Skip to main content
NCBI home page
Psychiatry, Psychology, and Law logoLink to Psychiatry, Psychology, and Law
. 2018 May 31;25(3):425–436. doi: 10.1080/13218719.2018.1463875

Unmaking a murderer: behaviour sequence analysis of false confessions

David A Keatley a,c,✉, Abbie Marono a, David D Clarke b
PMCID: PMC6818295  PMID: 31984030

Abstract

The causes of false confessions are an important issue in legal studies and forensics. Recent advances in DNA testing have increased the number of proven false confessions; however, there are probably many cases without scientific evidence to refute the guilty verdict. The current research provides a novel approach to understand the structure and process of police interrogation techniques, in the United States behaviour sequence analysis. This method allows complex interactions in interrogations to be broken down and the progression of techniques analysed clearly. A case study is provided of an individual who confessed to a series of very serious crimes. The results show that several psychological techniques, such as leading statements, pressure, empathy, and inducements, are used, in increasing frequency, which resulted in the suspect's confession. This research provides support for behaviour sequence analysis as a new method to understand the structure of police interrogations and how psychological techniques may be used to gain false confessions.

Keywords: behaviour sequence analysis, false confession, interrogations, legal system, Reid Technique

Despite the recent rise in media portrayals of false confessions, many people are still unable to understand why an innocent individual would confess to a crime they did not commit. Indeed, as the seriousness of the crime increases (e.g. homicide), people find it even harder to understand and accept that an individual would falsely confess to the crime (Drake, 2011; Henkel, Coffman, & Dailey, 2008; Leo & Liu, 2009; Woestehoff & Meissner, 2016). To date, however, the Innocence Project (2006) states that more than 220 individuals have been exonerated and released from prison, after post-conviction DNA testing. While this raises awareness of cases that could be scientifically overturned by DNA evidence, it leaves the question of a much larger issue: how many guilty verdicts have been made due to false confessions. Understanding the processes of why and how individuals may give false confessions is a central concern in legal systems around the world, especially in America (Kassin, 2008; Leo & Ofshe, 1998). False confessions obviously have the immediate issue of an innocent person being wrongly convicted and sent to prison; however, a further corollary is that a guilty person is left free to commit further crimes.

Confessions hold considerable weight in the legal system, especially in influencing jury members’ verdicts (Leo, 2008; Leo & Davis, 2010; Leo & Ofshe, 1998). The police investigation approach in the United States typically begins with police conducting interviews to gather information and assess the guilt of the suspect, known as the Behaviour Analysis Interview (BAI; Vrij, Mann, & Fisher, 2006). A major issue with this stage of the investigation is that police officers have an overconfidence in their own ability to detect deception and presume guilt; therefore, many police wrongly believe they only interrogate guilty suspects (Gudjonsson & Pearse, 2011; Kassin & Gudjonsson, 2004). The Reid Technique (Inbau, Reid, Buckley, & Brian, 2001), which is the main training manual for most criminal interrogators in America, states that police officers can be trained to detect suspects’ deception in the BAI to a high degree of accuracy, around 85% (Meissner & Kassin, 2002), though empirical evidence suggests that there is no evidence for this (Vrij et al., 2006). However, because Inbau and colleagues (2001) claim that the BAI leads to investigators making accurate decisions on suspects’ guilt, they then claim that only guilty suspects are interrogated. As only presumed guilty suspects are interrogated, this changes the police approach from information gathering to extracting a confession from the guilty suspect. This is an example of the principle of explosion: ex falso sequitur quodlibet ∼ from falsehood, anything follows; while investigators are proceeding under the false assumption that their ability to detect deceit is higher than it actually is, this opens the possibility for innocent people to be wrongly moved from the BAI into the second stage of the interrogation process: the adversarial interrogation.

