This item is only available as the following downloads:
1 DOES A CONFESSION AFFECT THE LIKELIHOOD OF ALIBI CORROBORATION FOR FRIENDS AND STRANGERS? By MEGAN R. KIENZLE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMEN TS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2011
2 2011 Megan R. Kienzle
3 To my parents, Carol and Doug, and my sister Emily
4 ACKNOWLEDGMENTS To begin, I would like to thank Dr. Lora Levett for all the support and guidance she has given me over the last 3 years here. I would not have gotten through this degree (or the next one) without her as my mentor, and more importantly as a friend. Further, I appreciate all the input and advice Dr. Lonn Lanza Kaduce and Dr. Lisa Hasel ha ve provided while working though each stage of this research project. Their comments and feedback made the project what it is, and I am grateful for their support in grad uate school as well. Further, I am very thankful of the American Psychology Law Socie in Aid which helped to fund this project. This program is responsible for helping get some really interesting projects off the ground, and I was lucky enough to be a part of this group. I would also like to thank each of the research assistants the hard work they put in as confederates, doing piles of data entry, or helping me run each sess ion. There are four in particular who deserve special mention : Josh Behl for running nearly all the second sessions, Brittany Ka ufman and Kelsey Henderson for giving up all those nights to help me run first sessions, and Nicole Barosso for helping work through the kinks when things did not go as planned. I a m so lucky to have such amazing RAs! Last, I want to thank my family for a lways supporting me and helping get me to where I am today. I especially want to thank my mom for listening to me complain each time I ran into a road bump and fo r ca lmly talking me through it. Special thanks to Kate Thompson, who will always be family t o me, for keeping me sane throughout this process (and life) by always being there for me when I need her most. And finally,
5 thanks to my dog, Nessie, for faithfully laying by my side wh ile I wrote this paper instead of playing with her
6 TABLE OF CONTENT S page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 Fal se Confession Research ................................ ................................ .................... 14 Structure of Confession Evidence ................................ ................................ .... 15 Confession Influence on Jurors ................................ ................................ ........ 18 Alibi Research ................................ ................................ ................................ ......... 22 Alibi Descriptions ................................ ................................ .............................. 22 Juror Perceptions of Alibis ................................ ................................ ................ 24 Alibi Generation ................................ ................................ ................................ 25 Interaction Effects ................................ ................................ ................................ ... 30 Overview of the Current Study ................................ ................................ ................ 35 2 ME THOD ................................ ................................ ................................ ................ 38 Participant Sample ................................ ................................ ................................ .. 38 Design ................................ ................................ ................................ ..................... 39 Independent Variables ................................ ................................ ............................ 39 Confession M anipulation ................................ ................................ .................. 39 Closeness M anipulation ................................ ................................ ................... 40 Dependent Measures ................................ ................................ .............................. 42 Alibi Questions ................................ ................................ ................................ 42 Alibi Corroboration Scales ................................ ................................ ................ 42 Wil lingness to be an alibi ................................ ................................ ........... 43 Alibi strength ................................ ................................ .............................. 43 Alibi depth ................................ ................................ ................................ .. 44 Alib i confidence ................................ ................................ .......................... 44 Memory for the event ................................ ................................ ................. 45 Viewing conditions ................................ ................................ ..................... 45 Co nfession Question Analyses ................................ ................................ ......... 46 Procedure ................................ ................................ ................................ ............... 46 The First Session ................................ ................................ ............................. 47 The Second Session ................................ ................................ ........................ 49
7 3 RESULTS ................................ ................................ ................................ ............... 52 Does Relationship to the Suspect and Confession Presence Affect Whether a Potential Alibi Wil l Proffer an A libi? ................................ ................................ ...... 53 Does the Relationship to the Suspect and Confession Presence Affect P arti orroboration? ................................ ............... 55 Does the Relationship to the Suspect and Confession Presence Affect P articipan e mory for the Event and Viewing C onditions? ................................ ................................ ................................ .......... 58 Does the Relationship to the Suspect and Confession Presence Affect onfession? ................................ ........ 60 4 DISCUSSION ................................ ................................ ................................ ......... 63 The Influenc C orroboration ................................ ................................ ................................ ....... 64 Confession Influence on Alibi C orroboration ................................ ........................... 66 Th e Interaction of Relationship to the Suspect and Presentation of Confession E vidence ................................ ................................ ................................ .............. 68 Implications ................................ ................................ ................................ ............. 69 Limitations ................................ ................................ ................................ ............... 72 APPENDIX A INFORMED CONSENT (P ART 1) ................................ ................................ .......... 76 B MILLER SOCIAL INTIMACY S CALE ................................ ................................ ...... 78 C GROUP CLOSENESS S CALE ................................ ................................ ............... 80 D SESSION 1 PROTOCOL ................................ ................................ ........................ 82 E LIST OF TEAM BUILDING ACTIVITIES ................................ ................................ 86 F CLOSENESS RATINGS F ORM ................................ ................................ ............. 89 G DEBRIE FING FOR P ARTICIPANTS (FAKE) ................................ .......................... 90 H FULL STUDY DEBRIEFING (REAL) ................................ ................................ ...... 91 I INFORMED CONSENT (P ART 2) ................................ ................................ .......... 93 J SESSION 2 PROTOCOL ................................ ................................ ........................ 95 K ALIBI QUESTIONNAIRE ................................ ................................ ........................ 99 L FINAL DEBRIEFING ................................ ................................ ............................. 107 LIST OF REFERENCES ................................ ................................ ............................. 110
8 B IOGRAPHICAL SKETCH ................................ ................................ .......................... 114
9 LIST OF TABLES Table page 3 1 Logistic r egression examining time 1 alibi willingness ................................ ........ 53 3 2 Logistic r egression examining time two alibi willingness ................................ .... 55 3 3 Main effect of relationship status on likelihood of corroboration ......................... 57 3 4 Main effect of confession status on likelihood of corroboration ........................... 58 3 5 Main effect of relationship status on memory and viewing conditions ................ 59 3 6 Main effects of relationship/confession status on belief in the confession .......... 61 3 7 Main effects of relationship/confession status on coer cion ................................ 62 3 8 Interaction between relationship and confession on coercion ............................ 62
10 LIST OF ABBREVIATION S CAQ Confession after questioning CBQ Confession before questio ning CI Confidence interval NC No confession NMFO Non motivated familiar other PA Participant alibis (the group of interest in this study) PF Participant friends (the pool of potential suspects)
11 Abstract o f Thesis Presented to the Graduate School of the University o f Florida in Partial Fulfillment of the Requirements for the Degree o f Master of Arts DOES A CONFESSION AFFECT THE LIKELIHOOD OF ALIBI CORROBORATION FOR FRIENDS AND STRANGERS? By Megan R. Kienzle August 2011 Chair: Lora M. Le vett Major: Criminology Law and Society Previous research has demonstrated several factors that contribute to wrongful convictions. However, researchers typically consider these factors independently of each other. In this study, we examined if one piec e of evidence (a confession) could alter another piece of evidence (the likelihood someone would provide an alibi). We manipulated both the relationship between the corroborator and the suspect and the presentation of suspect confession info rmation. Res ults indicated that knowledge of a confession from the suspect does result in lower likelihood of alibi corroboration and friends were more likely to be alibi corroborators than strangers Implications for police and courts are discussed.
12 CHAPTER 1 INT RODUCTION We know from past research that several factors can increase the likelihood that a person may be convicted of a crime they did not commit. Mistaken eyewitness identification, false confessions, snitch testimony, fa ulty science, incompetent couns el and many other things can all lea d to a wrongful conviction ( Scheck, Neufeld, & Dwyer, 2000; Wells et al 1998; Douglass & Steblay, 2006; Schoenfeld, 2005; Ceci & Bruck, 1993; Neuschatz, Lawson, Swanner, Meissner & Neuschatz, 2008; Tyler, 2006; Smith, Patry & Stinson, 2007 for a few examples). However, more often than not, cases of actual innocence have been attributed to several of these factors, causing us to question how so many things can go wrong in a single case (www.innocenceproject.org). It i s possible that these pieces of faulty evidence could interact with each other to result in these wrongful convictions (Hasel & Kassin, 2009; Kassin, 2009). The chance that one piece of bad evidence could alter the way investigators, witnesses, attorneys, and jurors examine, collect, or interpret other pieces of evidence may compound the effect of that initial, faulty piece of evidence and increase chances of conviction even though the suspect is actually innocent. Past research has shown that one piece of inculpatory evidence can contribute to the creation of more inculpatory evidence in that eyewitnesses who learned about a identification decisions to be consistent with the confe ssion evidence (Hasel & Kassin, 2009). It is possible that inculpatory evidence could also change exculpatory evidence. Specifically, whether or not a person is willing to offer an alibi may be affected by the potential alibi corroborator the other evidence in the case. Similar to
13 eyewitnesses, potential alibi corroborator s may change their alibi decisions to reflect other evidence in the case. To date, few studies have examined alibi generation, specifically none have looked at the possib ility that an alibi might change when presented with other types of evidence regarding the case. A good real world example of this possible evidence interaction is in the case of New York v. Kogut (2004). John Kogut was convicted of the rape and murder o f a 16 year old girl. Kogut spent nearly 20 years in jail before DNA testing was able to conclusively show that the semen found on initial investigation, Kogut went t hrough an 18 hour interrogation and was presented with false evidence about his case (police officers also used several other interrogation tactics). Even though he verbally maintained his innocence the entire time, the officers produced a confession (wri tten by the officer, not Kogut) that Kogut reluctantly signed after the fifth draft (not because he was actually admitting to the crime or believed he was guilty, but instead presumably to end the hours long interrogation). Allegedly, Kogut had been at his offense and theoretically should have had several people to serve as possible alibi corroborators (Kassin, 2009). However, when these potential alibi corroborators were questioned by police and told that Kog ut had confessed to the crime, they began to question if and when they saw him. In this case, investigators tol d the potential alibi corroborators about what they believed was a true confession. This may have caused the potential corroborators to question their memories for the night in question. Thus, potentially good alibi evidence was altered by another piece of faulty evidence. This shows first hand how one piece of faulty evidence (the false confession) can potentially
14 change to another piece of evi dence (the memories), and in at least one case, result in changing exculpatory evidence that may have cast reasonable doubt on In this study, we will examine the interaction between confession evidence and alibi corrob orators. In other words, we will see if knowing about confession evidence makes alibi corroborators more or less willing to provide their corroboration throughout the trial process. By manipulating whether a potential alibi corroborator is presented with confession evidence and the relationship between the alibi corroborator and the suspect, we will inspect what affects whether a potential alibi corroborator is likely to give an alibi, and subsequently, whether the decision to proffer an alibi is affected by other evidence in the case (namely a confession). Before delving into the current study further, we will examine how confession evidence is able to have this crucial impact on other types of evidence and the way that jurors perceive it. False Confessio n Research Some researchers posit that false confessions may have been an adaptive strategy for suspects in the past when there was no DNA testing or appeals court to validate their innocence (Bering & Shackelford, 2005). That is, if the suspect knows tha t the investigators already think he is guilty, he may believe he is better off in the long run to confess even though he did not actually commit the crime. In this case, he just wants to end the interrogation and can often make a plausible story by using the details of the case already supplied to him throughout the process of questioning. The suspect believes the likelihood of getting his named cleared by maintaining his innocence is futile and it seems probable that jurors who hear the evidence will th ink so as well.
15 Structure of Confession Evidence Confessions can be derived in 2 ways: (1) Internally, from actual guilt within the individual because they truly did commit the crime, or (2) externally, from other peripheral factors that occur during the q uestioning. For example, fear of violence, threats of a harsher sentence, guilt presumptive questioning tactics, or promises of leniency made by interrogators who come into the situation with pre conceived notions of guilt can lead the s uspect to admit to something he did not actually do (Kassin, 2005; Russano, Meissner, Narchet and Kassin, 2005; www.innocenceproject.org). These tactics are low in diagnosticity because even though they increase the likelihood of getting a true confession, they also increa se the likelihood of getting a false confession. Therefore, a possibly damaging outcome arises: a false positive, in which we get a confession from a suspect who is not guilty (Kassin, 1997). These tactics increase the likelihood of having the suspect co nfess, but this does not necessarily mean that those confessions are always truthful. The external influences (use of force, threats, deprivation of basic needs like food or water, promises of less severe punishments, etc.) may also make a confession inad missible in court, essentially stopping this evidence from ever being heard by jury members (Kassin, 1997). However in many cases, exclusion of a possibly coerced confession does not occur, which leaves jurors to decide how much weight to assign that piec e of evidence on their own. Previous research shows that jurors are not good at weighing this evi dence properly (Kassin, 1997). In the early studies of confessions research, three different categories of false confessions were termed (Kassin & Wrightsman, 1985). First, a voluntary confession is one that a suspect offers without any coercive tactics used by the interrogators. This
16 might happen when a suspect is trying to cover for a friend or family member or desires the fame and/or notoriety that go with confessing to the crime. Next, coerced compliant confessions are confessions in which a suspect confesses to a crime primarily to stop the interrogation process, l ike in the Kogut case. With this type of false confession the suspect knows that he/she d id not commit the crime, but because of the situation (perhaps being denied basic food and water for hours on end), the suspect says he/she committed the crime to escape the questioning. Last, coerced internalized confessions are confessions in which the suspect actually starts to believe that they have committed the offense. This is the most dangerous of the false confessions because it can alter knowledge of the true events to c onvince the investigators of the suspect innocence with or to explain in court. Another reason why some people may confess to crimes they did not commit is the belief that they may get an easier sentence if they cooperate with the investigators and do not When the suspect thinks he or she is going to get a prison sentence no matter what, it make s sense for him to try and get the least amount of time possible and this is more likely to happen by accepting th e blame and the subsequent punishment (even though in this case he should not). Further, i f the sus pect admits to the crime early on in the interrogation as opposed to hours later, then jurors typically see the suspect as already having suffered enough (G old & Weiner, 2000). They are then more likely to feel sorry for him and thereby reduce the punishment that he would be sentenced to because they think the suspect has learned his le sson and will not do it again.
17 Much of the previous research on false con fessions has focused on the interrogation tactics that can lead to false confessions, some even demonstrating that new, false memories can be formed during this process leading people to c reate vivid activities that never happened (Coff man and Henkel, 2004). In a Netherlands case study of two innocent men imprisoned for a murder, the suspects started with no knowledge about the crime and during the interrogation they became convinced they had committed it though the use of deceptive tr icks by investigators (Wagenaar, 2002). During interrogations, the suspects were fed information about the crime scene and murder weapon, and after hours of questioning they began to recall this information as memories from their own past actions. This c ase study demonstrates how, through the use of deception and other techniques used by investigators, nave suspects can actually begin to believe that they are responsible for a crime they did not commit. Another aspect of false confessions concerns how in terrogations are structured. One of the techniques most commonly used in questioning a suspect is the presentation of false incriminating evidence (e.g., presenting the suspect with evidence that an eyewitness placed him at the crime scene or telling the suspect that his/her DNA or fingerprints were found at the crime scene, when in reality none of this actually occurred). Investigation tactics like these are some of the more effective ways of producing internalized guilt for the offen se (Kassin & Kiechel 1996). The Reid encourages officers to use tactics like telling the suspect about evidence incriminating them in order to obtain a confession from uncooperative suspects (D rizin & Leo, 2004).
