). The mean age of the group was 22.47 (SD = 4.36). Finally, the lay participant group was mostly female (56.2%), Hispanic (80.1%) and non-Hispanic White (12.3%), and of typical college age (M = 23.56; SD = 7.18). The vast majority of students (>92%) were U.S. citizens.

The ethnicity differences for the groups were significant, χ²(10, N = 350) = 77.73, p < .001, Φ = .47. However, this result should be interpreted cautiously as nearly half of the cells failed to meet minimum counts. Furthermore, age of the members in each group was significantly different, F(2, 347) = 139.67, p < .001, η² = .45. A post hoc test revealed that the current officers were significantly different from the other two groups, but the other groups were not significantly different from each other. The gender composition of the groups was also significantly different, χ²(2, N = 350) = 47.90, p < .001, Φ = .37.

Materials

The packet of questionnaires included a demographic form, the Alibi Beliefs Questionnaire, and the Alibi Statement Belief Form. In addition to biographical information, the demographic questionnaire included items about service in law enforcement, interrogation experience, whether the students intended to enter law enforcement, and if so, what branch (federal agency or state) the individual was considering.

The Alibi Beliefs Questionnaire (Culhane, Hosch & Kehn, 2008) was then presented. The original ABQ consisted of 23 items. “It provides a layperson’s view on a variety of topics related to alibis from the generation of a statement to its actual content” (Culhane, Hosch, & Kehn, 2008, p. 183). The data (Culhane & Hosch, 2005) from Culhane and colleagues’ (Culhane, Hosch & Kehn) 2008 article were subjected to a factor analysis revealing that the questionnaire had two primary factors: items designed to index respondents’ beliefs about particular types of alibi corroborators (4 items) and the participant’s own propensity to provide a false alibi (3 items). Two additional questions were added to one existing item to index participant beliefs about suspects/defendants who change alibis (a total of 3 items). Although participants completed the entire ABQ, we focused attention on these three factors. The alibi questionnaire was coded so that higher scores were indicative of (a) increased skepticism of alibi corroborators, (b) increased skepticism of changes made to alibis, and (c) decreased propensity to provide false alibis. The ABQ was scored on a 1 (strongly agree) to 6 (strongly disagree) scale. This measure was administered to extend the work of Culhane, Hosch & Kehn (2008) with samples other than undergraduates, but also to identify general attitudes about changed alibi statements.

Next, the packet included the Alibi Statement Belief Form. This one-page questionnaire had participants imagine that an individual is being interrogated about whom he (all suspects were males) was with at the time of a robbery. For example, “Suspect 9 continuously stated that he had no witness and he was alone.” The participant gave an opinion (9-point scale) about how very poor (1) or very good (9) the story was. The alibis were based on two of the possibilities that may occur when an alibi is changed (the alibi can be either strengthened or weakened) and the additional alternative that one’s alibi remained unchanged. Three versions of each possibility were presented for a total of nine alibi statements. Therefore, the participants rated the first of the nine possible alibi statements and then moved on to the next statement for evaluation.

Strengthened alibi statements consisted of (a) changing from a motivated alibi corroborator to an unmotivated alibi corroborator (best friend to a neighbor), (b) changing from no alibi corroborator to a motivated alibi corroborator (girlfriend), and (c) changing from no alibi corroborator to an unmotivated alibi corroborator (co-worker). Those with motivation to lie have consistently been shown to be perceived as weaker alibi corroborators (Culhane & Hosch, 2004; Hosch et al., 2011; Olson & Wells, 2004); hence we identify a change from motivated to unmotivated as a strengthened statement. Likewise, we consider the opposite (unmotivated to motivated) to be a weakened statement. Weakening alibi statements consisted of (a) changing from an unmotivated alibi corroborator to a motivated alibi corroborator (neighbor to a first cousin), (b) changing from a motivated alibi corroborator (first cousin) to no alibi corroborator, and (c) changing from an unmotivated alibi corroborator (co-worker) to no alibi corroborator. Finally, three types of maintaining or unchanged statements were used: (a) asserting no alibi corroborator consistently, (b) keeping the same unmotivated alibi corroborator (neighbor), and (c) keeping the same motivated alibi corroborator (half-sister). Multiple types of unmotivated and motivated corroborators were chosen because of the within-subjects methodology. We hoped this would have prevented participant bias and reduced the opportunity to note that the statements were similar to each other. Again, the different types were matched using skepticism data from Hosch et al., (2011). This gives us confidence that the chosen relationships are indeed motivated (cousin, half-sister, girlfriend, and best friend) and unmotivated (co-worker and neighbor). The nine statements were subsequently averaged and collapsed into their respective type (strengthened, weakened, maintained).¹

Procedure

The introductory psychology participants were recruited via Internet sign-up forms, and they completed the questionnaires in a laboratory setting. Some of the criminal justice students who completed the measures in class were strictly volunteers, but most were offered extra credit toward their coursework by the professor of the class. All participating students signed a consent form and were handed the packet of questionnaires. After they participated, they were debriefed, they had any questions answered, and they were dismissed.

Officers’ data were collected via two strategies. Some completed the survey at the monthly meeting of the El Paso Municipal Police Officers’ Association, but the majority of participants were recruited at the officers’ respective stations or offices. The current officers’ data were collected in three locations: El Paso, Texas; Lake City, Florida; and Totowa, New Jersey. Multiple locations had to be used since there were not enough willing participants from any one area to meet the a priori power analyses requirements. In addition, these areas agreed to cooperate in the research.