Recent developments in interrogation techniques have moved away from physical threat and coercion to psychological techniques and processes, designed to gain confessions (Kassin, 2017; Kassin & Kiechel, 1996; Leo, 1992; Leo & Davis, 2010). The issue remains that many lay people and juries do not believe that psychological persuasion techniques are able to cause false confessions; indeed, the developers of the modern interrogation Reid Technique have explicitly claimed that psychological techniques cannot result in false confessions (Inbau et al., 2001). A number of techniques are used in police interrogations, including: isolating the suspect; placing the suspect in a stress- and anxiety-raising environment; interrogators exhibiting confidence in their assumption of guilt; and use of themes (e.g. moral excuses for suspect committing the crime). Investigators are also at liberty to use deception and trickery to convince a suspect that there is strong, scientific evidence to prove their guilt. DNA evidence, polygraph results, and various other forms of (pseudo-)scientific evidence such as statement validity analysis and computer voice stress analysis (CVSA) may be alluded to during the course of an interrogation to convince the suspect that denial is futile. Investigators may then use a number of other techniques to encourage suspects to begin confessing to a crime, including: good-cop bad-cop routines; minimising the seriousness of the crime; and changing suspects’ perceptions of severity of the crime and punishment.

Investigators are aware that they cannot force or directly coerce confessions from suspects; however, it is fair to say that many investigators may not be aware of the implicit influence of their language. For instance, understanding the difference between an open and a leading question may require further training, and even then it is easy to understand how a statement may be made during an interrogation that may be leading or otherwise present information that, when echoed by the suspect, appears to show privileged knowledge. Statement validity analysis (SVA) has been claimed to provide clear and accurate indicators of a suspect's guilt based on the type of language they use (Adams, 1996). However, recent research has called into question the validity of SVA, and indicated it is not an accurate means of identifying guilt (Miller & Stiff, 1993; Shuy, 1993). Understanding the structure and process of interrogations, however, remains an important area of research. Interrogations are intrinsically dynamic interactions between two or more individuals. Understanding the sequence of interactions and which psychological techniques are used throughout the course of the interrogation may begin to highlight the process through which (false) confessions are created. What is required, therefore, is a method for mapping the dynamic sequence of police interrogations, highlighting the transition between different psychological techniques, suspects’ responses, and eventual confessions.

Behaviour sequence analysis

Behaviour sequence analysis (BSA; Keatley, Barsky, & Clarke, 2017; Marono, Clarke, Navarro, & Keatley, 2017), also called lag-sequence analysis (LSA), allows temporal chains of (verbal and nonverbal) behaviours to be mapped and presented in clear diagrams that can be used to show which psychological techniques are used in interrogations. BSA approaches allow complex episodes in time to be broken down, or parsed, into discrete categories of behaviours or events. Categories should be mutually exclusive and exhaustive (Bakeman & Gottman, 1986; Bakeman, Gottman, & Osofsky, 1987). The categories are then analysed in terms of their sequential order: for instance, interrogator expresses empathy followed by suspect answers question. Long sequences or chains of categories can then be built, based on mathematical Markov models (Ivanouw, 2007).

In lag one BSA, which is the most common form of sequence analysis and the one used in the current study, transitions between categories that directly follow each other are calculated, for instance, A→B. In this example, A (which represents a particular behaviour or event) is the antecedent, and B (which represents another behaviour or event) is the sequitur. Sequence analysis allows for recurring pairs of behaviours: for instance, if an interrogator repeatedly makes leading comments, then chains of A→A may occur. Over long durations, like interrogations, lengthier chains of behaviours may occur: A→B→C→n. In lag one BSA, sequences are analysed in one-step transitions (i.e. A→B, B→C etc).1. Lag one BSA tests whether transitions between pairs of behaviours occur above the expected level of chance (Bakeman & Gottman, 1986).