18 F urther, the Supreme court of Sou th Carolina upheld that confessions derived using the Reid technique are acceptable in court ( State v. Myers, 359 S.C. 40) This means these types of influences that can lead to false confessions are s till being used to gain a confession and are then left to the jurors to understand the coercive nature the information may have. because they are the ones most often doing th e interrogating of suspects (Elaad, 2003) Researchers found that officers are highl y overconfident in this ability. However, they are not highly accurate, even less so than college students with whom they we re compared (Elaad, 2003). Because they are s o overconfident, this can lead to officers of denial in the first place). Further, this guilt, making officers more likel y to resort to extreme measures during interrogation in order to elicit the confession that want (which may or may not be truthful, though they believe is due to the suspect true guilt for the crime ) Confession Influence on Jurors Once a person confesse s, whether the confession was truthful or not, that piece of evidence will play a role in the trial outcome. Jurors are unable to separate a confession (even one that is likely false) from their verdict decisions, even in the case in which they think they can separate it (Kassin & Sukel, 1997; Kassin, 1997). In addition, when a confession is introduced at trial, it becomes the single most important piece of evidence (Drizin & Leo, 2004). That means this one part of the trial can be more influent ial on ju ror decisions than every other piece of evidence sh innocence. F ormer United States Supreme Court justice William Brennan recognized
19 the influence a confession can have in a trial by stating so profoundl pg 340 ). This starts a chain reaction through the legal system ending in incarceration, demonstrating how a false confession can be the only incriminating piece of evidence needed to convict. One of the most salient examples of the problem is the case of the Central Park Jogger. In this case, 5 boys were convicted of rape and murder because of their own (false) confessions. The boys later recanted their confessions and the attorneys had no other physical evidence to prove their g uilt besides the confessions. The boys served 12 years behind bars before DNA testing eventually exonerated them after a man (in jail for a different crime) admitted to the offense and his DNA was matched to the DNA found at the crime scene. Even in a case in which there is little other evidence to prove guilt beyond a reasonable doubt, if a person confesses to a crime, jury members place a lot of em phasis on that single statement. O nce it enters in the court proceedings (regardless of the veracity of the confession) it can be nearly impossible to erase this effect. S significant role in the outcome of these trials, but this is also not the case. When told to disregard false conf ession evidence, jurors said they understand the statement was made involuntarily and that they did not let it affect their trial decisions (Kassin & Sukel, 1997; Kassin, Meissner, & Norwick, 2005). However, the rates of guilty verdicts were significantly higher in conditions with a confession (even if jurors believed that confession was coerced) than conditions in which confession evidence was not involved at all. Simply having a confession raises conviction rates, regardless of what jurors believe about whether the confession was coerced. Many people assume that
20 laypeople, as potential jurors, would be able to spot a false confession if presented with one, however, this is often not the case (Kassin and Wrightsman, 1980; 1981). No matter how much juror s may think they can mentally separate their trial decisions from that piece of evidence, research shows that a confession (even if the jurors think it is coerced) still has an effect on their decision at the end of the trial. Some studies have examined using expert testimony to successfully warn the jury about confession evidence ( Woody & Forrest, 2009 ) as well as the potential helpfulness of videotaping interrogations for later viewing at trial (Ayling, 1984). Taping the interrogation not only allows for the corroboration of the confession in court, but also safeguards the rights of the accused in that a videotape exposes to the jury how suspects ar e treated during questioning (Loewy, 2007). However, videotaping does not always function as a safeguard in cases of possibly coerced confessions. In some cases, recaps of what was said during the interrogation or abbreviated versions of the long interrogations (including just the end result of the questioning and not the steps leading to it) have been used in order to not overwhelm the jury with lots of extr aneous information to remember. However, tapes like this are usually biased against the offender, presented on behalf of the prosecution, and fail to accurately show the deceptive or coercive nature of the questioning (Ware, Lassiter, Patterson & Ransom, 2008) Past research has determined that if an interrogation is to be taped without bias, then the camera need s to be focused on both the interviewer as well as the suspect, rather than solely on one or the other to ensure the jury views the evidence as fairly as possible (Ware et al 2008). In addition, the videotape of the confession should be a video of the complete interrogation. Taping interrogations can help jurors better
21 understand what the susp ect had to withstand and thus may be able to help jurors determine how much faith to put in the confession. I n some cases of wrongful convictions, a false confession could have affected other pieces of evidence as well especially g iven the possibility of false confessions in dencies to believe a confession More than half the known exoneration cases contain multiple causes and we are still unaware of exactly what happens in a situation with more than one piece of evidence pre sent (Kassin, 2005; www.innocenceproject.org) It is possible that one piece of faulty evidence could have affected the collection of other evidence in those cases. For example, a damaging piece of evidence like a confession may affect acti ons throughout the investigation process as well. It may affect the investigative questioning of potential suspec ts, eyewitnesses, or even alibi corroborators to make people view the suspect as more guilty than if the suspect had not confessed. Indeed, r esearch shows that learning about a confession has the potential to lead eyewitnesses to change their identification decisions (Hasel & Kassin, 2009). It is possible that learning about confessions could change other forms of potentially exonerating eviden ce as well, such as whether a potential corroborator will remember seeing the suspect at the time of the crime. In the next section, we will first examine how alibi evidence is generated. Then we will examine how alibi corroborators as another form of w itnesses in the courtroom somewhat similar to eyewitnesses as both are forms of person evidence that are susceptible to influence may proffer different testimony based on whether a suspect has confessed.
22 Alibi Research Alibi Descriptions One of the most i nfluential pieces of research examining alibis was the first to describe the taxonomy of evidence used to explain corroborating alibis physical and person evidence (Olsen & Wells, 2004). A defendant can have physical evidence that demonstrates that he or she was at a specific place during a specific timeframe. Cash or credit card receipts, videotaped security camera footage, and much else could be used to prove the defendant was at a different place during the time the crime was committed. This type of evidence can vary on a continuum from strong ( for example, videotaped footage) to weak ( such as a cash receipt) as a function of how easy or to use a combination of these ( for example a credit card receipt as well as the tim e stamped camera footage) because a combination like this would be the most difficult to obtain if it did not actually occur Person evidence is t he other form of potential corroborating evidence f or an alibi, and is somewhat less influential than physical evidence (Olson & Wells, 2004). This type of evidence is when a person other than the defendant acts as a witness who vouches for the whereabouts of the defendant at the time of the offense, ther eby elim inating the possibility he could h ave committed the crime. P erson alibi evidence can take on two forms: motivated or non motivated others (Olsen & Wells, 2004). Motivated others are people who have an interest in keeping the defendant out of trou ble (typically friends or family), and non motivated others are typically strangers who we assume have no reason to lie for the defendant. Alibi corroboration also varies on a scale from strong to weak, with motivated others serving as weak corroborators and
23 non motivated others serving as strong alibi corroborator s. For example, a defendant would rath er have a stranger serving as a corroborator than their mother because people assume that the strang er has no reason to lie for him. H is mother however, h as an interest in protecting her son and keeping him out of prison, and thus is a less reliable form of person evidence. Given this information, an ideal person to testify would be a non motivated familiar other because they not only know the suspect well enough to recognize them and remember seeing them after some time delay, but they also lack the motivation to keep the defendant out of trouble like friends or family do. Howe ver, the concept of non motivated others is hard to pinpoint and little researc h has looked at it to date. confession is the relationship between that person and the suspect. When examining alibi evidence, this relationship between the corroborator and the suspect also plays a crucial role in determining which alibis are more likely to be believed by others (Culhane & Hosch, 2004). So, the interplay of closeness between these two people could, in this case, affect both the likelihood of a possible corrob orator believing the suspect confessed, as well as the likelih ood that a jury attributes the testimony to the accurate memory for the timeframe (something we will discuss further later) rather than their interest in doing whatever it takes t o keep the suspect out of the criminal justice system. M ost research has focused on evaluation of the alibi but there has also been a small body of research looking at how o r why people decide to become a corroborator As more cases with failed alibi cor roboration come to light with the help of the
24 Innocence Project, a greater emphasis has been put on why there are not more alibis as researchers look into the way alibis are generated, it becomes clear how difficult (or becoming an alibi) can be. Juror Perceptions of Alibis Person alibi evid ence can be very influential to jurors; however, this depends on the relationship perceived between the defendant and the person who gives the alibi (Culhane & Hosch, 2004). Jurors tend to believe an alibi given by a person who would not be motivated to c over for the defendant (a non motivated other) and this is shown through the higher rates of acquittal found for those defendants whose alibis have no relationship versus those with a relationship to the defendant, while the opposite is true for motivated others. The problem with person evidence is that the most probable alibi is also the least believable. The people who see the suspect on a regular basis and are the most likely to recognize the person and remember his or her whereabouts are the motivate d others who would also be the most likely to lie and cover for him had he committed a crime; therefore we cannot always trust that wha t they say is the truth. N on motivated others who do not have the impulse to cover for the suspect are just that: strang ers. People are less likely to remember seeing a person they have never met before and are especially unlikely to be able to recall the date, time and place of this encounter. To jurors, this means that if a non motivated other does recall seeing the sus pect, we can be more certain of their testimony because they have nothing to gain from lying for the suspect but jurors fail to account for the chance that the corroborator may not be accurate in their identification of the suspect
25 Another way jurors weig h alibi evidence is by gauging the amount of time the suspect waited to offer anyone the alibi (Connelly, 1983) Some argue that the length of time waited before telling the investigators about the alibi is correlated to the weight jurors associate with i t; the earlier a suspect reveals that he has an alibi, the more weight jurors will assign to that piece of evidence (Connelly, 1983). This generally serves one of two purposes: First, it can increase the perceived accuracy of the alibi; there is more time to investigate its truth and if there were anything questionable about it, jurors assume the prosecution would have presented this at trial. Second, it will save the suspect the time, money, and embarrassment of going to court should the alibi be able to be proven true. This hurts the accuracy of the alibi since people often believe if the case gets to court, then the alibi must not be true. Alibi Generation Like other witnesses, it is possible that alibis may change form when presented with other types of evidence. In the Kogut case, there were initially alibi corroborators, but the confession evidence essentially eliminated the alibi evidence, which demonstrates how multiple pieces of evidence might interact (Kassin, 2009). When the people who saw Kog ut at the party on the night of the crime were questioned about their memory but then were told that he had confessed, they began to question when and where they saw him, essentially probing their internal, autobiographical memory for that timeframe. Then, they did not feel strong enough to testify as an alibi corroborator We know fro m past research that there are two views on autobiographic memory: 1) memory is essentially a reconstruction of what actually occurred (in other words, a truthful representat or events ( Markus, 1977 ; Tulving, 1972). So, if we assume the first explanation to be
26 corroborator s should have remembered seeing him at the party and not have had any reason to drop out or question their memory for that night. However, the second explanation fits more with what happened in this case; corroborators originally remembered seeing him at the party, but after probing that memory further they began to question how truthful it was and eventually dropped out because they could not be sure. The corroborators most likely based their memory on their schema for a party like the one that night, and the n assumed Kogut would be there. In this case he actually was however their memory is based on typical occurrences and not on the ir memory for the actual events that took place that night Therefore we hypothesize that hearing about the confession is what actually caused the corroborators to begin to question their memory in the first place. These types of social influences have been demonstrated many times over in research and have the potential to cause people to alter their original decision because of some kind of contextual informatio n ( e.g., Festinger, 1954; Semmler, Brewer, & Wells, 2004 ). Festinger developed the idea of cognitive dissonance in the early 1950s, explaining that people feel uncomfortable when their behaviors do not match their beliefs. Most commonly, this leads indiv iduals to change their beliefs to align with the behaviors in order to reduce this feeling of discomfort. This change can be due to authority figures, trustworthiness, or other factors that the recipient values, and as such, leads the person to question t he validity of their initial opinion. In the Kogut case, had the alibi corroborators not been told of confession, most likely they might have felt more confident of the timeframe, and then be able to assist in Kogut exculpat ory alibi evidence. This evidence may have given jurors reasonable doubt of
27 guilt and prevented his conviction; however the social influence (hearing about his confession) caused them to alter their first decision of serving as an alibi corroborato r The behavior of confessing to the crime was inconsistent with the beliefs that Kogut was with them during the time of the crime, so they changed their beliefs or memories to align with that information. If one piece of evidence was alte happening in other cases where two or more pieces of evidence are used to convict Thus, evidence like a false confession could not only have an adverse effect on the jury; it co uld also cau se the investigators to stop searching for potentially exonerating evidence as well. One study had participants try to gather types of evidence (physical or person) to corroborate their own whereabouts for a specific timeframe either before or after rating time participants had to retrace in order to corroborate ( either 3 days or 30 days), sly had to generate their own alibi. It appears that having gone through the corroboration process makes people sensitive to the di fficulty of gathering proof. A s such, they are less strenuous in the expectations they have for what evidence others should produce. The influenc e of experiencing how difficult the alibi corroboration process can be for other people to provide an alibi corroboration Another study found that nearly 80% of participant s who were required to find evidence to prove their whereabouts for a certain time initially provided an alibi that they
28 later changed even when told the first time that they could say they did not remember (Charman, Cahill, Leins, & Carol, 2009). Student s filled out a questionnaire that asked where they were at during several points of time and should return a few days later with any evidence they could find to corroborate these whereabouts, though in the process of gathering this evidence many realized t heir original statements were incorrect. Additionally, they were highly confident in the initial alibi (which was wrong in some cases) because they reported what they typically do at that time every day. Participants reported using scripts of their norma l behaviors rather than basing the alibi on the actual memory of what they were doing at that specific time, but then had to change their story once the participants attempted to gather evidence to support the alibi. Because people appear to rely on somet hing other than actual memory for an event, then if the a time when something out of the ordinary occurs, a script will lead to an initially incorrect account of that timeframe which, as we learn later, leads police investigators to discount the alibi altogether and prevent it from getting to trial (Dysart, 2009). Additional resea rch may be beneficial here in trying to find a way to get alibis to switch the type of memory they are accessing during the corroboration process to rely on genuine memory rather than their normal day script. Without this research we are assessing what people think they would do if asked to be an alibi corroborator for someone and not what they would a say if they actually remembered the specific timeframe. Many have assumed that friends would be more likely to provide an alibi for a suspect than strang ers simply because they are inherently more invested in the
29 process, but a recent study found this was not the case: friends (people manipulated to alibi (Marion & Burk e, 2009). In this study, friends were manipulated by inducing perceived similarity and completing tasks together, but this may not have really tapped into the actual qualities that might make friends more likely to be an alibi corroborator (such as knowin g about the suspects character over an extended period of time). The current study sought to strengthen this manipulation by looking at pairs of previously existing friends rather than trying to induce a friendship during the study timeframe. In that wa y, we expected to be better able to differentiate between alibi suspect pairs of friends or pairs of strangers. Another study looked at the types of alibis police investigators often come in contact with and their opinions on them (Dysart, 2009). Only 23% of these investigators had training on how to interview alibi corroborator s, and 82% of police investigators interviewed said anytime the person changes their story, it is because they are lying in support of the suspect (either out of fear of the suspect or to protect them). Police indicated that alibi corroborations from non motivated others, or corroborations that contained physical evidence of whereabouts at the time of the crime were the most believable, but these types of situations w ere hard to come by because, most often the corroborators the police encountered were family or friends. Overall, poli ce generally do not trust alibi evidence and 24% of the time, police reported failing to investigate the alibi fully. This research sugg ests that even when an alibi is possible, it may not make it through the generation process that would allow the person to testify in
30 Interaction Effects With a few exceptions, psycho legal research to date has neglected the possibil ity that the independent pieces of evidence used to convict a person may interact to work in ways that exacerbate the effects each piece would have individually Given that 52% of the 271 wrongful conviction cases (to date) have more than one listed contr ibuting factor, we can no longer assume that to be true ( www.i nnocenceproject.org ; Kassin, 2009 ). A recent study by two psychology researchers is o ne of the only experiments that has taken this possibility of interdependence into account (which will be di scussed in detail below), and provided evidence that this interaction effect is indeed happening (Hasel & Kassin, 2009). I t is therefore feasible to think that other types of evidence may interact to result in a similar pattern. T here has been some eviden ce that social variables can affect the way that both fingerprint and DNA analysts make match decisions. Dror and colleagues had latent fingerprint examiners study their own previous matches and non matches a second time (Dror, Charlton, & Peron, 2006; Dr or & Charlton, 2006). Five years after the print was first classified examiners consented to being tested using their previous matches sometime within the next year. When these prints were presented in combination with contextual information (like the s uspect had confessed or was in police custody during the time of the crime) indicating that the print match decision should be in the opposite direction of their original decision, examiners changed their original decision more often than when given no con textual information. That is, receiving unrelated information indicating their initial decision was incorrect caused the examiners to change their match decisions ( even when the information provided was fictional).