Data analysis

Correlations were obtained for the three factors of the ABQ. Because of the discrepancy of age and gender between some of the groups, preliminary correlations between the demographics, the ABQ, and alibi ratings were conducted. Although gender was not significantly correlated with any factor of the ABQ, largest r(348) = .062, p = ns, participant age was a significant correlate of the propensity to provide a false alibi, r(347) = .31, p < .001. Because of the significance of this correlation, age and gender were used as covariates in analysis of covariance (ANCOVA) for the ABQ.

The analysis of the alibi belief statement form was a 3 (Current Officers, Future Officers, and Laypersons) × 3 (Strengthen, Weaken, Maintain) mixed design. The between-subjects factor was the participant groups. Within-subjects factors were the change/no change categories.

Four planned comparisons were conducted using a Bonferroni correction. The comparisons were chosen based on the variables of interest. Two comparisons examined the between-groups factor. The officers and future officers were compared to laypersons; and the officers and future officers were evaluated via a planned comparison. For the within-subjects factors, there was a change versus no change comparison, which combined the strengthening and weakening cases and compared them to the maintain cases. The strengthening and weakening cases were then compared to each other.

Results

Alibi Belief Statement Form

The mean believability of each alibi statement from the three groups is presented in Table 1. The overall mean of the ratings was 4.26 (SD = .95). A mixed-design analysis of variance (ANOVA) indicated a significant within-subjects variable difference for the three types of statements, F(2, 684) = 831.18, p < .001, η² = .71. There was a clear benefit to maintaining one’s alibi. Maintained alibis (M = 6.59) were viewed far more positively than were alibi statements that were changed (M = 3.09), t(344) = 32.93, p < .001, η² = .76. The next planned comparison was between the ratings of the three strengthened and three weakened alibi statements. This test produced a small, yet statistically significant, difference for the strengthened (M = 3.18) and weakened alibi changes (M = 3.01), t(347) = 3.28, p = .001, η² = .03. This is most likely because of the within-subjects nature of the analysis.

Table 1.

Mean Beliefs of the Alibi Statements in Experiment 1

	Overall		Laypersons		Future Officers		Current Officers
Alibi	M	SD	M	SD	M	SD	M	SD
Strengthened	3.18	1.26	3.16	1.28	3.21	1.23	3.15	1.31
Maintained	6.59	1.59	6.75	1.56	6.53	1.67	6.39	1.49
Weakened	3.01	1.30	3.08	1.35	3.04	1.22	2.80	1.34
			4.33	0.95	4.26	0.94	4.12	4.12

Note. Scores are on a 1 (very poor) to 9 (very good) rating scale.

Next, the planned comparisons of officers/future officers versus laypersons and officers versus future officers were conducted. Contrary to expectations, there was no overall difference in belief judgments between the three groups, F(2, 342) = 1.16, p = ns, η² = .01. Neither the comparison of officers/future officers versus laypersons, t(341) = 1.11, p = ns, nor the officers versus future officers, t(200) = 1.05, p = ns, showed significant differences. Likewise, there was no significant interaction between the two variables, F(4, 684) = .76, p = ns, η² = .00.

Alibi Beliefs Questionnaire

The raw mean scores on the ABQ subscales are presented in Table 2. With age and gender entered as covariates, the future officers were significantly more skeptical of alibi corroborator types and changes to alibi statements, F(2, 345) = 4.43, p = .013, η² = .01 and F(2, 343) = 13.27, p < .001, η² = .07, than were laypersons. Neither was different from current officers. An ANCOVA test found no differences for the propensity to provide false alibis, F(2, 337) = 2.61, p = .075, η² = .02.

Table 2.

Mean Scores for ABQ Subscales in Experiment 1

	Laypersons		Future Officers		Current Officers
Factor	M	SD	M	SD	M	SD
ABQ Witness	11.40	3.02	12.43	3.20	11.69	3.32
ABQ Change	11.12	2.75	12.13	2.68	12.23	2.84
ABQ Lie	13.09	3.35	13.67	3.51	15.45	2.86

Note. ABQ = Alibi Belief Questionnaire.

The ABQ subscales were correlated with the Alibi Statement Belief Form evaluations. Skepticism toward alibi statements that changed was significantly correlated with the Alibi Statement Belief Form ratings of alibis that were strengthened, r(340) = -.34, p < .001, and weakened, r(340) = -.35, p < .001, but not correlated with the belief ratings of alibis that were maintained, r(340) = -.07, p = ns. The skepticism about the alibi corroborators factor correlated with strengthened, r(340) = -.13, p < .01, and maintained statements, r(340) = -.24, p < .001, but not the weakened alibi ratings, r(340) = -.07, p = ns. The negative correlations indicate that as the factors of skepticism toward changed alibi statements and skepticism about alibi corroborators increased among participants their ratings of the alibi story decreased and vice versa. The factor of propensity to lie was not correlated with any statement ratings, largest r(340) = -.07, p = ns.