Sequence analysis has previously been used in a variety of criminal and legal contexts, including: drink driving (Keatley et al., 2017), behavioural deception detection (Marono, Clarke, Navarro, & Keatley, 2017a, 2017b), violent episodes (Taylor, Keatley, & Clarke, 2017), rape cases (Ellis, Clarke, & Keatley, 2017; Fossi, Clarke, & Lawrence, 2005; Lawrence, Fossi, & Clarke, 2010), homicides (Keatley, Yaksic, & Reid, 2017), and courtroom dynamics (Gnisci & Bakeman, 2007). In relation to the present research, to the authors’ knowledge, there is no previous research using BSA to analyse legal interrogations. Therefore, the current research should be seen as the foundation and first step towards building a library of cases resulting in false confessions.

Present study

The aim of the present research is twofold: first, to show how BSA can be used to show patterns in psychological techniques used in interrogation methods; second, to provide the basis for future research to build on. The BSA approach is an additive research paradigm, allowing future research to be combined with existing research findings to build more complex patterns and pathways. Therefore, this research can be extended and grown in the future. The current study focused on a false confession made by an individual with learning difficulties (i.e. an IQ of 70). Given the wealth of research showing the vulnerability of young individuals (Gudjonsson, Sigurdsson, & Sigfusdottir, 2010) and individuals with learning or mental difficulties (Redlich, Summers, & Hoover, 2010), the present research focuses directly on arguably one of the biggest areas of concern. Although the present research is exploratory in its approach, a number of hypothesis about the types of psychological techniques expected to be shown can be made. Techniques such as persuasion, leading comments, empathy, and proclaiming to know the truth are key techniques used in the Reid Technique. All of these techniques are likely to occur in the current case; however, the order and repetition of these techniques are not clear.

Method

Sample

The current study involved a 17-year-old male, at the time of interrogation, with an IQ of around 70, who was originally arrested for suspicion of being party to first-degree murder, mutilation of a corpse, and second-degree sexual assault, when he was 16. The defendant was interrogated on four recorded occasions; though other interviews and interrogations are known to have taken place, no records have been made.2. All police interrogations were made without the presence of a legal representative, parent, or other adult supervision on behalf of the defendant. The defendant, though denying involvement to begin with, eventually made a full oral and written confession to the crime. The defendant later recanted the confession; however, the confession was still provided at trial, which resulted in a guilty verdict on the counts of intentional homicide, rape, and mutilation of a corpse. The defendant's intellectual limitations were ruled out during the court proceedings, as the cut-off is 70, and the defendant was said to be in the borderline range.

The detectives involved in the interrogation that is being analysed in the current study followed the Reid Technique throughout the interrogation. This needs to be highlighted, as the current research is based on a standard Reid Technique interrogation and, therefore, is similar to other interrogations using this approach. It would be wrong and oversimplified to explain the current false confession simply as a one-off, unique case. The detectives involved followed a procedure that is widely used and supported throughout the United States.

Materials

Legal transcripts of the first interrogation was gained from verified legal sites. Written transcripts made from recordings were used in the current analyses. As written transcripts were analysed, exact timing of categories could not be ascertained; however, the sequential nature of interrogation transcripts does allow for accurate sequences to be made. Though timings are not available for the current transcripts, the average length of the first interrogation was approximately 40 min.

Coding procedure

Coding was deliberately made as simple as possible in the current research. Over-complicated coding schemes, though providing more idiosyncratic details of an investigation, would probably reduce generalisability to future cases. The coding list (see Supplementary Materials S1) used in the current case was largely based on the techniques outlined in the Reid Technique, including: leading facts (split into leading facts related to ‘locations’, ‘times’, ‘facts’); punishments; empathy; reassurance; rewards; and so on. Two coders independently developed the coding list and then checked to ensure agreement on speech acts in the current interrogations. Ambiguous cases were discussed and resolved through agreement of when particular speech categories (i.e. ‘leading location’) began and ended. There were no disagreements about codes, per se, just ensuring that start and end points aligned. The codes for each interrogator were duplicated, so that analyses could clearly show which interrogator was talking and using different verbal techniques. A third coding list was developed for the defendant, indicating whether they were agreeing, providing new facts, or repeating facts given by the interrogators. To emphasise the effects of leading statements, categories of ‘echo’ (indicating the defendant was immediately echoing what he had just heard) and ‘retell a story/script’ (indicating the defendant was repeating an earlier piece of information given by the detectives) was included in the coding scheme.