31 Similar results were found when researche rs used biasing information that contained emotional information about the crime as opposed to contextual factors; examiners were more likely to change their previous match/no match decision in the direction of the information provided than if they were gi ven no emotional information (Dror, Peron, Hind, & Charlton, 2005). Examiners also seem to be aware of the influence, as even when they were provided with this emotional information and did not change their decisions, they still said the emotional informa tion affected them (Hall & Player, 2008). These studies demonstrated how easily simple contextual information can affect an outcome in the case of a fingerprint match decision. Our study hypothesizes that contextual information may affect other outcomes a s well specifically, a decision whether to be an alibi corroborator for a suspect. DNA analysts are vulnerable to many of the same influences that fingerprint analysts are when subjected to extraneous information about a case. Contrary to popular beli ef, DNA matching is not infallible and is subject to human error (Thompson, 1997). In fact, DNA examiners can be influenced by extraneous factors like social information, and this information can play a role in the decisions DNA examiners make about whethe r a DNA sample matches DNA taken from a crime scene. Some researchers posit that these types of analysts essentially work for law enforcement (Thompson, 1997). T here are analysts who flat out lie to help the police department, but even when misconduct is not that obvious and it is not known the analyst has a history of such behavior there is still the possibility that other factors could influence an match dec ision. Even seemingly small comments or a directional
32 hypothesis can bias examiners to decide in one way or the other. Essentially, if one conceptualizes the DNA test as an experiment, revealing the hypothesis of the experiment (by telling the DNA examine r the expected outcome of the test either to include or exclude a suspect) can result in investigator bias (Thompson, 1997). That is, the DNA examiner is then not blind to the hypotheses of the police investigators, and is susceptible to seeing what the y expect to see when interpreting results. This is especially common when the DNA is not a clear inclusion or exclusion (meaning it is ambiguous) and thus requires the analyst to use their own human judgment to make a subjective assessment of how close a match it is to the comparison (Risinger, Saks, Thompson, & Rosenthal, 2002). This acceptance of human decision making as the primary method of determination can leave this type of evidence open to the possibility of interference from other factors like th e social influences previously discussed. Even though social influence information can be a problem in DNA matching and fingerprinting, th ese type s of evidence allow for correction if social information has influenced their interpretations Specifically, w ith both DNA and fingerprint evidence, the investigators are able to go back later to double check whether analysis was completed correctly. In other words, the actual evidence (i.e., the fingerprints or DNA sample) is kept intact for future reference. T his is not the case with witness evidence like eyewitness identifications or alibi corroboration preserved. That is, the presenta tion of social information can permanently alter the memory of the witness. Once an alibi corroborator has been told about a confession,
33 and their beliefs about th eir own memories may be permanent; their memory might be forever modified. The recent study by Hasel and Kassin demonstrates how pieces of contextual has the potential to c hange other evidence (2009). Specificall y, the researchers were interested in h ow learning about a confession w in that decision when choosing from a lineup. In the study, participants witnessed a staged crime (a laptop theft), wer e immediately given a 6 person suspect absent lineup, asked give a confidence rating associated with their decision The same participants returned two days later and were randomly assigned to be told one of four conditions: 1) the person the witness selected from the initial lineup confessed, 2) all suspects denied involvement, 3) the selected person denied involvement, or 4) another person (not the one the eyewitness initi ally selected from the lineup) confessed to stealing the laptop. When relevant, participants were shown a signed confession. If they did not select someone from the initial lineup, they were randomly assigned to one of the three conditions: 1) an unidenti fied person confessed, 2) all suspects denied involvement, or 3) a specific person confessed. All participants were asked to reconsider their decision from the initial lineup (even if they did not choose) and again gave confidence ratings. Of the particip ants who chose a person from the first lineup, 2.44% changed when told their initially selected person confessed, 11.62% changed their decision when told all suspects denied having been involved 27.91% changed when told their selected person denied the cr ime, and 60.86% changed their lineup selection to a different
34 person when told that person confessed to the crime. A similar pattern was also found with those who did not select a person from the initial lineup. Not one person changed their selection whe n told everyone denied involvement, 44.67% changed when told an unidentified person confessed, and 50.00% changed when told a specific other confessed. who selected from the lineup mo re confident when they were given confirming feedback (the chosen person from the lineup confessed to the crime) than those who were told nothing at all. The witnesses were also likely to change their identification choices when faced with disconfirming f eedback through a confession from another person in the lineup (or from anyone when they had initially said the suspect was not present); it follows that those witnesses had low confidence ratings. Because reported confidence by an eyewitness has such a b ig influence on jurors when testifying in a courtroom recorded before any potentially biasing statements are said or heard If an eyewitness account changes in both lineup selection as well as the like a confession, it is possible that other piece s of evidence may change shape as well. More specifically, alibis with person evidence and eyewitness account s can be conceptualized as similar time of a crime (Olsen & Wells, 2004). With eyewitness identification, the eyewitness is remembering the situation and the face of the pe rpetrator to prove the suspect is the actual perpetrator whereas with an alibi a potential corroborator is remembering the
35 description and face of another person to prove they could not have committed the crime. Thus, it is possible that alibi corroborat ors may be as likely as eyewitnesses to change their accounts when presented with disconfirming evidence. Overview of the Current Study In light of all this past research, we created a study to examine if evidence interaction can happen when alibi corrobor ators hear about confession information. If an alibi corroborator changes their recollection when presented with disconfirming feedback (that the suspect confessed to the cri me), then this is an additional interaction effect we should be aware of during b oth the investigation and trial phases of a criminal case. If a person falsely confesses and this causes the alibi corroborator(s) in the case to withdraw evidence, the defendant not only has the possibly false confession acting as inculpatory evidence, b ut they also no longer have the corroborator(s) to serve an exculpatory function. To test this research question, we manipulated both t he relationship between the suspect and corroborator, and the timing of presentation of the confession evidence. W e empl oyed a two part experimental design During the first session, participants and a friend interacted with several other pairs completing team building activities. This served as the timeframe that a crime supposedly took place. Approximately two or three days later, t he participants returned to the lab to participate in the second phase Each person was randomly assigned to a condition in which they were told that either their friend, a non motivated familiar other, or a stranger (one of the people they met during the previous session ) was suspected of stealing an portable music player during the study a few days earlier. The research assistant, blind to the relationship between the suspect and the potential alibi corroborator acted as an
36 investigator b In conditions containing false confession evidence, the potential corroborators were told the suspect had confessed earlier in the day and also presented with a writ ten confess ion from the suspect. In one condition, the corroborator was asked if he would be willing to serve as an alibi corroborator for that pe rson prior to hearing about a confession (then subsequently asked again after hearing such). In the other confession co ndition, the corroborator was given all information (the suspect and the confession) at once and then asked about their willingness. In the control condition the potential alibi corroborators were asked about their willingness to serve as such without he aring any information about a co nfession Participants also filled out an alibi questionnaire assessing several qualities described below, and then were fully debriefed about the purpose of the study (and the need for using dece ption during the first sess ion). Therefore, this study was a 3 x 3 between subjects factorial design. First, w e predicted a main effect for both manipulations. For the relationship manipulations, we expected participants in t he friend condition would be more likely to p rovide an a libi for the suspect than those in a condition with a lesser relationship (the non motivated familiar other or stranger conditions, respectively ) For the confession manipulations, we predicted that this kind of evidence interaction would happen; potentia l alibi corroborators who were presented with confession evidence in either confession condition would be less likely to provide an alibi than those who were given no confession information Within this manipulation, we thought the participants who never heard the confession information would be most likely to provide an alibi, and those who heard all the information at once would be least likely to provide an alibi. In
37 the third confession manipulation, we want to see if we can get participants who initi ally provide an alibi when they have not heard the confession to switch their decision and no longer want to provide such once this information is presented to them. Further, we expected these main effects would be qualified by a higher order interaction b etween the two independent variables; t hose who were friends with the suspect (in other words, motivated corroborators) would be less likely to believe the confession evidence and therefore maintain their corroboration than strangers or non motivated famil iar others would be. This would demonstrate a potential problem for the court system because alibi corroborators who are most likely to be correct in their memory (and therefore should be corroborating the alibi) are the motivated others because they are more likely to remember the suspect and less likely to be influen ced by the confession evidence. H owever we know these are the people who are the least likely to be believed in a court setting (Culhane & Hosch, 2004). N on motivated familiar others or str angers may be likely to believe the confession evidence (and thus drop out as an alibi corroborators) because they may be less likely to remember the suspect, but they would be the preferred person to testify in court because they are more likely to be bel ieved by jurors than by motivated alib i corroborator s.
38 CHAPTER 2 METHOD Participant Sample 469 students (263 participant alibi s (PAs) and 206 participant friend s (PFs) ) were recruited through the use of a participant pool in the department of Sociology an d Criminology & Law, or in classes with the permission of the instructor. PAs participated in exchange for course credit, and alternative assignments were available if they did not want to participate in either phase of the study. PAs were all asked to b ring a friend (PF) with them to the first session and all participants were over 18 years old. Fifty seven PAs were eliminated from analyses because they did not bring a friend to the first session, and an additional 39 PFs wished not to be used for the s econd part of the study (thus the PAs could not be randomly assigned to relationship condition). All the participants removed until this point were done prior to the random assignment to conditions. Eleven PAs who were suspicious of the actual aim of the study were eliminated from analysis. No PAs revealed that they learned about the purpose of the study from the friend they brought to the first session. In total, four PAs failed the confession evidence manipulation check (described below) and were not in cluded in the analyses. One PA opted to withdraw his data at the end of the study so his infor mation random assignment took place (out of 166 fully completed participants) which is an attrition rate of less than 10% This left us with the final sample of 302 participants (151 PAs recruited through UF and 151 PFs brought to the first session by the PAs). The participants of inte rest in this study were the PAs.
39 Our final sa mple of PAs consisted of 66% females and 34% males which is consistent with the makeup of the Criminology undergraduate population at UF. The ages ranged from 18 32, with a mean age of 20.39 years. We had students from each class (9% freshmen, 23% sophom ores, 40% juniors, and 28% seniors). Most of our participants were White (52%) with Black second (21%), Hispanic a close third (15%), followed by Asian (4%), Cuban (4%), Latino (2%) and Other (1%), and 1% did not select a racial category. Design We used a 3 (suspect alibi relationship: friend v. non motivated familiar other v. stranger) x 3 (suspect confession: corroborator learned about a confession before initial questioning (CBQ) v. corroborator learned about a confession after initial questioning (CAQ) v. no confession (NC)) between subjects factorial design. Independent Variables Confession manipulation In the no confession condition (as further indicated by NC) the researcher showed a picture of the suspect and asked if they would be willing to be an alibi for this suspect, then wrote down their dichotomous Yes/No response. In the confession before questioning condition (indicated by CBQ) PAs were shown the written confession from the suspect along with their picture and subsequently asked about the ir willingness to be an alibi. In the confession after questioning ( indicated by CAQ) condition, the researcher presented the PA with the picture, asked if they would be willing to serve as her then explained that
40 the PA. The experimenter then again asked whether the PA was willing to be an alibi corroborator for that suspect. In the CBQ condition, the exp erimenter informed the PA that the suspect had been questioned earlier in the day and signed a written confession accepting responsibility for their actions and any punishment deemed fair by the principle investigator. They were shown the written confessi on and asked if they would like to be an alibi for that person. If PAs were in conditions without a confession, this information was not present. We included a manipulation check question in the alibi questionnaire to ensure that this manipulation worked by asking the participant to respond to the following statement on a six point Likert type scale (one indicating strongly disagree and 6 indicating strongly agree): I was told during questioning that the suspect confessed to taking the portable music play er (Appendix K Part C). We split this scale at the midpoint to dichotomize the variable, and any participant in confession conditions that responded with a one, two, or three (indicating disagreement) were removed from analyses, and any participants in t he NC condition who responded with a four, five, or six (indicating agreement) were also removed. In total, four participants failed this confession evidence manipulation check. Closeness manipulation During the second session, PAs were randomly assigned to learn that the person suspected of stealing the missing portable music player was either (1) the friend they brought to the first session, (2) a non motivated familiar other (like a classmate) or (3) a stranger (someone the participant had never previo usly met) PAs were asked during the first session to rate how close they felt to each person present on a one to seven Likert type scale (with one indicating stranger and seven indicating friend ; included in
41 Appendix F ). For participants assigned to the friend condition, the suspect was the person they brought with them to the first session. For participants in the non motivated familiar other condition, the suspect was chosen by taking the average of their ratings for each person present and selected a PF from that session that was closest to that average. Last, for participants assigned to the stranger condition, the suspect was the PF from the first session who the PA rated lowest on the scale. We included a closeness level manipulation check question in the alibi questionnaire (Appendix K Part C). Participants were asked to agree with the statement type scale ranging from strongly disagree to strongly agree. A one way ANOVA using relationsh ip closeness as measure showed our manipulation was mostly successful; there was an effect of F (2, 148) = 267.69, p <.01. Post hoc tests using least squares difference indicated that participants in the friend condition were more likely than participants in either the non motivated familiar other condition or the stranger condition to agree that the suspect was a friend (friend: M = 5.71 ( CI 5.47 5.95), non motivated familiar other: M = 1.78 ( CI 1.47 2.09), stranger: M = 1.79 ( CI 1.52 2.05). However, participants in the stranger condition did not significantly differ from participants in the non motivat ed familiar other condition which indicates that we may not have adequately captured the non motivated familiar other category since both that group and the stranger group tended to disagree w ith the above statement equally. This could also be a floor eff ect in that people view both friends and strangers equally on our measures.
42 Dependent Measures Alibi Questions yes/no response served as our primary dependent variable. For participants in the NC condition, this question was asked after learning about the portable music player theft and the identity of the suspect. For participants in the CBQ condition, this quest ion was asked after hearing about the portable music player confession. For participants in the after condition, this was asked after learning about the portable music player theft and the identity of t he suspect, but prior to hearing about the confession information. These responses serve as the alibi willingness for time one. Participant in the CAQ condition were aske d to respond to the alibi willingness question a second time after being t old about the confession. The two alibi question responses from participants in the CAQ condition were used to compute a variable measuring whether participants changed their alibi decision as a result of hearing about the e also asked to respond yes or no to the same question on the alibi questionnaire (administered after eac h manipulation was completed). Alibi Corroboration Scales We then created 6 main scales out of the questions from the alibi q uestionnaire. A full copy of a ll dependent measures from the alibi q uestionnaire as they appeared to partic ipants is included in Appendix K Unless otherwise indicated, participants were asked to provide agreement with all subsequent items, rated on Likert type scales ranging from 1 to 6, with one indicating strong disagreement and six indicating strong
43 agreement. All scaled items were summed and averaged to create one scale score. Reverse coded items are indicated by (R). Some scales include all reverse coded or no reverse coded questions because the order was mixed to encourage accurate reading of each question prior to giving a response. Some questions that did not factor onto the hypothesized scales were dropped from analyses. Willingness to be an alibi The first scale focused on general alibi corroboration willingness with higher scores indicating the participant was more willing to provide an alibi for the suspect. .77): a) I am sure about m during the night in question; b) I am a little unsure if this suspect is the actual perpetrator (R); c) There is no doubt in my mind about my decision to offer an alibi for this suspect; d) I h ave some hesitation in providing an alibi for this person because I think they might have actually committed the act (R); e) I question if serving as an alibi for this person is something I should do (R). Alibi strength We also created a scale to assess the strength with which the participants offered their alibi (or how sure the participant was that he or she would be an alibi for that person). All three items were reverse coded to ensure that higher numbers indicated more strength or confidence in offe the suspect is the person who took the portable music player during the first session (R); b) There is nothing that could convince me that the perpetrator was anyone other than this suspect (R); c) If I were on a jury hearing this case, I would convict the suspect (R).