The correlations among ABQ subscales for the three groups are presented in Table 3. In addition, the internal consistencies (α) of each measure are presented in the diagonal of the table. For laypersons, the subscales of the ABQ were intercorrelated as expected. The internal consistencies of the three ABQ factors ranged from .73 to .77. The future officers had internal consistencies ranging from .76 to .86. However, unlike the laypersons, the intercorrelations of the ABQ subscales produced a significant relationship only between skepticism toward types of alibi corroborators and skepticism toward changes in alibis. Finally, the internal consistencies for current law enforcement officers ranged from .82 to .85. Only the correlation of the propensity to provide false alibis and skepticism toward changes in alibis was not significant for current officers.

Table 3.

Intercorrelations Between the ABQ Factors

	Laypersons (n = 139)			Future Officers (n = 126)			Current Officers (n = 68)
Factor	1	2	3	1	2	3	1	2	3
ABQ Witness	(.73)			(.76)			(.84)
ABQ Change	.35***	(.74)		.43***	(.79)		.39***	(.82)
ABQ Lie	−.18*	−.15*	(.77)	−.13	.09	(.86)	−.24*	−.06	(.85)

Note. Cases excluded listwise. Internal consistencies are in parentheses. ABQ = Alibi Belief Questionnaire.

p < .05. ***p < .001.

Summary of Experiment 1

Experiment 1 was the initial investigation of the impact of an alibi statement change. Three groups of individuals were tested: current law enforcement officers, future law enforcement officers, and laypersons. The various groups showed no difference in their evaluations of alibi statements. The results of Experiment 1 indicated that suspects who changed their alibi statement were viewed with increased suspicion. This suspicion of a changed alibi seemed to have negated the “positive” of an improvement in a suspect’s alibi (e.g., changing from no alibi corroborator to having an alibi corroborator). The cynicism toward suspects’ statements was best seen in the future officer group. There is a strong possibility that experience on the force can reduce initial biases. The factors of the ABQ were shown to correlate with alibi skepticism. This initial research was then followed by Experiment 2, in which new samples of the same three groups of individuals were tested to determine how they would react to alibi statement changes made in a simulated police interrogation.

Experiment 2

The purpose of Experiment 2 was to compare the reported beliefs of the groups from Experiment 1 and their judgments of simulated interrogations in which alibis were changed or maintained. Participants in Experiment 2 were a new subset of police officers, criminal justice students, and psychology students. They evaluated a series of criminal suspects’ interrogations, made judgments about the suspect’s guilt or innocence, and rated personal characteristics of the suspect.

Hypotheses

We expected that all groups would be more punitive with their verdicts and perceive the individual as more likely to have committed the crime for suspects who changed their statement, no matter the type of change, as opposed to those who consistently maintained their statements. Based on the results of Experiment 1, the comparison of officers/future officers to laypersons was not expected to be significant. Likewise, there was no anticipated difference between officers and future officers. These comparisons were done to replicate and extend those of Experiment 1.

Method

Participants

Based on a GPOWER a priori analysis (ANOVA F test for differences among means) with power equal to .80, Type I error rate of .05, and medium expected effect size, 53 participants from each group were needed to assess differences among the groups (Faul & Erdfelder, 1992). The recruitment proceeded identically to that in Experiment 1. Laypersons, law enforcement officers, and students who planned to become law enforcement officers were solicited as participants.

A total of 211 individuals completed the study. Of those, 52 were current law enforcement officers (40 police officers, 11 sheriff’s department, and 1 military police). The remaining participants were criminal justice and introductory psychology students. Seventy-six were classified as future law enforcement and 83 were categorized as laypersons. The majority of future law enforcement participants reported intentions of working in federal positions (49), followed by police/sheriff departments (16), and other fields (11). Participants recruited from psychology and criminal justice classes received credit toward their class research requirement or extra credit toward their course grades. The current law enforcement officers were each paid ten dollars ($10) for their participation.

The overall mean age of the participants was 26.85 (SD = 9.53). The self-reported ethnicity of the groups was: 77.3% Hispanic, 17.5% non-Hispanic White, 3.8% non-Hispanic Black, and 1.4% other ethnicities. The gender composition of the sample was 48.3% female and 51.7% male. The demographics of the three groups are now discussed separately.

The current law enforcement officers had a mean age of 38.81 (SD = 8.58), and the majority of the sample was male (92.3%). The ethnicity of the officers was mostly Hispanic (67.3%), followed by non-Hispanic White (26.9%), and other ethnicities (5.7%). The officers had an average of 12.23 years (SD = 8.47) on the job. Finally, the officers had a great range in the number of interrogations they reported having participated in (as low as 1 and as many as 1,000). The officers averaged 188 (SD = 243.68) interrogations.

Future law enforcement officers were mostly male (53.9%) and were mostly Hispanic in this sample (84.2%). The second most frequent ethnic group was non-Hispanic Whites (14.5%). The mean age of the group was 22.09 (SD = 4.35). Finally, the lay participant group was mostly female (75.9%), Hispanic (77.1%) and non-Hispanic White (14.5%), and college age (M = 23.70; SD = 6.91).

The ethnicity differences for the groups were not significant, χ²(10, N = 211) = 15.30, p = ns, Φ = .27. However, this result should be interpreted cautiously as more than half of the cells failed to meet minimum counts. Furthermore, age of the members in each group was significantly different, F(2, 208) = 114.10, p < .001, η² = .52. A post hoc test revealed that the current officers were significantly different from the other two groups, but the other groups were not significantly different from each other. The gender composition of the groups was also significantly different, χ²(2, N = 211) = 59.82, p < .001, Φ = .53. Because of the discrepancy of age and gender between some of the groups, these variables were entered as covariates in the analyses.