To ensure that the complete interrogation was captured in the sequence analysis, categories were included to be used when conversation was not directly related to anything involved in the case, so-called ‘small talk’. Interrogators are trained to use small-talk and non-accusatory talk to build rapport. Therefore, ‘talk’ was included as a category to stand for any conversation act that did not appear to be directly relevant to the case, and can be understood as a form of rapport building that does not fit into any other defined category. This approach to categorisation, use of an ‘other’ category, is typical and is encouraged in sequence analysis research, so as to provide exhaustive lists and overviews of analysed episodes.

Statistical analysis

The interrogation transcripts were analysed using a lag one sequence analysis approach. This means that transitions between pairs of categories were analysed, and then put into longer chains. This is important to remember when reading the results. The lag one analysis is between pairs of behaviours, A→B, B→C, and so on. It is possible to connect these pairs into longer chains, A→B→C, and indeed the state transition diagram does this. However, interpretation should be made between pairs of categories only, in single-transition steps. The sequences were input into the statistical software R (R Core Team, 2013), using a program developed by the research team to analyse the data. The program first outputs frequencies of individual categories, followed by the lag one sequence analysis.

Results

Within the first interrogation, the frequencies of each category are shown in Table 1. The table indicates that one of the interrogators, W, asked a lot of factual questions (n = 82). The next most frequently observed categories were: F (the second interrogator) offering reassurance to the suspect (n = 66), and the suspect saying yes/agreeing with the interrogators (n = 66). It should be noted, however, that in the current dataset leading questioning was split into leading questions based on different subjects (i.e. time, facts, location). If the leading questioning is grouped at a meta-level, ignoring specifics, then a total of 71 leading speech acts are made (36 by interrogator F and 35 by interrogator W). There are also 46 times when either F or W answers for Brendan, thus possibly feeding him information that he later retold in a confession, which appears to be first-hand knowledge of the crime. Indeed, there are 15 times when Brendan immediately echoes information he has been given, 7 times where he repeats the story or script he has been given, and 7 times when his story explicitly changes. None of these factors alone point definitively towards guilt or innocence; but, it does show that the process of the interrogation altered the story being told by Brendan.

Table 1.

Frequencies of individual categories used in the interrogation.

Behaviour Frequency
W: ask fact 82
B: yes/agree 66
F: reassurance 66
F: ask fact 60
B: silence/unheard 56
B: answer; W: reassurance 40
F: leading fact; W: leading fact 32
F: answer for Brendan 29
B: no/disagree 25
W: ask location 24
W: open question 21
F: empathy; F: open question 20
F: proclaim to know truth; W: yes 18
W: answer for Brendan 17
F: something bothering you 12
B: echo fact; B: state location; F: ask time; F: Help yourself/tell truth 11
B: state fact; F: interrupt Brendan; W: Help yourself/tell truth; W: proclaim to know truth 10
B: unrelated; F: ask location; W: ask time 9
B: state time; W: interrupt Brendan 8
P: retell story/script 7
W: empathy; W: no/deny 6
F: interrupt each other; W: interrupt each other; W: turn against Steven 5
B: echo time; B: maybe; F: social pressure 4
B: state ‘Steven’; B: story change – time; F: turn against Steven 3
B: answer Teresa; B: story change – fact; B: story change -‘Steven’; F: leading ‘Teresa’; F: no/deny; F: threaten; F: yes; W: leading time; W: something bothering you 2
B: correct statement; B: incorrect statement; F: ask fact; F: leading location; F: leading ‘Steven’; F: punish; F: reward; W: leading location; W: reward; W: social pressure 1
Open in a new tab

Note: F = second interrogator;   W = first interrogator;   B = Brendan. Categories with the same frequency have been grouped together.