44 Alibi depth The next scale we created assessed the depth of corroboration through the criminal justice system that the suspect would be willing to go through to maintain their alibi providi situations (each in its own question) increasing in stress to the corroborator related to completing this task, ultimately ending in testifying for that person in court as that would situation. Higher scores on this scale indicate willingness to proceed further into the criminal justice process in order to maintain their alibi for this suspect. The questions I would be willing to tell the suspect that I believe h e was NOT the one who took the portable music player nificant other that I would serve as an alibi corroborator for the suspect; c) I would be willing to provide an during a police interview; e) I would be willing to provide an alibi for the suspect by talking with both the trial attorneys; f) I would be willing to provide an alibi for the suspect that required me to take a polygraph test; and g) I would be willing to provide an alibi for the suspect by t estifying in a criminal trial. Alibi confidence Following each of the depth questions above, each question also had a paired item to assess how sure the PA was that he or she would actually be willing to provide an alibi in that particular situation. The responses were specific to how sure the participant was that he or she would be willing to go through that particular criminal justice stage in order to maintain their alibi corroboration, not just how sure they were
45 about being an alibi in general (this was assessed by scale 2). Again, higher scores indicate more confidence that the participant would provide an alibi throughout the involved in the alibi corroboration process at all (R); b) I would feel comfortable talking question whether vouching for this suspect to his employer is something I want to do (R); d) I am unsure if I would put myself through police questioning in order to be a problem for me to do; f) I hesitate somewhat to submit to a polygraph test in order to vouch for this suspect (R); and g) I would worry about going to court in order to testify for this suspect (R). Memory for the event by scaling responses to the following 6 items, higher scores indicated be lief their believe my memory for the event in question is good; b) I am not sure if I remember exactly what happened during the first session (R); c) I can account for the actions during the entire first session; d) During the first session, there are more than 15 everything that happened during the first session; and f) I am conf ident that my account of what happened is accurate. Viewing conditions Th the first session. Higher scores indicated the PAs believed they had better viewing
46 conditio My view of the suspect and his or her actions were clear during the entire first session; ession NOT have a clear view of the suspect for most of the first session (R); f) My view of the suspect was obstructed during some portion of the first session (R) ; and g) I had a good view of the suspect taking the portable music player Confession Question Analyses We also created a two su ), with higher numbers indicating more belief that the suspect actually confessed to stealing the portable music player This scale was comprised of the following questions: a) I believe the suspect confessed to the crime in que stion; and b) I believe the suspect did not confess to the crime in question. In addition, we asked the following question: I think the suspect was coerced into confessing to taking the portable music player Procedure Participants signed up in the Crimin ology, Law and Society Participant Pool for a 2 part study called Team Building in Friends and Strangers. The first session served as the timeframe that the so called crime took place (though no crime actually occurred, no portable music player was stolen ). T he second session was used to probe PAs for their to serve as an alibi corroborator for the suspect in question. PAs were also asked to bring a friend with the m to the first session and those who failed to do so were not included in the analyses to ensure each person had an equal chance at being randomly
47 assigned to any of the relationship conditions. The second session was completed individually (no PFs were p resent and the PAs were run one at a time). The First Session Upon signing up for the study online, PAs and PFs showed up to a classroom in groups of up to 20 people (10 PAs and 10 PFs). We also included at least two confederates from our lab that acted a s participants throughout each session for two reasons. First, confederates were in cluded to ensure any group did no t get too small; the smallest group that was tested included 8 people, although the majority of groups (20) included 18 20 members (enough t o make the room crowded). Of the total 33 groups, there was: 1 group of 8, 1 group of 9, 2 groups of 12, 2 groups of 14, 2 groups of 15, 3 groups of 16, 2 groups of 17, 9 groups of 18, 2 groups of 19, 5 groups of 20, and 4 groups of 22. We also instructe d the confederates to leave the room (for the restroom, a drink of water) at some point during the activities for a few minutes to mimic (attached in Appendix A) explaining that t hey would take part in some team building activities and complete closeness scales ( as a ruse for creating the event for the first session ) Once signed, everyone filled out two closeness scales to increase the believability of the fake study: one to asse ss the level of bond they feel with thei r friend (attached in Appendix B friends who they had not previo usly met (attached in Appendix C ). Then, they engaged in several team building activities while interacting with both the friends as well as the strangers by forming smaller groups made up of 5 8 people that changed with each new activity (script for session one attached in Appendix D, list of team building activities attached in Appendix E ). This portion of the study lasted about an hour and a half, and
48 the activities were timed to ensure each group completed exactly 45 minutes of interaction. At times PAs were in groups with their PF, but at other times they were no t. Near the end of the session, everyone again filled out the same closeness scales (Appendix B and C previous hours as a result of the team building activities. We also took a digital ph oto of each person without smiling (later used in the second session to identify the suspect), made a PowerPoint show with each picture on a slide, then had all participants rate how close they felt to every other person in the room on a one to seven Liker t scale (with one indicating the person was a stranger and seven indicating the person was a friend ; attached in Appendix F ). This scale was the basis for our relationship condition assignments (and to ensure a range of relationship levels were present fo r each person).Last, we separated the PAs and PFs into two rooms where the debriefing was completed. The PAs were given a fake de briefing (attached in Appendix G ) saying that we hypothesized team building activities to increase bonds with strangers, but n ot with friends. The debriefing also mentioned that we were interested in whether this was a lasting effect and thus the reason for the second session (to collect closeness scale information after a 2 3 day delay). They were thanked for their participati on, awarded class credit for the current session, and reminded of their time for the second session a few days later. PFs (in the other room) were also thanked for their participation, were told the full, final study debriefing (require d by IRB, attached i n Appendix H ), and explicitly told NOT to discuss any details with the PAs over the next week as it was crucial to our study that PAs not learn about the actual purpose of the study prior to coming to the second
49 session. We stressed the importance of the study to the John Kogut case described in the introduction of this thesis and explained how any leaked information could significantly compromise our results. We also asked the PFs to write out a short confession to show the PAs if they were in the condit ions in which the PA was told the portable music player during the session on (date) and am willing to accept any punishment the d by the PF at the bottom of the page. PFs had the option to indicate that they did not want their name, picture, or confession used in the second session. Thirty nine PFs opted to do this; the data from the PAs who brought these PFs to the study was not u sed because of the lack of ability to randomly assign the PA to the relationship condition. The Second Session Two or three days later, PAs who choose to continue with the study reported to the lab individually for their second session timeslots. They wer e randoml y assigned to one of the nine cells in the study design. They were again given an informed consent document (maintaining the team building cover story Attached in Appendix I ) and filled out both the friend and team closeness scales (Appendices B and C ). After finishing both scales, PAs were told that an portable music player was taken during the team building session (which did not actually happen). They were told that the research assistant wanted to question the PA about their memory for that timeframe (script for the second session attached in Appendix J) All materials were prepared ahead of time by the principle investigator, and the research assistant was blind to which relationship con dition the PA was assigned to. T he researchers only k new which confession condition to use. The research assistant presented the picture of the suspect taken
50 during the first session to help the PA recall the suspect and to make the questioning similar to a real criminal investigation where a mugshot might be presented The research assistant then told the PA that this person had been accused while the research assistant was questioning other PAs earlier that day In conditions with confessions, the investigator also presented the written confession as well either before asking the PA if he would serve as an alibi in the CBQ condition (giving all the information at once), or following the alibi willingness question in the CAQ condition (telling them who the suspect is and asking about alibi willingness, the n showing the confession and asking about willingness again). PAs then filled out the alibi quest ionnaire (attached in Appendix K ) designed to further examine those decisions and assess the depth of their commitment to providing the alibi corroboration for the suspect. The second session typically took about 20 minutes. We also checked for leaked information from friends on this questionnaire by explaining that it was completely ok if they had learned information from their friend, but we just needed to be able to account for it. PAs were assured they would not lose any credit by revealing that they were told anything by their friend. Afterwards PAs were thanked, awarded class credit, and debriefed again this time with the full study debriefing that thei r PFs received after t he first session (Appendix L ). To ensure that no harm was done to participants we also asked them to indicate whether they felt distressed or upset at any point during the session and if so, to describe what made them feel this way a nd if there is anything we could do to keep that from happening (required for approval by IRB). All but two people indicated they had a positive, enjoyable experience and felt little to no stress throughout either part of the study.
51 Those two participant s indicated moderate stress from having to be questioned about a theft.
52 CHAPTER 3 RESULTS Before running any analyses, all data was entered twice into two separate datasets. These were merged to compare each data line for incons istencies and mistakes in entry and then compared back to the original data when any discrepancies were found. We al so ran frequency analyses on each variable to find any responses that were out of range and to ensure each of the variables were entered in the correctly coded dire ction. After running the hypothesized analyses reported below, each of our analyses was run again including the covariates race, gender, which researcher ran the second session, and if the student had previously taken Psychology and Law to ensure random a ssignment was completed correctly and that there were no differences between analyses completed with these controls and the ones reported below We did find an interaction between experimenter and confession status in the coercion analysis, however we had a large number of experimenters t hat only ran a few sessions. Because of that, it is possible these results are most likely not reliably demonstrating any bias (and because this was the only significant result in all the tests of covariation ; it is possi ble it is spurious). Principle components factor analyses with a varimax rotation were run on each of the hypothesized scale items to ensure each scale was assessing the construct it was designed to assess. We kept any items that loaded above .7 on the f the dependent measures section of the method). Results were reported using a 95% confidenc e interval ; this is standard practice for most psychology studies. Any findings that were significant at the 0.10 confidence level were reported as marginally significant.
53 Does R elationship to the S uspect and C onfession P resence A ffect W hether a P otential A libi W ill P roffer an A libi? backwards, stepwise logistic regression with relationship status, confession status, and the ir interaction as independent variables. The stranger and NC conditions were coded as comparison group s The final model included two 2 (4) = 55.86, p <.01, R 2 = .31. There was a main effect of rel ationship status on participant s decision s to be an alibi ( T able 3 1) ; friends were more likely than strangers to offer an alibi (89% were willing to be an alibi versus 32%, respectively). T here were no differences between the non motivated familiar others (30%) and strangers on this measure There was an overall main effect of confession status on the whether be an alibi as well. Participants who heard confession evidence before being asked to serve as alibis were significantly less likely to do so than those who did not (39% were willing to be an alibi versus 60%, respectively). Recall that this question was asked prior to when participants in the CAQ condition were told about the confession. Thus, participants in the CAQ condition should look exactly the same as the participants in the NC condition. Accordingly, t here were no significa nt differences in decisions whether to be an alibi between participants in the CAQ condition (52% were willing to be an alibi corroborator ) and participants who did not hear about a confession. Table 3 1 Logistic r egression examining time 1 alibi willing ness Variable St. Error p>lzl Relation 32.68 <.01 NMFO .13 .44 .09 .76 .87 Friend 3.05* .58 27.28 <.01 21.10 Confession 6.28 .04 CBQ 1.27* .51 6.27 .01 .28 CAQ .47 .47 .98 .32 .63 indicates significance
54 After hearing about the confession, participants in the CAQ condition had the option to switch their alibi corroboration decision. To assess whether people in the CAQ condition would be more likely to change their alibi decision in the stranger relationship condition versus the friend relationship condition, we ran a Chi Square Test to libi. We isolated those in the CAQ condition that changed their decision for this test (they were the only participants asked whether they would like to switch their alibi decision. We predicted that strangers would be more likely than friends to change from being willing to provide an alibi to no longer wanting to serve as such when confronted with the confession. However these results were not significant, 2 ( 2 N =161 ) = 1.13, p = .57. While all people who changed the decision went from providing an alibi to not providing one, p articipants did not differ in their likelihood to change their alibi decision whether they were in a friend (27%), non motivated f amiliar other (13%), or stranger (25%) condition; participants were equally likely to change from offering an alibi to withdrawing that alibi regardless of their relationship to the suspect. A second question asking participants to respond dichotomously (y es/no) whether they would provide an alibi for the suspect was asked on the alibi questionnaire (after the participants in the CAQ condition heard about the confession following their initial decision whether to be an alibi). We again ran a backwards, ste pwise logistic regression regressing the alibi question onto confession status, relationship status, and the interaction between these variables The stranger and NC conditions were again coded as the comparison groups. The overall model was also signifi cant, 2 (4) = 63.65, p <.01, R 2 = .34 (Table 3 2). There was an overall effect of relationship status on
55 more likely than strangers (28%) to provide an alibi, however non m otivated familiar others (21%) were not different t han strangers in this respect. willingness to be an alibi at time two as well. Participants who heard about confession evidence p rior to being asked to serve as alibis (39%) were significantly less likely than participants who heard no confession (60%) to serve as alibis There was also significant difference between the CAQ group (34%) and the NC group (again, 60%) that we expecte d given that this question was asked after the CAQ group heard about Participants in the CAQ group looked more like the CBQ group during the time two questioning because both were significantly less likely to offer an alibi for t Table 3 2. Logistic r egression examining time two alibi willingness Variable St. Error p>lzl Relation 38.42 <.01 NMFO .40 .48 .70 .40 .67 Friend 2.95 .55 28.63 <.01 19.07 Confession 10.80 <.01 CBQ 1.29 .51 6.38 .01 .28 CAQ 1.61 54 9.05 <.01 .92 indicates significances Does the R elationship to the S uspect and C onfession P resence A ffect P arti O verall L ikelihood of C orroboration? We predicted that the level of bond the PA felt with the suspect would affect the ing for that person, so we ran a MANOVA to test the effect of relationship status and confession status and the interaction between these two alibi, strength of alibi, alibi depth, and alibi confidence. These scales were all included
56 in the same analyses because they assess very similar concepts (likelihood of corroborating). There was an overall significant effect of relationship status on the scales, = .53, F (8, 27 2) = 12.92, p <.01, 2 = .28. Tests at the scalar level showed a main effect of relationship status on each of the 4 scales (Table 3 3 for means and significant differences for each group). Overall, there was an effect of relationship status on participan F (2, 139) = 41.09, p <.01, 2 = .37. Post hoc tests using LSD showed that participants in the friend condition were more willing to be an alibi than participants in the non motivated familiar other condition and the stranger condition. Participants in the stranger and non motivated familiar other conditions did not differ in their willingness to be an alibi. to the alibi strength scale, F (2, 1 39) = 4.03, p <.02, 2 = .06. Participants in the friend condition offered stronger alibis than participants in the non motivated familiar other or stranger conditions; however, the difference between friend and stranger was only marginally significant. Again, participants i n the stranger and non motivated familiar other conditions did not differ in the strength with which they offered their alibis. depth scale as well, F (2, 139) = 53.94, p <.0 1, 2 = .44. Participants in the friend condition scored higher on the alibi depth scale than both non motivated familiar others and strangers. Participants in the stranger and non motivated familiar other conditions did not differ in their alibi depth scor es. the alibi confidence scale, F (2, 139) = 34.57, p <.01, 2 = .33. Specifically, friends
57 scored higher on alibi confidence than non motivated familiar others or strangers Again, participants in the stranger and non motivated familiar other conditions did not differ in their alibi confidence. Table 3 3. Main effect of relationship status on likelihood of corroboration Strangers (CI) NMFO (CI) Friend (CI) Wil lingness to be an alibi 3.59 a (3.32 3.87) 3.47 b (3.20 3.75) 5.09 ab (4.91 5.37) Strength of alibi 5.19 a (4.98 5.40) 5.06 b (4.85 5.27) 5.49 ab (5.27 5.70) Alibi depth 2.67 a (2.30 3.05) 2.74 b (2.36 3.11) 5.13 ab (4.75 5.51) Alibi confidence 3.3 5 a (3.04 3.65) 3.40 b (3.10 3.71) 4.97 ab (4.66 5.28) Shared subscripts within a row indicate significant differences, however a in strength of alibi is marginally significant We also predicted that knowledge of a confession would have an effec t on the likelihood of corroboration. The scale questions were collected after all condition manipulations were completed, so we expected that the CAQ group would look similar to the CBQ group since both had received the confession information at that poi nt. Further, both of these groups were expected to differ from the NC condition in which PAs never heard any confession information. We found a main effect of confession status on the scales, = .75, F (8, 272) = 5.30, p <.01, 2 = .14, indicating that t he timing alibi in support of the suspect (Table 3 4 for means). There was an overall effect of F (2, 139) = 7.07, p <.01, 2 = .14. Post hoc tests using LSD showed that participants in the CAQ group were less willing to offer an alibi for the suspect than those in the CBQ or NC groups. Participants in the NC and CBQ group s did not differ in their willingness to be an alibi. There was an overall main effect of confession status on alibi strength, F (2, 139) = 17.57, p <.01, 2 = .20. Post hoc tests using LSD showed that participants in the CAQ
58 condition rated the strength of their alibi lower than participants in the CBQ or N C conditions. Again, participants in the CBQ and NC groups did not differ in the strength of alibi they reported. The main effect of confession status on the alibi depth scale was only marginally significant, F (2, 139) = 2.78, p =.07, 2 = .04. Post hoc tests using LSD revealed PAs in either confession group were less likely than those given no confession to proceed through the entire criminal justice system in order to maintain their alibi. Participants in the CBQ and CAQ group did not differ in these r esponses. alibi confidence scale, F (2, 139) = 5.16, p <.01, 2 = .07. Post hoc tests using LSD showed participants in the CAQ condition reported less confidence in proceeding through the entire criminal justice system than those in the NC or CBQ conditions. Participants in the NC and CBQ conditions did not differ in this regard. The effect of the interaction between these variables on their likelihood of corroboration was not significant, =91, F (16, 416) = .85, p = .63, 2 = .02 Table 3 4. Main effect of confession status on likelihood of corroboration NC (CI) CBQ (CI) CAQ (CI) Willingness to be an alibi 4.41 a (4.15 4.68) 4.07 b (3.78 4.35) 3.68 ab (3.40 3.96) Strength of alibi 5.58 a (5.37 5.78) 5.44 b (5.22 5.66) 4.73 ab (4.51 4.94) Alibi depth 3.87 ab (3.5 1 4.23) 3.34 b (2.96 3.73) 3.33 a (2.94 3.71) Alibi confidence 4.21 a (3.91 4.50) 3.99 b (3.68 4.30) 3.52 ab (3.20 3.83) Shared subscripts within a row indicate significant differences, however b in alibi depth is only marginally significant Does the R elationship to the S uspect and C onfession P resence A ffect P ar B eliefs A bout T heir M e mory for the E vent and V iewing C onditions? We predicted that participants would report differences in their memory or viewing conditions from the timeframe o f the first session based on the relationship they had
59 with the suspect. Specifically, we thought those in the friend condition would report they had better memory and a better view of the suspect actions than those in non motivated familiar other and s tranger conditions. We ran a second MANOVA with the independent variables and the interaction between them as predictors and the viewing conditions and memory scales as the dependent measures to test this hypothesis. There was an overall significant effec t of relationship status on the DVs, = .68, F (4, 284) = 14.93, p <.01, 2 = .18 (Table 3 5 for group means). At the scalar level, there F (2, 142) = 26.53, p <.01, 2 = .27. Post hoc tests using LSD sh owed that PAs in the friend condition reported they believed they had better memory for the first session than PAs in the non motivated familiar other or stranger conditions. P A s in the stranger and non motivated familiar other conditions did not differ i n their belief in their own memory. the viewing conditions scale, F (2, 142) = 30.12, p <.01, 2 = .30. Post hoc tests using LSD revealed friends reported they believed they had better viewing conditions of the suspect than did non motivated familiar others or strangers. Again, participants in the stranger and non motivated familiar other conditi ons did not differ in their report s of their viewing conditions. Table 3 5. Main effect of relationship status on memory and viewing conditions Strangers (CI) NMFO (CI) Friend (CI) Memory 3.26 a (2.97 3.55) 3.36 b (3.06 3.66) 4.67 ab (4.36 4.97) Viewing conditions 2.55 a (2.29 2.81) 2.51 b (2.25 2.77) 3.81 ab (3.54 4.01) Shared subscripts within a row in dicate significant differences We also predicted that participants who heard about the confession evidence in either t he before or after condition would report having a worse memory for the event
60 and having worse viewing conditions than those who never hear d about the confession evidence. However, the effect of confession evidence on the DVs was not significant, = .96, F (4, 282) = 1.64, p = .17, 2 confession during the second session reported no better memory for the first session timeframe, nor better viewing conditions of the suspect than those who never heard abo ut confession evidence. The interaction between relationship and confession status on their responses to the memory and viewing conditions scales was not significant, =98, F (8, 282) = .28, p = .97, 2 < .01 Does the R elationship to the S uspect and C onf ession P resence A ffect P articipant B eliefs A bout the S C onfession? We ran two additional ANOVAs with our independent variables and the interaction between the m as predictors on a two item scale of belief in the confession and a question about pos sible coercion to confess. We hypothesized that participants in the friend condition would be less likely to believe the confession than those in the non motivated familiar other or stranger conditions. Overall, there was a main effect for relationship st atus on the belief in confession scale, F (2, 141) = 2.54, p <.01, 2 = .04. Post hoc tests using LSD showed that non motivated familiar others were less likely than strangers to believe the suspect confessed to the act (Table 3 6 for means). However, ther e were no differences between strangers and friends, or friends and non motivated familiar others. Further, we predicted that people who heard about the confession in either condition would be more likely to believe it than those who never heard the confes sion evidence. We found a main effect for confession status on the belief in confession scale, F (2, 141) = 20.07, p <.01, 2 = .22. According to post hoc tests using LSD, both
61 the CBQ and the CAQ groups were more likely to believe the suspect confessed than the NC group was (also in Table 3 6). The two confession conditions did not differ. The effect of the interaction of to the belief scale was not significant F (4, 141) = 1.05, p = .38, 2 = .03. Table 3 6. Main effects of relationship/confession status on belief in the confession Stranger (CI) NMFO (C I) Friend (CI) Belief in confession 3.14 a (2.78 3.50) 2.59 a (2.23 2.94) 2.73 (2.36 3.09) No Confess. (CI) Confess. Before (CI) Confess. After (CI) Belief in confession 1.89 ab (1.55 2.24) 3.33 a (2.97 3.6 9) 3.23 b (2.86 3.60) Shared subscripts within a row in dicate significant differences We ran another ANOVA with the independent variables and the interaction suspect was coerced into confessing to taking the portable music player responses to the coercion scale, F (2, 141) = 14.03, p <.01, 2 = .17. Again, post hoc tests using L SD showed that friends were more likely than both non motivated familiar others and strangers to believe the suspect was coerced into confessing to the crime. Strangers and non motivated familiar others did not differ on their belief in coercion o f the su spect. coercion scale, F (2, 141) = 36.53, p <.01, 2 = .34. Post hoc tests using LSD revealed that both CBQ and CAQ conditions were more likely than participants in the NC condition to believe the confession was coerced, (Table 3 7 for means). Participants in the two confession conditions did not differ in whether they believed the suspect was coerced into confessing.