Materials

The participants completed the same demographics form used in Experiment 1. Nine interrogation transcript synopses were read by the participants. Although it was not made explicitly clear that these were interrogations and not interviews, the person of interest was referred to as a suspect and was directly accused of the crime. Again, the manipulations were based on two of the possibilities that may occur when an alibi is changed (the alibi can be either strengthened or weakened) and the additional occurrence of one’s alibi remaining unchanged. Three types of each possibility (manipulating the corroborator motivation) were presented for a total of nine interrogations. The alibis mirrored those from Experiment 1.

The stimulus interrogations were developed from actual police interrogations, but they were altered to achieve experimental control. The packet described their task as that of a “detective” who was investigating robbery crimes and that they would read a series of interrogation excerpts that were chosen because of their similarities. The packet also contained the following statement, “All suspects presented may have been found to be innocent, all suspects may have been found to be guilty, or some may be guilty and some may be innocent. It is up to you to make the determination.” All interrogations included a denial of guilt by the suspect and two pieces of circumstantial evidence (e.g., a gun similar to the type of one used in the robbery, large amount of cash found on suspect) against the suspect. The only true difference in the transcripts was the suspect’s alibi information. Alibi statements were presented via the questioning of the interrogating officer within the transcript. Specifically, the officer pointed out the alibi and whether it was changed. There was no indication of whether the suspect changed the story as a result of his own memory, or through the detective’s questioning. All change conditions included the suspect’s explaining the change as, “I made a mistake.” The transcript makes changes in the alibi statement seem as if they have taken place during the same interrogation session. In the maintained condition, the suspect agreed with the officer’s alibi clarification.

After reading the three-page synopsis for each interrogation, the participants read a summary of the charge (all armed robbery, but an explicit definition of armed robbery was not provided), alibi evidence (a repetition of the information in the transcript), and circumstantial evidence. The page following the summary contained a judgment of guilt or innocence and the participant’s confidence in that decision on a 1 (very uncertain) to 5 (very certain) scale. These two items were followed by a likelihood of crime commission item. This item had a range from 1 (very unlikely) to 9 (very likely). Finally, a series of characteristic judgments (also using a 1 to 9 format) about the suspect that included honesty, credibility, believability, persuasiveness, knowledge, competence, intelligence, and detail was completed. All scenarios were pilot tested using introductory psychology students for believability. As expected, the various interrogations were rated above seven on a nine-point scale for believability. We did not pilot test information about the alibi corroborators, as extensive data collection from similar pools of participants was collected by Hosch et al. (2011).

Procedure

The data were collected in small groups ranging in number of participants from 2 to 30. The packet of experimental materials was administered, which contained a demographics questionnaire and the interrogation excerpts. After reading each individual transcript, the participants provided their dichotomous guilt or innocence decision, their confidence in that decision, their judgment of the likelihood that the accused actually committed the crime, and the series of characteristic judgments about the suspect.

Data analysis

As with Experiment 1, the complete design was a 3 (Current Officers, Future Officers, Laypersons) × 3 (Strengthen, Weaken, Maintain) mixed design. The between-subjects factor was the three groups. Within-subjects factors were the change/no change categories. There were three statements per change/no change category. However, they were averaged and collapsed into a single value for each category.

The same four planned comparisons that were conducted in Experiment 1 were repeated, once for likelihood of commission and once for judgments of guilt or innocence. Again, current officers and future officers were first compared to laypersons. Then, current officers and future officers were compared to each other. The within-subjects comparisons included the change versus no change comparison and the strengthened versus weakened cases comparison. The process was the same for guilt/innocence judgments.

Results

Conviction rates

As Table 4 shows, the collapsed conviction rates of the suspects ranged from 45.9% to 72.4%. The overall conviction rate was 60.65% (SD = 22.85). The participant’s average confidence in their verdicts was not correlated to their average verdict, r(201) = .10, p = ns. There were significant differences among the three groups in their confidence judgments, F(2, 200) = 3.66, p = .027, η² = .04. Controlling for age and gender, the future officers (M = 3.69) were significantly more confident in their overall decision than the laypersons (M = 3.49). That is, regardless of guilty or innocent judgments, the future officers were more confident. Current officers (M = 3.59) were not significantly different from either student groups.

Table 4:

Conviction Rates (in %) as a Function of the Suspects’ Alibi Statements in Experiment 2

Alibi	Overall	Laypersons	Future Officers	Current Officers
Strengthened	62.7	58.1	61.8	71.2
Maintained	50.6	45.9	49.6	60.0
Weakened	69.2	65.0	72.4	71.2
		56.1	61.3	67.1

As was found in Experiment 1, the overall mixed-design test showed a significant within-subjects difference for the various statement types, F(2, 408) = 24.19, p < .001, η² = .11. The planned comparison of change (M = 65.7%) and no change (M = 50.6%) was significant, t(206) = 6.82, p < .001, η² = .18. However, the comparison of verdicts for strengthened and weakened alibi statements was not significant in this study (with the Bonferroni correction), t(208) = 2.65, p = ns, ² = .03.