A lag one behaviour sequence analysis was conducted on the interrogation transcripts. The analyses provides a transition frequency matrix (see Supplementary Material S2), and this table can be plotted in a state transition diagram (see Figure 1). The state transition diagram shows the significant transitions between pairs of behaviours. Only those transitions that occur significantly above the level of chance are included in the state transition diagram. This provides an illustration of techniques in the interrogation as a whole, and shows how behaviours are connected. The diagram should be read one step at a time between behaviours.

Figure 1.

Figure 1.

Open in a new tab

State transition diagram of behaviours used throughout the interrogation. F = second interrogator; W = first interrogator; B = Brendan. Key provided in diagram, representing standardised residual values. Figure shows only behaviours of frequency 3 and above.

The strongest transition in the BSA, measured by the standardised residual (SR), was F repeating W's statement that the witness should turn against the other suspect in the crime (n = 2, SR = 15.30). Effectively, whenever W suggested that the suspect should turn on the other named suspect, F immediately repeated this statement. In terms of other behaviours related to the Reid Technique, the data show that F was likely to apply social pressure to the suspect, and immediately follow this with sentiments that the suspect should help himself by helping the investigators (n = 2, SR = 8.79). Essentially, this shows a build-up of techniques occurring, rather than one technique being used and then waiting for a response. The interrogators layer or chain psychological techniques outlined in the Reid Technique. Other examples of this include both F (n = 24, SR = 8.66) and W (n = 12, SR = 12.63) repeating multiple reassuring comments in sequence.

The data show that on multiple occasions F asks a factual question, and Brendan disagrees (n = 11, SR = 7.19). There is also a significant transition from Brendan replying with a ‘no’ or disagreement leading to W then answering for Brendan (n = 3, SR = 3.66). There are also multiple occasions when the interrogators ask a leading factual question, and Brendan retells the story/script he has been given earlier by the interrogators (n = 2, SR = 3.49). The fine-grained approach to categorising behaviours can sometimes make clear patterns harder to see in the diagrams. Therefore, to reduce complexity, categories of similar behaviours can be grouped together. For instance, if all instances of leading questioning are combined for each interrogator, then we see how leading questions are typically stacked in sequence (see Figure 2). For example, F making a leading statement is significantly followed by: F making another leading statement straight away (n = 4, SR = 2.11); W making a leading statement straight away (n = 3, SR = 1.34); F immediately proclaiming to know the truth, thus making denial futile (n = 3, SR = 2.67); and F encouraging Brendan to help himself by telling the truth, thus essentially agreeing with F (n = 2, SR = 2.34). W in comparison, after he makes a leading statement, is typically followed by: Brendan agreeing (n = 7, SR = 2.74); W offering reassurance (n = 4, SR = 1.94); F making a leading statement (n = 4, SR = 2.18); W making a further leading statement (n = 5, SR = 3.10); W immediately answering for Brendan (n = 2, SR = 1.63); and Brendan retelling the story that he has just heard from W (n = 2, SR = 3.30).

Figure 2.

Figure 2.

Open in a new tab

State transition diagram of combined leading statements and sequiturs. F = second interrogator; W = first interrogator; B = Brendan. Key provided alongside diagram, representing standardised residual values.

Discussion

The aim of the present research was to present a novel way of analysing complex police interrogations. Behaviour sequence analysis (BSA) is a useful method to clearly map and show patterns of interaction in dynamic episodes, such as interrogations. It should be noted from the outset that the current research does not make definitive claims about the validity of the Reid Technique in interrogations or the innocence or guilt of Brendan Dassey. Every indication, at the time of writing, suggests that Brendan Dassey provided a false confession, and legal scholars have converged towards the opinion that the interrogation process was problematic in a number of ways. The current research makes no claims to support or diminish the views of legal scholars; however, the current research does have an impact in providing a clearer map of the pathways that lead towards the final (false) confession made by Brendan Dassey.