62 Table 3 7 Main effects of relationship/confession status on coercion Stranger (CI) NMFO (CI) Friend (CI) Coercion 3.11 a (2.77 3.45) 3.33 b (2.97 3.66) 4.34 ab (3.99 4.6 9) No Confess. (CI) Confess. Before (CI) Confess. After (CI) Coercion 2.42 ab (2.09 2.75) 4.00 a (3.65 4.34) 4.36 b (4.01 4.72) Shared subscripts within a row indicate significant differences We found a signif icant effect of the interaction between the independent variables F (4, 141) = 7.21, p <.01, 2 = 17 Test of the simple main effect of relationship status within confession status usin g LSD revealed that within the confession conditions, participants in the friend condition were more likely than participants in the non motivated familiar other and stranger conditions to believe the suspect was coerced to confessing to taking the portabl e music player (Table 3 8 for means) However, this relationship was not observed in the conditions with no confession. Table 3 8. Interaction between relationship and confession status o n coercion Strangers (CI) NMFO (CI) Friend (CI) NC 2. 37 (1.81 2.92) 2.78 (2 .21 3.35) 2.12 (1.53 2.70) CBQ 3.24 a (2.65 3.82) 3.38 b (2.77 3.98) 5.38 ab (4.77 5.98) CAQ 3.73 a (3.11 4.36) 3.82 b (3.24 4.41) 5.53 ab (4.91 6.16) Shared subscripts within a row i ndicate significant differences
63 CHAPTER 4 DISCUSSION In this study, students interacted with several people for 90 minutes and were questioned about the story of a fake theft that occurred during the event. We manipulated whether the suspect was a friend, a non motivated familiar other, or a stranger, and whether the potential alibi was told the suspect had confessed to the act or not Given that no theft occurred, the potential alibi corroborators should be able to proffer an alibi corroboration because the potential alibis were with t he suspect during that entire timeframe. However, similar to the Kogut case, potential alibi corroborator s (friends, strangers, and non motivated familiar others) were less likely to want to be an questioning. Thus, the current contamination m ight have occurred and allowed us to experimentally examine whether learning about a confession can cause alibi corroborator s to alter their memories for the event. We expected that having a prior relationship with the suspect would affect alibi corroboration, and those who were friends with the suspect were more likely to be an alibi for the suspect than non motivated familiar others or strangers were. We further expected that hearing information about the suspect having confessed would affect alibi corroboration, and those who heard confession information were more likely to be an alibi than those who do not. Last, we expected partici pants in the stranger condition would be most likely to change their alibi corroboration after hearing the confession information and those in the friend condition with no confession information to be least
64 likely, however we did not find this result. It i s possible that power for these tests was not sufficient to demonstrate such results c orrect ly. The I nfluence of the A R elationship to the S u spect on L ikelihood of C orroboration Previous research that has looked at alibi generation in terms of who is likel y to testify for a suspect has not found that friends were more willing to be an alibi corroborator than strangers who had no connection to the suspect (Marion & Burke, 2009). However, in the current study; we found that friends were more likely to offer an alibi for the suspect than were either non motivated familiar others or strangers. It is possible that a stronger manipulation of relationship may be what made these results significant in comparison to past research In the prior study, closene ss was induced during the experiment and seemed to be unable to truly shift a previously unknown pair of participant and confederate into strong enough friends that one would be willing to go through the alibi corroboration process. Our study used pre exi sting relationships, and as such we were able to have years of commitment through their friendship to each other as a backbone for our relationship manipulation. The process of serving as an alibi for someone is not as simple as borrowing notes for a clas s which a student might do with any acquaintance; being a corroborator likely uses some knowledge about the which cannot be determined during an hour long study. Future researchers should consider this in designing alibi studies if th ey want the relationship between the alibi and suspect to be strong. We failed to differentiate between our stranger and non motivated familiar other group in this study, so researchers need to keep the relationship strong (perhaps by using already existi ng friends rather than
65 inducing friendship during the study) so that pairs are not made up of strangers who are Unfortunately, research has also shown that friends are viewed as less trustworthy alibis than strangers when corroborators testify in court (Culhane & Hosch, 2004). This presents a problem for a defendant in that the most likely alibi (a friend) would be viewed poorly by ju rors compared to a stranger. In contrast, the best alibi for court, a stra nger who remembers seeing the defendant during t he timeframe in question, would no t be likely to be an alibi corroborator in the first place. T his study demonstrated that friends are more likely than strangers to offer an alibi and to believe that they ha ve a better memory for the timeframe in question (regardless of whether that memory is truthfulness than were strangers. Friends believed they had a better memory for the e vent and a better view of the suspect during the event compared to strangers, when in fact would have varied across conditions equally). It is possible that because they bel ieved they had better memories, they were more likely to proffer an accurate alibi, but this does not mean jurors will trust their explanation of what happened This may account for the second explanation of autobiographical memory discussed in the introdu ction: that for that specific timeframe. Friends spend more time together so their schema for a normal day includes one another, whereas a stranger may be more likely to refuse to be an alibi corroborator because their schema for that time does not include the suspect (but this is dependent on whether the timeframe in question is an atypical one). The
66 problem is that most people (including those serving as jurors) beli eve we can accurately recall any timeframe, which shows the inconsistency between what we expect to happen and what actually does occur Confession I nfluence on A libi C orroboration In this study, when participants heard that the suspect confessed, they were less likely to serve as an alibi corroborator than when they heard no informat ion about a confession. However, they did not rate their memory or viewing conditions any lower than people who did not hear about a confession for the time of the crime. We kn ow from past research that confession evidence can change the decisions of both eyewitnesses and fingerprint analysts from their original choices to one aligning with the confession evidence (Hasel & Kassin, 2009; Dror & Charlton, 2006). The current study shows that alibi corroborators are also affected by learning confession information. With each new piece of research that adds to this list of possible contaminations it becomes more and more evident that person evidence needs to be treated more carefull y than we currently are doing Memory is fallible and needs to be protected against contamination by other pieces of evidence. This is especially true when confession evidence is the contaminating piece because so much research has demonstrated the influ ence even a single sentence confession can have on the outcome of a trial (Drizin & Leo, 2004; Kassin & Sukel, 1997; Kassin, 1997). This study has given even more piec e of evidence (Leo, 2009). There was a difference between the two confession group in most of the scales assessing overall likelihood of corroboration which indicated the CAQ group was l ess likely to corroborate an alibi. This may be due to the timing of presentation.
67 Participants reported the same memory and viewing conditions for the event regardless of hearing or not hearing confession evidence (and regardless of when they heard the confession evidence), but they reported lower confidence, less stre ngth, and less willingness to provide an alibi when told about the confession after they were initially asked about their willingness to be an alibi compared to when they were told about the confession before being asked for their initial willingness to be an alibi. It may be that giving the confession information after asking them to provide their initial alibi corroboration decision led potential alibi corroborators to be less sure of the corroboration compared to others, whereas potential alibis who got all the pieces of information about the investigation at once were more confident in their alibi decisions. This is something futu re research may want to address; often investigators may draw out the process of revealing information to suspects (and may use this same approach when dealing with alibi corroborator s because most do not have formal training on ho w to interview them ( Dysart, 2009). It may foster more of a collaborative process between corroborator and investigator when all the information is given to the corroborator up front rather than assessing their willingness to be an alibi after each piece is revealed. However, some research has found that using strategic disclosure of evidence (revealing one item at a time) when interviewing suspects results in better ability to detect deception (Hartwig, Granhag, Stromwall & Kronkvist, 2006), but potential alibi corroborator s are inh erently different from suspects, which this may expl ain our opposing results Revealing evidence piece by piece with su spects allows for the opportunity to those inconsistencies. However, this is not the goal of an alibi interview which focuses
68 more on gathering all the information possible. These two objectives may need different strategies to be effective. It is also possible that alibi s who heard about the confession after initially making an alibi corroboration choice viewed the confession as competing information with their original belief of what happened, and so they lowered their confidence, strength, and willingness to be an alibi in order to relieve the conflicting views, consistent with the cognitive dissonance theory described earlier (Festinger, 1954). The I nteraction of R elationship t o the S uspect and P resentation of C onfession E vidence Friends were more likely than both strangers and non motivated familiar others to believe the suspect was coerced into confessing to the crime, but only in the conditions where confession information wa s presented. Research has shown that confession evidence can still have an effect on trial outcomes, even when jurors say they know the confession was coerced and will not consider it in their decision (Kassin & Sukel, 1997; Kassin, Meissner, & Norwick, 2 005). It appears as though friends may experience some cognitive dissonance when told their friend has confessed to stealing that non motivated familiar others and strangers do not experience. It is possible that their initial belief of the friend being a good person who would not steal is confronted with this confession which indicates the opposite, so they alter their belief about the confession However, since a j ury is made up of strangers they do not have the prior knowledge about the suspect, and as such, do not experience the cognitive dissonance that would explain why a confession might be present when the suspect actually did not commit the crime
69 Implication s One of the biggest issues in collecting person evidence like alibi or eyewitness evidence is that investigators only have one chance to get an accurate account of what happened. As mentioned earlier, with physical evidence, it is possible to return to t he original source and retest the results to see whether social influence had an effect in the match or exclusion made by an investigator. However, we do not have that luxury with person evidence. Because of this, it is crucial to learn everything we can about that piece of evidence before any other pieces of evidence have a chance to change what the person may say. We also must document the original account of what took place to have a record that can be referenced when s tatements change in light of other evidence. A very common recommendation coming out of confession research is to videotape the interrogation, and we assert the same should be true for alibi corroborator questioning as well. If the interactions with thes e corroborators are able to be shown at trial, jurors would be better able to understand exactly what took place during questioning and an expert would be able to explain that any information presented to the corroborator by investigators may have an influ ence on defense team in that they are now able to present evidence of an alibi (and probably an expert to explain why the evidence is in video form ) when that evidence would be unavailable to them had the questioning not be videotaped. This recommendation also serves to keep the police investigators honest in that a videotape would be evidence of investigators trying to eliminate any alibi corroborator s from testifying in order to strengthen the case. If jurors are able to watch exactly what is being said to a potential
70 corroborator they stand a better chance of accurately deciding how much influence was w such testimony. This study is additional support for the second conceptualization of autobiographical memory described earlier; we base our recollections on what we assume happened, not what actually occurred (Tul ving, 1972; Markus, 1977). The idea that memory is fallible has been an issue for memory researchers for years and is applicable when switching to research on alibi corroborators who are susceptible to social information changing their account of a timeframe (Robinson & Swanson, 1990). T he prob lems associate d with memory have been demonstrated over an extended time, in multiple forms of people evidence, and throughout several different fields of study, investigators and jurors need to be more aware of how fallible memory can be and put less emph evidence or additional witnesses/alibis (all individually collected, of course). Further, it would be beneficial for police investigators to treat each piece of evidence independently to pre vent this kind of evidentiary interaction from occurring. Specific to this research, that means not using the knowledge of a confession from the suspect to discourage a potential alibi corroborator from serving as one. Many of the tactics commonly used b y police investigators may be more harmful than helpful, timeframe. Research has revealed information about the kinds of interrogation tactics that are likely to cause f alse confessions, but it is still possible that some people ( particularly people in more vulnerable populations like the young and mentally disabled though anyone is susceptible ) will confess to crimes they did not actually
71 commit (Kassin, 1997; Kassin & Wrightsman, 1985). When this happens, it is crucial to keep this information separate from people who could serve as alibi corroborators for the suspect to prevent the possibility that their memory for the time in question could be tainted. It is very dam aging in a court setting to argue th at a confession is likely false and would be even worse for the defendant to lose those people who could account for his whereabouts and conclusively say that he could not possibly be the one who committed the offense (L eo, 2009; Kassin & Sukel, 1997; Kassin, 1997). The effect this has on jurors may be compounded; having a false confession and no alibi to demonstrate it is false may be exponentially more harmful to the suspect than jurors hearing either of those pieces a lone. If a justice says confession evidence can be extremely prejudicial, what can be said about the compounded effect of confession evidence on the collection of other evidence? Again, more research needs to be done to see how much more damaging any com bination s of evidence pieces may be on jurors be fore researchers draw any conclusive opinions but it seems possible from past research that police investigators may be unaware of how any two pieces of evidence fit together to alter decisions (Hasel & Kass in, 2009; Kassin, 2009; Dror, Charlton, & Peron, 2006; Dror & Charlton, 2006). Even more so, researchers are unaware of how others (specifically jurors who hear the case) may or may not be able to evaluate those pieces of evidence when presented in combin ation. use of DNA (though long after his conviction) however most cases do not have DNA evidence available to them. If police who use techniques that present information (li ke a
72 confession) to potential corroborators change their practices to protect against possible evidence contamination then we can prevent futur e cases from this problem. W hat about the past cases that lead to a suspect behind bars because the alibi corro borator s were not willing to go to court for him or her? In this situation, judges need to be very careful in how they respond to those cases that originally had corroborators who subsequently withdrew their stories especially when in combination with co nfession evidence. If confessions really are as prejudicial as Justice Brennan says they are then judges may need to relook at each case with a confession to determine if the possibility for contamination occurred. Limitations One potential problem with the study is that we currently have low power which may be increasing the likelihood of a type II error. Some effects were marginally significant so it is possible that with more participants we would have stronger evidence of the possibility that alibi corroborator s are influenced when told about confession evidence. In addition, the interaction effects were almost all not significant in this study. These are the tests in the study that require t he most statistical power. I ncreasing the power in the stu dy would de crease the likelihood that the n on significant effects of the interaction on our dependent measures are type II errors. Currently, we have between 17 21 people per cell, but we plan to recruit at least 25 participants in each condition to incre ase power. Additionally, using a college sample for this study may limit the generalizability of our findings. I t is commonly known that most interaction s people have with the criminal justice system takes place in young adulthood (around the age of most college students), however college students and community members are two distinct groups of
73 people. C ollege students most likely differ from the kind of person who is often involved in crime so generating an alibi for a student participating in a school sponsored study may be inherently different than generating one for a person outside of the educational setting. It is also possible that the setting for college students is different from what community members experience (Culhane, 2009). College studen ts spend the majority of their time on campus and with their friends (typically other young people), whereas community members spend most of their time either at work or with their families. T he types of alibis experienced by these two groups are likely v ery different; however, the memory of people in either group should be similar. In this way, the results from a college sample should hold true when transitioned to a community setting, though this is something future research should examine as well. Anot her problem with the current study is a lack of significant differences and a success ful manipulation check between two of our closeness manipulation groups. As mentioned earlier, we may not have adequately captured the non motivated familiar other group because in this study they tend to behave like strangers. It is possible this is a flaw in the study design; there may be something about our manipulation that did not differentiate between students that know the participants well enough to recognize the suspect and remember seeing him but were not friends to the extent that the corroborotor has an interest in keeping the suspect out of the criminal justice program altogether. It is also important to note that strangers may not always be non motivated, s o this may have been a trait we were unable to tease apart from these two groups (Dysart, 2009).