Unlike Experiment 1 there was a significant difference in guilty judgments for the three groups, F(2, 204) = 3.72, p = .026, η² = .04. As expected, the planned comparisons of future officers/current officers to laypersons and future officers to current officers were not significant, t(205) = 2.32, p = ns, η² = .03, and, t(124) = 1.38, p = ns, η² = .02, respectively. However, a post hoc test of the ANOVA revealed that the overall guilty judgments of current officers (67.1%) differed from those of laypersons (56.1%). As in Experiment 1, there was no significant interaction between the two variables, F(4, 408) = 1.22, p = ns, η² = .01.

Likelihood of commission

Recall that participants were also asked to make a judgment of the likelihood that the suspect actually committed the crime as well. Higher numbers represent a greater suspicion that the suspect committed the crime. The same seven planned comparisons were conducted with these judgments. Table 5 presents the mean responses to this measure for each alibi. There was a significant difference in the change (M = 5.68) and no change comparison (M = 5.19), t(176) = 5.19, p < .001, η² = .13. Like verdicts, there was no significant difference for likelihood judgments of strengthened or weakened alibis, t(182) = 1.11, p = ns, η² = .01.

Table 5:

Mean Likelihood of Commission Judgments in Experiment 2

	Overall		Laypersons		Future Officers		Current Officers
Alibi	M	SD	M	SD	M	SD	M	SD
Strengthened	5.63	1.27	5.54	1.19	5.67	1.20	5.72	1.49
Maintained	5.25	1.32	5.29	1.25	5.09	1.23	5.43	1.56
Weakened	5.79	1.26	5.66	1.24	6.02	1.25	5.64	1.27
			5.47	0.87	5.52	0.95	5.72	1.12

Note. Scores are on a 1 (very unlikely) to 9 (very likely) rating scale.

The next two comparisons examined the between-group factor for differences in likelihood judgments. Again, neither the comparison of officers/future officers to laypersons, t(132) = .29, p = ns, η² = .00, nor the comparison of officers to future officers, t(105) = .29, p = ns, η² = .00, was significant.

Suspect ratings

Recall that participants were asked to rate each of the nine suspects on eight traits: credibility, honesty, persuasiveness, knowledge, competence, intelligence, believability, and detail. The eight traits were averaged for all nine suspects and entered as predictors in two hierarchical regressions, one for verdicts (the nine verdicts averaged) and one for likelihood of commission collapsed across the nine. For verdicts, step one with age and gender entered as predictors was not significant, R² = .007. The inclusion of the eight evaluation judgments resulted in a significant increase (ΔR² = .443), F(8, 173) = 17.44, p < .001. Only believability, β = -.515, and honesty, β = -.311, were significant, however.

For likelihood of commission, the results mirrored the verdict regression. The first step with age and gender was not significant (R² = .013) and there was a significant change in R² (.347) in step two, F(8, 150) = 11.49, p < .001. Again, only believability, β = -.438, and honesty, β = -.395, were significant.

Summary of Experiment 2

Experiment 2 is further confirmation that changes made to an alibi statement are detrimental to the credibility of the suspect’s story. Experiment 2 showed that when the suspect made a change to his alibi statement, all but one case resulted in the participants (a) judging the suspect to be guilty more often and (b) raising their judgment of the likelihood the suspect committed the crime. Interestingly, guilty judgments were significantly different in two of the groups. The current officers had higher rates of guilty decisions than did laypersons. It is important to note that there were always two pieces of circumstantial evidence used, but the evidence had to change for each case. If it did not change, we ran the risk of the cases being too similar and possibly allowing participants to uncover the true nature of the manipulation. Still, the circumstantial evidence may have been the cause of the findings.

Experiment 3

A third study was conducted to replicate the findings from Experiments 1 and 2 in a between-subjects methodology. The within-subjects data was compelling, but the design was also a limitation. Additionally, participants may have used relative judgment strategies in the first two studies, whereas, the between-subjects design will permit for absolute judgments. Because participants will see one, and only one, possible alibi corroborator scenario, they will make decisions based on it alone. This will also increase the ecological validity of the materials, because jurors would never be asked to determine which of nine alibi statements was most credible.

Method

Participants

Four hundred and twenty three undergraduates from a Rocky Mountain institution comprised the final sample. We did not collect police officers or future police officers for this project, as student decisions were very similar in Experiments 1 and 2. The sample was mostly female (64.8%) and of typical college age (M = 19.81; SD = 3.04). Participants self-identified their ethnicity as follows: 88.2% non-Hispanic White, 6.9% Hispanic, 1.9% Asian, 1.7% African American, and 1.4% other. Students participated as part of their introduction to psychology course requirement.

Materials and procedure

One transcript from Experiment 2 was randomly chosen and used as the interrogation. Four motivated corroborators (girlfriend, half-sister, cousin, and best friend) and two unmotivated corroborators (neighbor and co-worker) were counterbalanced and used in the various scenarios.² The possibility of no alibi corroborator was also included, and a total of 35 possible conditions were developed. As in Experiments 1 and 2, Experiment 3 utilized strengthened, weakened, and maintained alibi statements. The impetus for why the change took place was given as the detective saying, “First you said __________. Now, when I question you about it for more details you say__________”. The interviewee says, “I just made a mistake”.