74 However, it is also possible that this group of people (familiar others who are truly non motivated) does not exist in the real world either particularly beca use this finding is similar to that of previous work (Olson & Wells, 2004) In that study, Olson and Wells found that participants who rated the credibility of alibi corroborators perceived no differences between non motivated strangers and non motivated familiar others (with the caveat that this may not hold up in a trial setting). If there are no differences in the two groups? Finding a person that is familiar enou gh with a suspect, but not so familiar that their motivation for involvement in an investigation changes, may be an extremely difficult endeavor. Psychological research also demonstrates that a sizable group of people do not form relationships that way o ne is nearly always motivated for some reason towards a particular act or behavior or person (Gilovich, 1991; Murray, 1999). If a stranger wants to believe that a suspect is good and would not steal anything, then the behaviors that are consistent with th is belief are what comes to mind when asked if the corroborator would provide an alibi. The same is true if a stranger wants to believe a suspect is bad; the behaviors that reinforce this notion are the ones that are recollected It is possible that trul y non motivated people do not exist. The recommendations put forth in this discussion are based on the fairly small body of research looking at alibi generation, along with some evidence interaction studies to demonstrate that confession evidence can effec t whether a person is willing to supply an alibi corroboration for a suspect. Because both areas are still very new in the field of psychology and law, more research needs to examine the reasons people decide to become an alibi and what types of evidence may change this decision. If
75 more research is done and findings from this study are replicated, recommendations like those above could be made to police departments about how to avoid potential for contamination of evidence (and safeguards to make sure al ibi evidence is collected appropriately) Presumably, the less interdependent individual pieces of evidence are, the more accurate jurors could be in making a decision at trial (although this remains a question for future research). If these recommendatio ns are framed in a way emphasizing that these policies are likely to increase the chances of getting justice for the victim(s), then we would expect police to be open to implementing such strategies in how they question potential alibi corroborator s in the future and ensure the work investigators do is seen as strong in a court setting. Even small adjustments to the way investigators question individuals involved in a case may have a drastic impact on reducing the number of people who are falsely accused a nd possibly convicted for acts they did not commit.
76 APPENDIX A INFORMED CONSENT (PA RT 1) Informed consent Protocol Title: Team Building with Friends and Strangers Please read this consent document carefully before you decide to participate in this study Purpose of the research study: The purpose of this study is to examine how team building activities enhance bonds differentially with friends or strangers. What you will be asked to do in the study: In this study, you will be asked to complete a cl oseness scale for both the friend you brought as well as the other people in this same session. Next, you will participate in team building activities in large groups, while mixing up members of those groups occasionally. Then you will complete the same c loseness scales once again (for your friend and for members of the larger group). At the end of the study, we will collect contact information in case you need to be contacted further and take a digital photo to better assess closeness levels between grou p members. Time required: 1 hours Risks and benefits: There is no more than minimal risk to you as a participant. As a benefit, you will learn about the research process. Compensation: You will be awarded credit through the Department of Criminolog y, Law and Society Participant Pool for participating in the study. Confidentiality: Your identity will be kept confidential to the extent provided by law. Any information you provide will not be linked to your personal information and all pictures wil l be destroyed after completion of the project. Voluntary participation: Your participation in this study is completely voluntary. There is no penalty for not participating. In addition, you do not have to answer any question you do not wish to answer. Right to withdraw from the study: You have the right to withdraw from the study at anytime without consequence. Whom to contact if you have questions about the study:
77 Box 117330, Gainesville, FL 32611, firstname.lastname@example.org Lora Levett, Ph.D., Assistant Professor, Department of Sociology and Criminology & Law, PO Box 117330, Gainesville, FL 32611, email@example.com Whom to contact about your rights as a research participant in the study: UFIRB Office, Box 112250, University of Florida, Gainesville, FL 32611 2250; ph 352 392 0433. Agreement: I have read the procedure described above. I volunt arily agree to participate in the procedure and I have received a copy of this description. Participant: _____________________________________ Date: _______________ Principle Investigator: ___________________________________ Date: _________
78 APPENDIX B MILLER SOCIAL INTIMA CY SCALE Miller Social Intimacy Scale Please read each question carefully, then circle the number that best represents how you feel about that statement. Very Some of Almost rarely the time always 1. When y ou have leisure time how often do you choose to spend it with him/her alone? 1 2 3 4 5 6 7 8 9 10 2. How often do you keep very personal information to yourself and do not share it with him/her? 1 2 3 4 5 6 7 8 9 10 3. How often do you show him/her affection? 1 2 3 4 5 6 7 8 9 10 4. How often do you confide very personal information to him/her? 1 2 3 4 5 6 7 8 9 10 5. How often are you able to unde rstand his/her feelings? 1 2 3 4 5 6 7 8 9 10 6. How often do you feel close to him/her? 1 2 3 4 5 6 7 8 9 10 Not A little A great much deal 7. How much do you lik e to spend time alone with him/her? 1 2 3 4 5 6 7 8 9 10 8. How much do you feel like being encouraging and supportive to him/her when he/she is unhappy? 1 2 3 4 5 6 7 8 9 10 9. How close do you feel to him/her most of the time? 1 2 3 4 5 6 7 8 9 10 10. How important is it to you to listen to his/her very personal disclosures? 1 2 3 4 5 6 7 8 9 10
79 (Continued) Not A li ttle A great much deal 11. How satisfying is your relationship with him/her? 1 2 3 4 5 6 7 8 9 10 12. How affectionate do you feel towards him/her? 1 2 3 4 5 6 7 8 9 10 13. How important is it to you that he/she understands your feelings? 1 2 3 4 5 6 7 8 9 10 14. How much damage is caused by a typical disagreement in your relationship with him/her? 1 2 3 4 5 6 7 8 9 10 15. How important is it to you that he/she be encouraging and supportive to you when you are unhappy? 1 2 3 4 5 6 7 8 9 10 16. How important is it to you that he/she show you affection? 1 2 3 4 5 6 7 8 9 10 17. How imp ortant is your relationship with him/her in your life? 1 2 3 4 5 6 7 8 9 10
80 APPENDIX C GROUP CLOSENESS SCALE Group Closeness Scale Please read each question carefully, then circle the number that best represents how you fe el about that statement. Very Some of Almost rarely the time always 18. How often would you choose to spend time with this group of people? 1 2 3 4 5 6 7 8 9 10 19. How often do you think the team intera cted in a positive manner? 1 2 3 4 5 6 7 8 9 10 20. How often would you show this group of people affection? 1 2 3 4 5 6 7 8 9 10 21. How often would you confide personal information to this group of people? 1 2 3 4 5 6 7 8 9 10 22. How often do you think you could understand 1 2 3 4 5 6 7 8 9 10 23. How often do you feel close ties with members of this group? 1 2 3 4 5 6 7 8 9 10 Not A little A great very deal 24. How likely are you to be encouraging and supportive to members of the group if they are unhappy? 1 2 3 4 5 6 7 8 9 10 25. How close do you feel to this group most of the time? 1 2 3 4 5 6 7 8 9 10 26. How important is it to you to listen to group 1 2 3 4 5 6 7 8 9 10 27. How s atisfying is your relationship with
81 this group? 1 2 3 4 5 6 7 8 9 10 (Continued) Not A little A great very deal 28. How much closer do you feel to the group after completing these activities? 1 2 3 4 5 6 7 8 9 10 29. How important is it to you that other group members understand your feelings? 1 2 3 4 5 6 7 8 9 10 30. How likely are you to spend time with this group after today? 1 2 3 4 5 6 7 8 9 10 31. How important is it for you to be a part of this group? 1 2 3 4 5 6 7 8 9 10 32. How important is your relationship with members of this group in your life? 1 2 3 4 5 6 7 8 9 10 No Some Yes what 33. Do you feel you belong with the other members of the group? 1 2 3 4 5 6 7 8 9 10 34. Do you feel the group worked well together as a team? 1 2 3 4 5 6 7 8 9 10 35. Did you enjoy the time you spent with the members of this group? 1 2 3 4 5 6 7 8 9 10 19. Did you feel the group made a cohesive team? 1 2 3 4 5 6 7 8 9 10
82 APPENDIX D SESSION 1 PROTOCOL Team Building in Friends and Strangers You need to arrive at least 10 minutes before the study is set to begin. Before participants arrive, make sure to get all papers/supplies ready in the order participants will need them. Then set up the powerpoint template for the 3 rd closeness scale. During the study it is crucial for the confederate RAs to pretend like they do not know each other or the RAs that are helping to run the experiment. You are filling in so there are extra you should excuse yourself to go to the bathroom or get a drink at a couple times throughout the study, then come back shortly. Once the participants and friends arrive, welcom e them to the study and have them sit veryone to arrive. We are taking photographs of everyone to use a little later in the study, so if s tarted the tasks, upload all the pictures to the computer. Arrange all 20 pictures in the template so that each is visible (shrink the pictures if you need to, but keep them all the same size) and letter A H (depending on how many people are there). now that everyone is here we can get started. The first thing I need you all to do is read through this informed consent document, and if you agree to participate in the study, sign and date at the bottom. It just explains a little more be doing tonight, so let me know if you have any questions. Also, please let me know if you would like a copy of the informed consent for Hand out the ICs, collect them all again once people have signed. Assign participant numbers while co llecting these sheets (1 P or 1 F for participant or friend, then 2 P etc) and have them remember to write this on the scale sheets. level of the bond that you share with ea ch person in this room. The one that says Miller social intimacy scale is to rate how close you feel to the friend that you brought with you here today. The one that says Group closeness scale is to rate how close you feel to this group of people here as a whole. Answer each of the questions on both of these scales as honestly as you can and remember that seeing them and please remember to write the participant number I just g ave you in the top right corner of each sheet. Pass out both closeness scales, then collect them again as people finish.
83 ***DURING THIS TIME, CONFEDERATE RAs SHOULD EXCUSE THEMSELVES TO USE THE RESTROOM OR GET A DRINK OVER THE COURSE OF THE SESSION*** The RAs that are helping to run the study should assist in separating into groups, explain the activity and then monitor the group they are in charge of. DO 45 minutes of team building games **S ee other sheet for list of team building activities and description of how to play and how to designate groups** complete the scales you filled out at the beginning of the study a second time. y If you rate them a 7, it would mean that they are a good friend. Be sure to write the letter that corresponds to the friend you came with today in the box in the lower ri ght corner. Pass out the number sheets, then collect their number sheets once people complete the task, then black out the screen so the pictures are no longer displayed. again. Remember the Miller scale refers to the friend you brought to the study with you, and the Group scale refers to this group as a whole. Please answer each question honestly as to how close you feel to these people right now, and do not simply answer consis of the study. Since we really want to get honest r esponses to how you guys separate those of you who signed up for the study from the friends you b rought with you, so lets have the participants follow the other researcher down the hall Have Participants go down the hall to another room, then complete debriefing for each as follows: Participants Hand out the sho read it aloud to them. create bonds between individuals. More specifically, we think that these types of games encourage bonds to form more quickly within stranger groups because there is no prior groundwork for a relationship to be formed on so even small, seemingly meaningless interactions have a large impact on the level of closeness two people feel. On the other hand, the friend you broug ht with you today had already built a relationship foundation and therefore small games like the ones
84 you played would not have had as great an impact on the level of your friendship as it did with the strangers. If you wish to have the data you provided n ot included in the study, you are free to withdraw your data from the sample. If you choose to do so, it will be destroyed immediately. If you have any questions feel free to ask them. Your results will be kept confidential to the experimenters. Thank you again for Remind each participant of their 2 nd session time by filling out the reminder sheet for each person individually. week, so just hang out a bit whil e I go around the room. The second session will mostly consist of filling out the closeness scales again to see if there is any time delay effects with how you feel towards these other people so it will most likely take a half hour at the longest. And re member that it will be in Walker Hall room Fill out the sheets SLOWLY! and say everything that you write down out loud. Double check the participants name with their participant number written on all their paperwork a nd their time for the second session. Ask if they all know how to get to Walker and with the friends group, so anything you can do to take longer is a good thing. Also, since this second session is done individually we really want to make sure discussion with your fri end might contaminate your true judgments on the closeness you feel with these people. Thank them again for their participation, then walk back towards the original room. Friends Hand out the full study debriefing and read it aloud to them. this study was an Innocence Project client who had been at a party with several people during the time of the crime he was accused of, which should have provided him with multiple alibi corroborators to stand up for him in court to maintain his innocence. However, because of extensive pressure during interrogation, he ended up confessing to the crime and once word of this got out, all of his friends from the party started to question when and where they saw him and no longer wanted to serve as an alibi fo r him in court. We think the confession is what caused these friends to go back and rethink about the night in question and to withdraw their original story. In this study, we are looking to examine whether or not a piece of evidence used during investiga tive questioning (the presence or absence of a confession from the suspect) is able to influence whether a person is willing to offer their alibi in the potential alibi corrob orator being a close friend, acquaintance, or a stranger could have an influence on that decision as well. So, in the second session, we are varying whether we tell your friends that either you or someone else they met during this session (a stranger or a cquaintance) is a suspect for stealing a portable music player We will also vary whether we tell them that the suspect
85 confessed to the act or not. Thus, we are able to see if the least believable alibi er) was less likely to be altered when told the suspect had confessed than when told he did not. If this is true, then strangers (the stronger alibi) are more likely than friends to be swayed into questioning their memory of what happened during the incid ent and then also more likely to choose to withhold their alibi testimony. extremely important that you do not reveal any of this information to your friend during the time he or she is waiting f or the second session. If they were to find out the true purpose of the study, they would likely not be able to accurately answer any of the alibi at the bottom of the study as a sort of spect in that session for one of the people in the other room. If you are, we are also asking that you write out a short confession on a piece of paper and sign it for us to use in the second session to increase believability that someone actually did con fess. If you allowing us to use your information and remember that everything will be kep t Hand out slips of paper for them to write on if they want and read what they should put portable music player during the study on (whatever date) and am willing to accept any punishment the investigator think have them sign below that sentence. really need you to keep these details about the study quiet until after your friend has finished their participation. The most important part of this research is how your friends react to being asked to serve as alibis for tonight remember in the Kogut case explained in the debriefing, the alibis could have saved him from wrongful conviction. If your friends are leaked any information about the true responses, which makes what you just participated in a waste of time. So, again, y is. You can talk just talk about it as if the study only consisted of team building activities). Thank you so much for your cooperation with this. Again, if you do not wish to be a suspect, please just indicate so on your form. Collect any written confessions and all the signed debriefi ngs, give copy of debriefing if they want it, then let them go.
86 APPENDIX E LIST OF TEAM BUILDIN G ACTIVITIES List of Possible Team Building Activities Human Scavenger Hunt Do this as 1 entire group. Hand out the sheets and pens to each person, then have them go around the room talking to everyone, trying to get signatures for each of the boxes. Line up Split into 2 groups by drawing numbers out of a hat. Then have them lineup up in specific orders like: Alphabetically by last name without talking Fr om shortest to tallest with eyes closed From oldest to youngest without talking 2 truths and a lie: Have everyone split into 2 groups (one of participants and one of friends). Then give each person an index card and have them write 2 truths and a lie on. Collect all the cards and put them in the middle, then have someone come select a card and guess who the person is, as well as which statement is false. Go around until all the cards have been chosen. Human Knot Have everyone count around 1 3, then split into those groups. Then stand in a circle facing each other and have everyone close their eyes and reach across the circle with their right hand and grab hands with someone. Then open their eyes and grab the left hand of someone different, going a Then the goal is to unravel the knot. Have them work on this by individual groups once, then make it a race between the 3 groups by having them wait until the PI says go to start unraveling the knot. T ele phone aka filter one large group I Will Survive In this team building game for the workplace, people are divided into groups and are presented with a scenario. The scenario can go something like this, 'You are stranded on an island, and you are given th e choice of ten objects for survival. Choose wisely.' in 3 groups by drawing numbers Make a castle one member of the team gets 30 seconds to make the castle out of cards then switch out first team to get 4 levels wins in 2 groups by lining up and tallest/ shortest
87 zoom each person gets a part of a comic strip and they have to put it in order without showing the pictures to anyone else in 3 groups by birth month Amoeba race every person in the group must be touching and 2 people cannot have either foot on the ground, then go a certain distance in 2 groups by drawing numbers step through paper game in 4 groups (split 2 groups in half) Icebreakers (in 3 groups by drawing numbers) 1. If you could have an endless supply of any food, what would you get? 2. If you were an animal, what would you be and why? 4. When you were little, who was your favorite super hero and why? 5. Who is your hero? (a parent, a celebrity, an influential 7. If they made a movie of your life, what would it be about and which actor would you want to play you? 8. If you were an ice cream flavor, which one would you be and why? 10. If you could visit any place in the world, where would you choose to go and why 12. Are you a morning or night person? 13. What are your favorite hob bies? 14. What are your pet peeves or interesting things about you that you dislike? 16. Name one of your favorite things about someone in your family. 17. Tell us about a unique or quirky habit of y ours. 19. If someone made a movie of your life would it be a drama, a comedy, a romantic comedy, action film, or science fiction? 20. If I could be anybody besides myself, I would be If you were a comic strip character, who would you be and why? What thought or message would you want to put in a fortune cookie? If you had to give up a favorite food, which would be the most difficult to give up? What is one foo If you won a lottery ticket and had a million dollars, what would you do with it? to? If you could be any superhero and have s uper powers, which one would you like to have and why?
88 Mount Rushmore honors four U.S. presidents: Washington, Jefferson, Lincoln, and Roosevelt. If y o u could add any person to Mount Rushmore, who would you add and why? What award would you lo ve to win and for what achievement? If you could transport yourself anywhere instantly, where would you go and why? In your opinion, which animal is the best (or most beautiful) and why? What is one item that you really should throw away, but probably never will? Growing up, what were your favorite toys to play with as a child? Four corners draw pictures in each of the corners of a piece of paper favorite hobby favorite vacation spot something you're scared of one of the most impor tant things in your life Then everyone shares with the group (in 4 groups by numbers 1 5, 6 10, etc) React and Act one person sees a paper with a description on it and has to act it out for others to guess Being surprised by a large, aggressive bear in the woods You just won the lottery You have just been proposed for marriage with an engagement ring You just got fired by an incompetent boss Making the game winning pass to win the Superbowl You just fell in love in 3 gro ups (by numbering off)
89 APPENDIX F CLOSENESS RATINGS FO RM Ratings form for everyone from first session P # Circle the number you believe best represents the relationship you feel with each person. Picture # Stranger Neutral Friend A 1 2 3 4 5 6 7 B 1 2 3 4 5 6 7 C 1 2 3 4 5 6 7 D 1 2 3 4 5 6 7 E 1 2 3 4 5 6 7 F 1 2 3 4 5 6 7 G 1 2 3 4 5 6 7 H 1 2 3 4 5 6 7 I 1 2 3 4 5 6 7 J 1 2 3 4 5 6 7 K 1 2 3 4 5 6 7 L 1 2 3 4 5 6 7 M 1 2 3 4 5 6 7 N 1 2 3 4 5 6 7 O 1 2 3 4 5 6 7 P 1 2 3 4 5 6 7 Q 1 2 3 4 5 6 7 R 1 2 3 4 5 6 7 S 1 2 3 4 5 6 7 T 1 2 3 4 5 6 7 Please write the letter of the person you came to the study with here
90 APPENDIX G DEBRIEFING FOR PARTI CIPANTS (FAKE) Team Building Debriefing In this study, we are looking to ex amine the way that team building activities create bonds between individuals. More specifically, we think that these types of games encourage bonds to form more quickly within stranger groups because there is no prior groundwork for a relationship to be f ormed on. So even small, seemingly meaningless interactions may have a large impact on the level of closeness two people feel. On the other hand, the friend you brought with you today had already built a relationship foundation and therefore small games like the ones you played would not have had as great an impact on the level of your friendship as it did with the strangers. If you wish to have the data you provided not included in the study, you are free to withdraw your data from the sample. If you ch oose to do so, it will be destroyed immediately. If you have any questions feel free to ask them and please know your results will be kept confidential. Thank you again for participating.
91 APPENDIX H FULL STUDY DEBRIEFIN G (REAL) Confession Influence on Alibi Corroboration Full Study Debriefing (for friends after 1 st session) The basis for this study was an Innocence Project client who had been at a party with several people during the time of the crime he was accused of, which should have provided him with multiple alibi corroborators to stand up for him in court to maintain his innocence. However, because of extensive pressure during interrogation, he ended up confessing to the crime and once word of the confession got out, all of his friends from the party started to question when and where they saw him and no longer wanted to serve as alibis for him in court. We think the confession is what caused these friends to go back and rethink about the night in question and to withdraw their original stories In this study, we are looking to examine whether or not a piece of evidence used during investigative questioning (the presence or absence of a confession from the suspect) is able to influence whether a person is willing to offer their alibi in support of corroborator being a close friend, an acquaintance, or a stranger could have an influence on that decision as well. So, in the second session, we are varying whet her we tell your friends that either you or someone else they met during this session (a stranger or an acquaintance) is a suspect for stealing an portable music player We will also vary whether we tell them that the suspect confessed to the act or not. T hus, we her) was less likely to be altered when told the suspect had confessed than when told he did not. If this is true, then strangers (the stronger alibi ) are more likely than friends to be swayed into questioning their memory of what happened during the incident and then also more likely to choose to withhold their alibi testimony. extremely imp ortant that you do not reveal any of this information to your friend during the time he or she is waiting for the second session. If they were to find out the true purpose of the study, they would likely not be able to accurately answer any of the alibi q uestions as they would have had they not been leaked information. By signing below you acknowledge that you have read the above debriefing and understand the procedures for this study and reason for using deception. You have also been explain ed the reasoning for using a written confession and understand why this was necessary. You have also asked any and all questions you may have about this first session as well as the second session that your friend will attend. I would like to have my wri tten confession and picture used in the second session and feel I have not been coerced into doing so Yes No I agree NOT about the follow up session to my friend Yes No
92 Participant: _____________________________________ Date: _______________
93 APPENDIX I INFORMED CONSENT (PA RT 2) Informed consent Please read this consent document carefully before you deci de to further your participation in this study. Purpose of the questioning: The purpose of asking you to come in today is to fill out the closeness scales again to see if there are any delayed time effects in your closeness ratings of your friend or the other group members. What you will be asked to do: You will be asked to take both the individual and team closeness scales again and to answer a few other questions about your memory from the first session. Time required: 1 hour Risks and benefits: There is no risk to you as an informer. As a benefit, you will learn about the research process. Compensation: You will receive 2 credits for your additional time here towards the Participant Pool research requirement. If you do not wish to continue wi th the study, there are alternative assignments for you to receive an equivalent 2 credits. Confidentiality: Your identity will be kept confidential to the extent provided by law. Any information you provide will not be revealed to any person you may sus pect of committing the incident and no further actions will be taken regardless of the information you provide here. All identifying information will not be linked with your data, and we will destroy your identifying information after the session is comple te. Voluntary participation: Your participation is completely voluntary. There is no penalty for not participating. In addition, you do not have to answer any question you do not wish to answer. Right to withdraw from questioning: You have the right to withdraw from the questioning at anytime without consequence. Whom to contact if you have questions about the study: Box 117330, Gainesville, FL 32611, firstname.lastname@example.org
94 Lora Levett, Ph.D., Assistant Professor, Department of Sociology and Criminology & Law, PO Box 117330, Gainesville, FL 32611, email@example.com Whom to contact about your righ ts as a research participant in the study: UFIRB Office, Box 112250, University of Florida, Gainesville, FL 32611 2250; ph 352 392 0433. Agreement: I have read the procedure described above. I voluntarily agree to participate in the procedure and I have received a copy of this description. Participant: _____________________________________ Date: _______________ Principle Investigator: ___________________________________ Date: _________
95 APPENDIX J S ESSION 2 PROTOCOL Team Building in Friends and Stra ngers Session 2 Protocol You need to arrive at least 10 minutes before the study is set to begin. Everything you need will be on a clipboard (the timeslot and name will be written on it) so double check the name when the participant arrives. There are 3 conditions for confession timing in this study so make sure you know which lines to read before the participant arrives. 1. No confession information 2. Confession information prior to alibi question 3. Confession information after alibi question (asked twice) Be sure to fill out the alibi half sheet as much as you can before the participant arrives (condition, your initials) then fill in other details as they happen (their answers). read through the informed consent document, then sign and date at the bottom. Give the participant an IC, then collect once signed. s that you did during the first session. Again, it is critically important that you really look inward and answer each question based on how you actually feel about that person (or group) TODAY. Please do not just answer based on how you filled it out du ring the last session. Just as a reminder, the Miller scale refers to the friend that you brought with you to the first session, and the Group scale refers to the group of Hand out both closeness scales. Col lect both papers once completed. **start memorizing** session. We had someone report that their portable music player was taken from their office down the hall during the fir the people who were there individually if you have any information about how that could have happened so we can try to fix this situation and return the portable music player further Show the participant the picture of the suspect. 1. In the No Confession condition: Ask: did not take the portable music player Hand them a blank piece of paper and say: 2. In the Confession Prior condition: Show the participant the written confessio n.
96 Say: portable music player when we questioned anyone that does not think this person is the one who stole the portable music player Would you be willing to serve as an alibi for him/her, saying that they did not take the portable music player Hand them a blank piece of paper and say: Y you made 3. In the Confession After condition: Ask: did not take the portable music player ? Hand them a blank piece of paper and say: Show the participant the written confession and say: taking it when we questioned him/her earlier. Does that change your a nswer nd Hand them another blank piece of paper and: If they changed their answer say: d If they answered the same, say: alibi for this person. Please answer each of the questions as honestly as you Hand them the Alibi Questionnaire and leave the picture of the suspect on the desk while they fill it out. Collect both papers when finished. **Stop memorizing** for this study to better explain Hand them a debriefing form and read it to them. innocence af ter being convicted of a crime. This man had been at a party with several people during the time of the crime, which should have provided him with multiple alibi corroborators to stand up for him by saying that he could not have been the one who committed the crime. However, because of extensive pressure during interrogation, he ended up confessing to the crime (which he later recanted) and once word of this got out, all of his friends from the party started to question when and where they saw him and bec ause of this they no longer wanted to be an alibi for him in court. We think the confession might be what caused these friends to go back and rethink about the night in question and to withdraw their original story. In this study, we are looking to examin e if that can actually happen; basically whether or not a piece of evidence used during investigative questioning (the
97 presence or absence of a confession from the suspect) is able to influence whether a person is willing to offer their alibi for that susp ect in court. We were also interested in whether the potential alibi corroborator being a close friend, acquaintance, or a stranger could have an influence on that decision as well. So the first session served as the timeframe that a crime could have occ urred and by having you interact with different groups of people we were ensuring that you party). The second session, when you learned about the fake story of a stolen portable music player was a form of questioning that police investigators might use to feel out potential alibis and we delayed this by a few days to allow for some memory confusion as to the exact events that took place during the first meeting. In this second session, we told you that either your friend or someone else you met during the previous session (a stranger or acquaintance) was a suspect for stealing an portable music player In some conditions, we also told you that the suspect confessed to the act and showed you a written confession. We hypothesized that you, as potential alibis, would be most likely to withdraw your alibi for the suspect when he or she was a stranger and y ou had been told about the confession. Further, we think you would be least likely to withdraw the friend and you were not given any confession information. Thus, we will be a ble than when not. If this is true, then strangers (the stronger alibi) are more likely th an friends to be swayed into questioning their memory of what happened during the incident and then also more likely to withhold their alibi testimony (and theoretically increase the chances of the wrong person being charged). Because alibi research involv you ahead of time what we were looking for because that would most likely have altered your behavior and answers to the questions during the second session. You would also be likely to watch others acti ons more closely during the first session and to question your friend about their behavior during the time between sessions. This delay time was needed to recreate a real world possibility of memory decay between the crime and being called as an alibi cor Point out the IRB questions on the second page, then read: in detail what it is that made you feel anxious or distressed during the study and what we can do to pr Give them time to answer these questions, then say: If th included in the results, you are free to withdraw your data from the sample. If you choose to do so, i t will be destroyed immediately with no penalty to you. If you have any questions feel free to ask them now. Because this study uses some
98 deception of the actual purpose, it is really important that you do not talk about it with anyone. This could severe ly damage our results if new participants were aware of the study hypothesis and procedures before beginning their own sessions. Your results will be kept confidential to the experimenters, and your data will not be identifiable by any of your personal in formation. The photographs taken during the first session will also be deleted immediately. Collect the debriefing form once completed (if they want one, they can keep a blank copy) and thank them again for their participation.