After reading the transcript, participants made a judgment of guilt or innocence, a judgment of likelihood of guilt (1 = very unlikely and 9 = very likely), and rated the suspect on a variety of characteristics (honest, trustworthy, credible, etc.). Next, participants completed a manipulation check sheet that asked specific questions about the suspect’s alibi. Questions included “Did the suspect have an alibi corroborator (someone he asserted was with him at the time of the crime)?” and “Did the suspect change his story?” Follow-up explanations were requested after answering these questions asking about who the suspect claimed to be with at the time of the crime. If a participant incorrectly answered these questions, his or her data were removed from the final count. Thus, 100% of the participants’ data are from those who correctly recalled the details of the alibi. Finally, participants completed the ABQ.

Results

Overall, the mock interviewers judged the suspect guilty 31.2% of the time. This reduction may be a function of the different sample composition. Participants rated the suspect guilty most often when a weakened alibi statement was presented (38% indicated guilty). The least frequent guilty verdicts were seen in the maintained condition (23%), followed by strengthened alibi statements (32%). Logistic regression of guilt/innocence judgments using dummy codes of the three statements (maintained statements as the referent) was significant, Model χ² (2, N = 423) = 7.44, p = .024, Nagelkerke R² = .025. The comparisons of statements indicated that the 15% difference between weakened and maintained stories was significant, B = .709, SE = .264, Wald (1) = 7.22, p = 007. The 9% difference between strengthened and maintained alibi statements was not significant, B = .428, SE = .269, Wald (1) = 2.54, p = 111.

Participants’ likelihood of guilt decision was entered into a one-way ANOVA. Again, there were significant differences by statement type, F(2, 421) = 3.14, p = .044, η² = .02. Specifically, the mean likelihood of guilt for maintained statements was 4.54, whereas strengthened and weakened were 4.92 and 4.95, respectively. The ratings of the interviewee were collapsed into a single variable, with higher numbers being more positive evaluations, and compared for the three types of alibi statements. Although the pattern was as expected (maintained = 4.65, strengthened = 4.48, and weakened = 4.37), the small differences were not significant, F(2, 420) = 2.24, p = .107, η² = .01.

General Discussion

Alibi Statements

There is a clear benefit to maintaining one’s alibi statement (i.e., participants viewed change negatively). Furthermore, the difference between evaluations of strengthened and weakened alibi statements was minimal at best in Experiment 1 and not significant in Experiment 2 or 3, suggesting that change is a negative regardless of type or direction. These findings support social psychological research that indicates that when a source is consistent, he or she is also more persuasive (e.g., Moscovici, Lage, & Naffrechoux, 1969).

The pattern of results among the three groups of participants was surprising. It was expected that the collection of officers and future officers would have more negative beliefs toward alibi statements than would laypersons. And, future officers were expected to have the most negative beliefs and behaviors when compared to the current officers. These hypotheses were not supported in Experiment 1. However, there were significant differences in the post hoc tests of participants’ behaviors. Specifically, the overall guilty judgments were higher for current officers than laypersons. The future officer group was not significantly different from either group. It is possible that the limited circumstantial evidence was an issue for the future officers, as they had not had experience with actual investigations. Furthermore, the results of Experiment 3 may be more exaggerated in samples of current officers, because their behaviors seem more punitive in interview scenarios. However, it is important to note that there were substantial differences in the conviction rates from Experiments 2 and 3. This speaks to the methodological importance of the between-subjects design, as participants were not able to use relative judgment strategies when deciding guilt.

As discussed earlier, Sommers and Douglass (2007) found that alibis were rated stronger in police investigation conditions. Thus, in our work we utilized detective scenarios for the changes made to statements. It may be the case that in a criminal trial, changes would be treated even more harshly when examining rates of convictions. It is possible that under instances of cross-examination, the changes made to a statement would be even more salient to decision makers (i.e., jurors) and evaluated with a greater degree of skepticism. Therefore, our results may be underestimations of the impact of changed alibi statements.

Limitations

As noted above, the alibi change measure in Experiments 1 and 2 may have fallen prey to a relative judgment strategy by participants or to order effects of the presentation. Because the participants saw all alibi stories together, they may have been basing their judgments on what they had indicated previously. Although order was randomly assigned, there was not a counterbalancing of the items. Thus, it was necessary to conduct a study using a between-subjects design. The reduced effect size of the differences in alibi statements may be a result of the elimination of relative judgments.

There is an additional limitation with the use of corroborator types. Different alibi corroborators were used in the roles of motivated and unmotivated alibi corroborators. They were not counterbalanced across conditions in Experiments 1 and 2 and, as a result, the same individuals were repeatedly used in the same conditions. The effects could be attributed to the person rather than the condition. However, we would like to emphasize that corroborators were chosen specifically because they were previously assessed as being the most neutral by other participants’ judgments of skepticism. Furthermore, the corroborator types were counterbalanced in Experiment 3, and no differences were found when comparing the results of each.

Future Research

As these three experiments were the initial attempt to understand the impact that a change in one’s story has on credibility, much more research is needed to better understand the nuances of alibis both changed and maintained. In addition, there are several other questions about alibis that have yet to be answered. For example, how important is the timing of the alibi? Kerans (1982) has argued that offering an alibi after the trial has started is disadvantageous for a defendant. However, recent research by Dahl and colleagues (2009) has found that evaluations of strong alibis depended on the order of presentation of evidence, as recency effects take place in the decision-making process. That is, the mock jurors were more influenced by a strong alibi when it was one of the last pieces of information presented.

The interrogation represents one of the earliest opportunities to produce an alibi. Would an alibi presented during interrogation actually produce a different impact than one produced after the start of a trial? No empirical research exists to suggest that alibis presented earlier in the criminal investigation process are better for the suspected individual. In fact, when one considers the frequent changes made to innocent person’s alibis, it may be beneficial for the suspect to wait until he or she can confirm the alibi, even if that means waiting until a criminal trial. In the same vein, there are a host of other possible changes that one could make to the story (e.g., changing the location or context of the alibi, but not the corroborator). Furthermore, there is the possibility that changes involving no alibi corroborators are categorically different from other changes. However, no research to date speaks to the various changes that could be made in one’s alibi statement.

It was also interesting that the evaluation of alibis may correlate with ethnicity. Undergraduates in Experiments 1 and 2 were primarily of Hispanic descent, whereas Experiment 3 used mostly Anglo participants. The evaluation of alibis in different cultures could yield different results if explored systematically. It may be the case that Hispanics are more understanding of motivated alibi corroborators, because of the cultural expectation that you spend more time with family. On the other hand, Anglos may be less punitive to those who have friends as corroborators, because of their own propensity to utilize such individuals in a generation study (Culhane, Hosch & Kehn, 2008). The question remains unanswered.

Finally, physical alibi evidence, evidence that demonstrates the individual was in fact not at the scene of the crime at the time it occurred (e.g., store receipt, ticket stub), is apparently more difficult to obtain than an alibi corroborator but is reported by jurors to be as important (Culhane, Hosch & Kehn, 2008). Olson and Wells (2004) included physical evidence as a fundamental part of their taxonomy, but tested it only in a limited scope. Sargent and Bradfield (2004) found that strong alibis (i.e., a video surveillance of the defendant at the time of the crime) were more convincing than weak alibis (no security camera present and no corroborator). Clearly such evidence can be an effective defense tool. Physical alibi evidence may be important or wanted by mock jurors, however empirical research has shown how infrequently it will actually be presented (Culhane, Hosch & Kehn, 2008). Given that this evidence seems to be an important expectation, the importance of physical alibi evidence needs to be empirically explored. There are a multitude of possibilities in juror situations, particularly when weak alibi corroborators are present. A strong piece of physical alibi evidence (e.g., time-stamped materials) should counteract the “penalty” of presenting a weak alibi corroborator.

Footnotes

This research was conducted as the first author’s PhD dissertation at the University of Texas at El Paso. Portions of this research were presented as a paper discussion at the 2005 annual American Psychology-Law Society conference in La Jolla, California, and as a poster at the 2006 annual American Psychology-Law Society conference in St. Petersburg, Florida. Thanks are extended to several members of the Psychology and Criminal Justice Departments at the University of Texas at El Paso. Also, we would like to thank William G. Weaver, Chief William Adcox, Lieutenant Michael Hanna, and Detective Earl Arbogast of the UTEP Police Department, Dean Kinder of the El Paso Municipal Police Officer’s Association, Captain Tim Culhane of the Florida Highway Patrol, Edward Culhane, and the numerous research assistants who assisted with data collection and entry. Finally, thanks to Adrienne Freng, Cary Heck, Eddie Muñoz, and Narina Nunez for their insightful comments on an earlier draft of this article. Experiment 1 was supported by a grant from the Graduate School of the University of Texas at El Paso. Experiment 2 was supported by the NIMH M-RISP Grant No. 2 R24 MH47167-11.

Notes

Scott E. Culhane is an associate professor in the Department of Criminal Justice at the University of Wyoming. He received his doctorate in legal psychology from the University of Texas at El Paso. His research program focuses on jury decision making with a particular interest in the presentation of alibi witnesses. He is also researching the formation of alibi statements, the ability to detect lying in such statements, psychometric evaluations of psychological scales in minority samples, and serial murder.

Harmon M. Hosch is the Helen M. C. and J. Edward Stern Professor of Psychology at the University of Texas at El Paso. He earned his PhD in social and personality psychology from the New School for Social Research in New York City. He was a Fulbright scholar in Chihuahua, Mexico, from 1982 to 1983 and a visiting scientist in the role of director of the Law and Social Science Program of the National Science Foundation from 1996 to 1998.

References

Brewer

Burke

(2002). Effects of testimonial inconsistencies and eyewitness confidence on mock-juror judgments. Law and Human Behavior, 26, 353-364. doi:10.1023/A:101538052272210.1023/A:1015380522722

Brewer

Hupfeld

R. M.

(2004). Effects of testimonial inconsistencies and witness group identity on mock-juror judgments. Journal of Applied Social Psychology, 34, 493-513. doi:10.1111/j.1559-1816.2004.tb02558.x

Burke

Turtle

Olson

(2006). Alibis in criminal investigations and trials. In Toglia

M. P.

Read

J. D.

Ross

D. F.

Lindsay

R. C. L.

(Eds.), Handbook of eyewitness psychology: Vol. 1. Memory for events. Mahwah, NJ: Lawrence Erlbaum Associates.

Culhane

S. E.

Hosch

H. M.

(2004). An alibi witness’s influence on juror’s verdicts. Journal of Applied Social Psychology, 34, 1604-1616.

Culhane

S. E.

Hosch

H. M.

(2005). Changed alibi statements in mock evaluator decisions. Unpublished raw data. University of Texas El Paso, Center for Law and Human Behavior.