99 APPENDIX K ALIBI QUESTIONNAIRE Alibi Questionnaire Instructions: Please indicate your opinion on the following questions by circling the best answer choice. PART A: Alibi Corroboration and Confidence 1. I would be willing to provide an alibi for this person. 1 2 Yes No 2. I am 100% confident that the suspect is the person who took the portable music player during the first session. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 3. I am unsure if this suspect is the actual perpetrator. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 4. I have doubt about whether this suspect was the only one who could possibly have stolen the portable mu sic player 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 5. There is nothing that could convince me that the perpetrator was anyone other than this suspect. 1 2 3 4 5 6 Strongly Disagree Disagree Somewh at Disagree Somewhat Agree Agree Strongly Agree 6. There is no doubt in my mind about my decision to offer an alibi for this suspect. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 7. I am hesitant to pro vide an alibi for this person because I think they might have actually committed the act. 1 2 3 4 5 6 Strongly Disagree Somewhat Somewhat Agree Strongly
100 Disagree Disagree Agree Agree 8. I question if serving as an alibi for this person is something I s hould do. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree during the night in question. 1 2 3 4 5 6 Strongly Disagree Disa gree Somewhat Disagree Somewhat Agree Agree Strongly Agree 10. If I were on a jury hearing this case, I would convict the suspect. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 11. I believe the suspect is innocent of what he or she is being accused of. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree PART B: Alibi Depth 12. I would be willing to tell the suspect that I believe h e was NOT the one who took t he portable music player 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 13. I am unsure of whether I would want to get involved in the alibi corroboration process. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree an alibi corroborator for the suspect. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Ag ree Strongly Agree what happened during the first session. 1 2 3 4 5 6
101 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 16. I would be wi 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 17. I question whether vouching for this suspect to his employer is something I want to do. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 18. I would be willing to provide an alibi for this suspect during a police interview. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strong ly Agree 19. I am unsure if I would put myself through police questioning in order to 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 20. I would be willing to provide an a libi for the suspect by talking with both the trial attorneys. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 21. I am sure talking with the trial attorneys would not be a problem for me to do. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 22. I would be willing to provide an alibi for the suspect that required me to take a polygraph test. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat A gree Agree Strongly Agree 23. I would hesitate to submit to a polygraph test in order to vouch for this suspect. 1 2 3 4 5 6 Strongly Disagree Somewhat Somewhat Agree Strongly
102 Disagree Disagree Agree Agree 24. I would be willing to provide an alibi f or the suspect by testifying in a criminal trial. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 25. I would worry about going to court in order to testify for this suspect. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree PART C: Memory for the Event and Viewing Conditions 26. I believe my memory for the event in question is good. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 27. I am NOT sure if I remember exactly what happened during the first session. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree first session. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 29. During the first session, there are more than 15 minutes in which I am unsure of where the suspect was. 1 2 3 4 5 6 Strongly Disagree Dis agree Somewhat Disagree Somewhat Agree Agree Strongly Agree 30. I believe I could recite everything that happened during the first session. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 31. I am confiden t that my account of what happened is accurate. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree
103 32. I had a good view of the suspect throughout the first session. 1 2 3 4 5 6 Strongly Disagree Disagree Som ewhat Disagree Somewhat Agree Agree Strongly Agree 33. My view of the suspect and his or her actions were clear during the entire first session. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 34. I am NOT session. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 36. I did NOT have a clear view of the suspect for most of the first session. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree S trongly Agree 37. My view of the suspect was obstructed during some portion of the first session. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 38. I had a good view of the suspect taking the portable m usic player 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 39. I consider the suspect a friend. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 40. I was to ld during questioning that the suspect confessed to taking the portable music player 1 2 3 4 5 6
104 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 41. I had NOT met the suspect prior to the first session. 1 2 3 4 5 6 S trongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 42. I believe the suspect did NOT confess to the crime in question. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 43. I bel ieve the suspect confessed to the crime in question. 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree 44. I think the suspect was coerced into confessing about taking the portable music player 1 2 3 4 5 6 Strongly Disagree Disagree Somewhat Disagree Somewhat Agree Agree Strongly Agree PART D: Demographics 45. What race do you identify with? 1 2 3 4 5 6 Asian/Pacific Islander Black Cuban Hispanic Latino White Other, please specify 46. Are you a US citizen? 1 2 Yes No 47. What is your age? 48. What is your gender? 1 2 Male Female 49. What is your current school classification? 1 2 3 4 5 6 Freshman Sophomore Junior Senior Grad Not
105 Student Enrolled 50. What is your current ma jor? 51. Do you have a job (besides being a student)? 1 2 Yes No Please specify
106 Were you at all suspicious of the true intent of the study? Yes No (if yes, please explain what you believe this study is investigating) Did your friend (or anyone else) reveal anything about the true study goals during the time between sessions? Yes No (if yes, please explain what the y said)
107 APPENDIX L FINAL DEBRIEFING Confession Influence on Alibi Corroboration Final Debriefing (for participants after 2 nd session) The basis for this study was an Innocence Project client who maintained his innocence after being convicted of a crime. This man had been at a party with several people during the time of the crime, which should have provided him with multiple alibi corroborators to stand up for him by saying that he could not have been the one who committed the crime. However, because of extensive pressure during interrogation, he ended up confessing to the crime (which he later recanted) and once word of this got out, all of his friends from t he party started to question when and where they saw him and because of this they no longer wanted to be an alibi for him in court. We think the confession might be what caused these friends to go back and rethink about the night in question and to withdr aw their original story. In this study, we are looking to examine if that can actually happen; basically whether or not a piece of evidence used during investigative questioning (the presence or absence of a confession from the suspect) is able to influenc e whether a person is willing to offer their alibi for that suspect in court. We were also interested in whether the potential alibi corroborator being a close friend, acquaintance, or a stranger could have an influence on that decision as well. So the f irst session served as the timeframe that a crime could have occurred and by having you interact with different groups of completely know the actions of every other person at a party). The second session, when you learned about the fake story of a stolen portable music player was a form of questioning that police investigators might use to feel out potential alibis and we d elayed this by a few days to allow for some memory confusion as to the exact events that took place during the first meeting. In this second session, we told you that either your friend or someone else you met during the previous session (a stranger or ac quaintance) was a suspect for stealing an portable music player In some conditions, we also told you that the suspect confessed to the act and showed you a written confession. We hypothesized that you, as potential alibis, would be most likely to withdra w your alibi for the suspect when he or she was a stranger and you had been told about the confession. Further, we think you would be least likely to withdraw the alibi (and u were not given any confession information. Thus, we will be able to see if the least is true, then strangers (the stronger alibi) are more likely than friends to be swayed into questioning their memory of what happened during the incident and then also more likely to withhold their alibi testimony (and theoretically increase the chances o f the wrong person being charged). tell you ahead of time what we were looking for because that would most likely have altered your behavior and answers to the questions during th e second session. You
108 would also be likely to watch others actions more closely during the first session and to question your friend about their behavior during the time between sessions. Th is delay time was needed to recreate a real world possibility of memory decay between the crime and being called as an alibi corroborator. The Institutional Review Board (IRB) is concerned that this study may cause some adverse reactions or trauma in some participants, so please fill out the following questions as fully and honestly as you can to better assess the effects this study had on you during your participation. Was there anything about this study that made you upset, distressed, anxious, or uncom fortable? Yes No (if yes, please explain what) (If you answered yes to this question, please do not hesitate to ask the researcher for information about the student counseling center or to further explain the need for deception in this study). Did you feel you were pressured into offering an alibi for the suspect? Yes No (if yes, please explain why) Do yo u think the procedures we used were justified to investigate this topic, given what you know now? Yes No (if no, please explain why not) Please rank the amount of stress you felt during this study (circle one) No Stress A little Stress Moderate Stress A lot of Stress Extreme Stress What was your reaction to the study like? (circle one)
109 Positive, I enjoyed my experience Negative, I did not enjoy my experience Again, if you wish to have the data from this (or any) part of the study not included in the results, you are free to withdraw your data from the sample. If you choose to do so, it will be destroyed immediately with no penalty to you. If you have any questions feel free t o ask them now. Because this study uses some deception of the actual purpose, it is really important that you do not talk about it with anyone. This could severely damage our results if new participants were aware of the study hypothesis and procedures be fore beginning their own sessions. Your results will be kept confidential to the experimenters, and your data will not be identifiable by any of your personal information. The photographs taken during the first session will also be deleted immediately. Thank you again for your participation. I have read the procedure detailed above. I understand the purpose and content of the study and have asked any questions I may have. I would like to have my data included in the study: Yes No Par ticipant: _____________________________________ Date: _______________
110 LIST OF REFERENCES Ayling, C. J. (1984). Corroborating Confessions: An Empirical Analysis of Legal Safeguard Against False Confessions. Wisconsin Law Review, 19844 1121 1204. The Ben jamin N. Cardozo School of Law at Yeshiva University. (2008). The Innocence Project. www.innocenceproject.org. Bering, J. M., & Shackelford, T. K. (2005). Evolutionary Psychology and False Confession. American Psychologist, 60 (9), 1037 1038. Ceci, S., & B ruck, M. (1993). Suggestibility of the Child Witness: A Historical Review and Synthesis. Psychological Bulletin 113(3), 403 439. Charman, S., Cahill, B., Leins, D., & Carol, R., 2009. [PowerPoint slides]. Retrieved during American Psychology Law Society A nnual Conference. 2009, Vancouver, Canada. Coffman, K. J., & Henkel, L. A. (2004). Memory Distortions in Coerced False Confessions: A Source Monitoring Framework Analysis. Applied Cognitive Psychology, 18 (5), 567 588. Connelly, P. (1983). Alibi: Proof of F alsehood and Consciousness of Guilt. Criminal Law Quarterly, 166(25), 165 178. Culhane, S., (2009) [PowerPoint slides]. Retrieved during American Psychology Law Society Annual Conference Journal of Applied Social Psychology 34(8), 1604 1616. Douglass, A., & Steblay, N. (2006). Memory Distortion in Eyewitnesses: A Meta Analysis of the Post identification Feedback Effect. Applied Cognitive Psychology 20(7), 859 869. Driz in, S. A., & Leo, R. A. (2004). The Problem of False Confessions in the Post DNA World. North Carolina Law Review 82, 891 1007. Dror, I. & Charlton, D. (2006). Why Experts Make Errors. Journal of Forensic Identification. 56(4), 600 617. Dror, I., Charlton D., & Peron, A. (2006). Contextual Information Renders Experts Vulnerable to Making Erroneous Identifications. Forensic Science International, 156, 74 78. Dror, I., Peron, Hind, & Charlton, D. (2005). When Emotions Get the Better of Us: The Effect of Con textual Top down Processing on Matching Fingerprints. Applied Cognitive Psychology, 19, 799 809.
111 Dysart, J., 2009. [PowerPoint slides]. Retrieved during American Psychology Law Society Annual Conference. 2009, Vancouver, Canada. Elaad, E. (2003). Effects o f Feedback on the Overestimated Capacity to Detect Lies and the Underestimated Ability to Tell Lies. Applied Cognitive Psychology, 17 (3), 349 363. Festinger, L. (1954). Social Psychology and Group Processess. Annual Review of Psychology, 187 216. Gold, G. J., & Weiner, B. (2000). Remorse, Confession, Group Identity, and Expectancies about Repeating a Transgression. Basic and Applied Social Psychology, 22 291 300. Hall, L. & Player, E. (2008). Will the Introduction of an Emotional Context Affect Fingerpri nt Analysis and Decision making? Forensic Science International, 181, 36 39. Hartwig, M., Granhag, P., Stromwall, L., & Kronkvist, O., (2006). Strategic Use of Evidence During Police Interviews: When Training to Detect Deception Works, Law and Human Behavi or, 30(1), 603 619. Hasel, L., & Kassin, S. (2009). On the Presumption of Evidentiary Independence: Can Confessions Corrupt Eyewitness Identifications? Psychological Science 20(1), 122 126. Kassin, S. M. (1997). The Psychology of Confession Evidence. Amer ican Psychologist, 52 (3), 221 233. Kassin, S. M. (2005). On the Psychology of Confessions: Does Innocence Put Innocents at Risk? American Psychologist, 60 (3), 215 228. Kassin, 2009.[PowerPoint slides]. Retrieved during American Psychology Association Ann ual Conference. 2009, San Francisco, CA. Kassin, S. M., & Kiechel, K. L. (1996). The Social Psychology of False Confessions: Compliance, Internalization, and Confabulation. Psychological Science 7, 125 128. 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 (2), 211 227. Kassin, S. M., & Sukel, H. (1997). Coerced Confessions and the Jury: An Experimental Test of the "Harmless Error" Rule. Law and Human Behavior 21, 27 46. Kassin, S. M., & Wrightsman, L. S. (1980). Prior Confessions and Mock Juror Verdicts. Journal of Applied Social Psychology 10(2), 133 146.
112 Kassin, S. M., & Wrightsman, L. S. (1981) Coerced Confess ions, Judicial Instruction, and Mock Juror Verdicts. Journal of Applied Social Psychology 11(6), 489 506. Kassin, S. M., & Wrightsman, L. S. (1985). Confession Evidence. In S. M. Kassin & L. S. Wrightsman (Eds.), The Psychology of Evidence and Trial Proce dure (pp. 67 94). Beverly Hills, CA: Sage Leo, R. A. (2009). False confessions: Causes, Consequences, and Implications. Journal of the American Academy of Psychiatry and the Law, 37(1), 332 343. Loewy, A. H. (2007). Systemic Changes that Could Reduce the C onviction of the Innocent. Criminal Law Forum, 18 (1), 137 149. Marion, S., & Burke, T., 2009. [PowerPoint slides]. Retrieved during American Psychology Law Society Annual Conference. 2009, Vancouver, Canada. Markus, H. (1977). Self schemata and Processing Information About the Self. Journal of Personality and Social Psychology, 35, 63 78. Miller, R., & Lefcourt, H. (1982). The Assessment of Social Intimacy. Journal of Personality Assessment, 46(5), 514 518. Neuschatz, J., Lawson, D., Swanner, J., Meissner, C., & Neuschatz, J. (2008). The Effects of Accomplice Witnesses and Jailhouse Informants on Jury Decision Making. Law and Human Behavior, 32, 137 149. New York v. Kogut. 800 N.Y.S.2d 353. (2004) Olsen, E., & Wells, G. (2004). What Makes a Good Alibi? A Pr oposed Taxonomy. Law and Human Behavior, 28(2). Olson, E., & Wells, G. (2010). The alibi generation effect: Alibi generation experience influences alibi evaluation. Legal and Criminological Psychology doi:10.1111/j.2044 8333.2010.02003.x. Risinger, D. M., Saks, M. J., Thompson, W. C. & Rosenthal, R. (2002). The Daubert/Kumho Implications of Observer Effects in Forensic Science: Hidden Problems of Expectation and Suggestion. California Law Review, 90(1), 1 56. Robinson, J., & Swanson, K., (1990). Autobiogra phical Memory: The Next Phase. Applied Cognitive Psychology, 25(1), 321 335. Russano, M. B., Meissner, C. A., Narchet, F. M., & Kassin, S. M. (2005). Investigating True and False Confessions Within a Novel Experimental Paradigm, Psychological Science 16(6 ), 481 486. Scheck, B., Neufeld, P., & Dwyer, J., (2000). Actual Innocence: Five Days to Execution, and Other Dispatches From the Wrongly Convicted. New York: Doubleday.
113 Schoenfeld, H. (2005). Violated Trust: Conceptualizing Prosecutorial Misconduct. Journ al of Contemporary Criminal Justice 21, 250 271. Semmler, C., Brewer, N., & Wells, G. (2004). Effects of Postidentification Feedback on Eyewitness Identification and Nonidentification Confidence. Journal of Applied Psychology, 89(2), 334 346. Smith, S., P atry, M., & Stinson, V. (2007). But What is the CSI Effect? How Crime The Canadian Journal of Police & Security Services 5(3), 1 10. State v. Myers, 359 S.C. 40 (2004). Thompson, W. C. (1997). A Sociological Perspective on the Science of Forensic DNA Testing. U.C. Davis Law Review 30(4) 1113 1136. Tulving, E. (1972). Episodic and Semantic Memory. In E. Tulving & W. Donaldson (Eds.), Organization of Memory 381 403. New York: Academic Press. Tyl er, T. (2006). Viewing CSI and the Threshold of Guilt: Managing Truth and Justice in Reality and Fiction. The Yale Law Journal 115, 1050 1085. Wagenaar, W. A. (2002). False confessions after repeated interrogation: The Putten murder case. European Review, 10 (4), 519 537. Ware, L., Lassiter, D., Patterson, S., & Ransom, M. (2008). Camera Perspective Bias in Videotaped Confessions: Evidence That Visual Attention Is a Mediator. Journal of Experimental Psychology: Applied, 14(2), 192 200. Wells, G., Small, M., Penrod, S., Malpass, R., Fulero, S., & Brimacombe, C. (1998). Eyewitness Identification Procedures: Recommendations for Lineups and Photospreads. Law and Human Behavior, 22(6), 1 39. Woody, W., & Forrest, K. (2009). Effects of False Evidence Ploys and Exp ert Testimony Evidence. Behavioral Sciences and the Law, 27, 333 360.
114 BIOGRAPHICAL SKETCH Megan Kienzle earned her Bachelor of Science degree in psychology (minoring in Cr iminal Ju stice Studies) from Iowa State University in 2008 After graduation, she moved to Florida and began work towards the Master of Arts degree at the University of Florida with plans of obtaining the Doctor of Philosophy degree from the same institution He r research interests generally lie in the intersection of psychology and law; more specifically, she researches jury decision making, the death penalty, and factors associated with wrongful convictions.