Culhane

S. E.

Hosch

H. M.

Heck

(2008). Interrogation technique endorsement by current law enforcement, future law enforcement, and laypersons. Police Quarterly, 11, 366-386.

Culhane

S. E.

Hosch

H. M.

Kehn

(2008). Alibi generation: Data from U.S. Hispanics and U.S. Non-Hispanic Whites. Journal of Ethnicity in Criminal Justice, 6, 177-199.

Culhane

S. E.

Kehn

Horgan

A. J.

Meissner

C. A.

Hosch

H. M.

(2008, March). Consistency in alibi generation: Data for true and false alibis. Paper presented at the annual meeting of the American Psychology-Law Society, Jacksonville, FL.

Dahl

L. C.

Brimacombe

C. A. E.

Lindsay

D. S.

(2009). Investigating investigators: How presentation order influences participant-investigators’ interpretations of eyewitness identification and alibi evidence. Law and Human Behavior, 33, 368-380. doi:10.1007/s10979-008-9151-y10.1007/s10979-008-9151-y

10.

Epstein

D. M.

(1964). Advance notice of alibi. Journal of Criminal Law, Criminology, and Police Science, 55, 29-38.

11.

Evans

B. J.

Coman

G. J.

Stanley

R. O.

(1992). The police personality: Type A behavior and trait anxiety. Journal of Criminal Justice, 20, 429-441.

12.

Faul

Erdfelder

(1992). GPOWER: A priori, post hoc, and compromise power analyses for MS-DOS [Computer program]. Bonn, Germany: Bonn University, Department of Psychology.

13.

Genz

J. L.

Lester

(1976). Authoritarianism in policemen as a function of experience. Journal of Police Science and Administration, 4(1), 9-13.

14.

Gooderson

R. N.

(1977). Alibi. London, England: Heinemann Educational.

15.

Hosch

H. M.

Culhane

S. E.

Jolly

K. W.

Chavez

R. M.

Shaw

L. R.

(2010). Effects of an alibi witness’s relationship to the defendant on mock jurors’ judgments. Law and Human Behavior. 35, 127-142. DOI: 10.1007/s10979-010-9225-510.1007/s10979-010-9225-5.

16.

Kassin

S. M.

Meissner

C. A.

Norwick

R. J.

(2005). “I’d know a false confession if I saw one”: A comparative study of college students and police investigators, Law and Human Behavior, 29, 211-227.

17.

Kerans

R. P.

(1982). Recent developments in the law concerning alibi and identification evidence. The Criminal Law Quarterly, 25, 47-65.

18.

Lindsay

R. C. L.

(1994). Expectations of eyewitness performance: Jurors’ verdicts do not follow from their beliefs. In Ross

D. F.

Read

J. D.

Toglia

M. P.

(Eds.), Adult eyewitness testimony: Current trends and developments (pp. 362-384). Cambridge, England: Cambridge University Press.

19.

Lindsay

R. C. L.

Lim

Marando

Cully

(1986). Mock-juror evaluations of eyewitness testimony: A test of metamemory hypotheses. Journal of Applied Social Psychology, 15, 447-459.

20.

Merriam-Webster (2012). Dictionary. Accessed at http://www.merriam-webster.com

21.

Milgram

(1977). The individual in a social world. Reading, MA: Addison-Wesley.

22.

Moscovici

Lage

Naffrechoux

(1969). Influence of a consistent minority on the responses of a majority in a color perception task. Sociometry, 32, 365-380.

23.

Narby

D. J.

Cutler

B. L.

Moran

(1993). A meta-analysis of the association between authoritarianism and jurors’ perceptions of defendant culpability. Journal of Applied Psychology, 78, 34-42.

24.

Olson

E. A.

Wells

G. L.

(2003, May). Alibi generation: Not so easy for the innocent. Poster session presented at the 15th annual meeting of the American Psychological Society, Atlanta, GA.

25.

Olson

E. A.

Wells

G. L.

(2004). What makes a good alibi: A proposed taxonomy. Law and Human Behavior, 28, 157-176.

26.

Pearce

P. L.

(1980). Strangers, travelers, and Greyhound terminals: A study of small-scale helping behaviors. Journal of Personality and Social Psychology, 38, 935-940.

27.

Sargent

M. J.

Bradfield

A. L.

(2004). Race and information processing in criminal trials: Does the defendant’s race affect how the facts are evaluated? Personality and Social Psychology Bulletin, 30, 995-1008.

28.

Semmler

Brewer

(2002). Effects of mood and emotion on juror processing and judgments. Behavioral Sciences & the Law, 20, 423-436. doi:10.1002/bsl.50210.1002/bsl.502

29.

Sommers

S. R.

Douglass

A. B.

(2007). Context matters: Alibi strength varies according to evaluator perspective. Legal and Criminological Psychology, 12, 41-54.

30.

Sullivan

T. P.

(1971). Presentation of the defense. Journal of Criminal Law, Criminology, and Police Science, 62, 153-172.

31.

Williams v. Florida, 399 U.S. 78 (1970).

32.

Ziegler

Diehl

Zigon

Fett

(2004). Source consistency, distinctiveness, and consensus: The three dimensions of the Kelley ANOVA Model in persuasion. Personality and Social Psychology Bulletin, 30, 352-364.