The current findings support previous research that has highlighted a number of psychological techniques and approaches that police investigators use in the United States (Kassin & Gudjonsson, 2004; Kassin & Kiechel, 1996; Leo & Davis, 2010). The present research documents how these methods are used sequentially. Techniques such as leading questioning and giving information that Brendan can and does echo are seen throughout the interrogation. Expressing empathy and offering reassurance to reduce Brendan's perceptions of the severity of the crime are also psychological techniques that are well documented in the interrogation literature (Kassin & Gudjonsson, 2004; Leo, 2008). There are other techniques, such as isolating the suspect, placing the suspect in a bare room, and sitting close to the suspect throughout the interrogation, which are used in interrogations, and should also be used to describe the environment of the current interrogation.

The current research shows one of two very different storylines in police interrogation practice. First, it may show the effectiveness of the Reid Technique in being able to make a guilty person finally confess to a number of very serious crimes. If this is the true case, then the method evidently has great benefit to the legal system. However, the second story is more problematic: what if the suspect is innocent? In this instance, the techniques and interrogation patterns shown in the current research indicate the risk of the Reid Technique making an innocent suspect confess to crimes they did not commit. The central issue becomes how investigators draw the line between ‘good’ and ‘bad’ interrogation practices, which is an issue perhaps beyond the scope of the current research. An issue that the current research can begin to elucidate is the sequence of behaviours that lead to false confessions being made. Though the present research is the foundational first step, the end result could be a better understanding of key pathways that are likely to lead to false confessions, in contrast to patterns in interrogations that lead to honest confessions. It may be that the methods and techniques forwarded by the Reid Technique of interrogation are valid and useful, but only if they are used in the correct order and manner. Essentially, the Reid Technique may propose methods that are incredibly useful in extracting confessions; however, particular patterns of the technique may lead to false confessions.

The analyses of the grouped leading statements, Figure 2, clearly shows the way in which psychological techniques of persuasion, leading statements, confidence in proclaiming to know the truth, reassurances, answering for the suspect, and interruption can be used in close proximity to each other in order to leave the suspect with the only options of agreeing, echoing the fact he has just been given, or repeating a statement he had previously been given by the interrogators. Again, this may indicate a prosecutorial process that is designed to not allow guilty suspects the chance to deny their guilt or make false statements of what occurred; however, it does appear in these sequences that the suspect is not given a chance to offer their version of events, amidst multiple leading statements and psychological techniques to make the suspect follow and agree with the interrogator’s storyline.

A limitation of the current research is, of course, argument about whether Brendan Dassey is innocent or guilty of the charges made against him. This, however, is also a strength as it is all too easy in research to tell ‘just so’ stories with data, once an outcome is known. For instance, analysing only those cases that have been overturned (perhaps through post-conviction DNA evidence) allows researchers to find post hoc patterns that prove innocence. The current research investigates a crime that is still slightly open to debate and, therefore, is conducted without the knowledge of hindsight. This is much more akin to the type of cases that the method could be used for in future. Though a database can be created and developed focusing on solved cases of true and false confessions, BSA can be used in real time, with unsolved or uncertain cases, as long as researchers and investigators remain vigilant of the limitations within such research. The results are only as valid as the data, and if previous interrogations or conversations are had between a suspect and the police, then by the time recorded transcripts are made and analysed, the researcher may not know the effect of previous encounters. At present, researchers are unable to give indicators of accuracy regarding sequence diagrams and conclusions of true or false confessions. What researchers and legal scholars can begin to clarify, however, is which techniques have been used and whether suspects’ stories have changed. Long interrogation interactions can be reduced to very simple, easy-to-follow sequence diagrams, which is particularly useful for interpreting complex results. Complicated statistical analyses of interactions can be shown in clear, sequential steps that jury members can easily follow and understand. The benefit of this may be that jury members begin to understand just how easy it is for police to stack psychological techniques and lead suspects to making false confessions.

Future research should complement the existing findings by including analyses of other interrogations. Research should focus on false confessions that have been proven false through DNA evidence, as well as other forms of post-conviction evidence. In addition, research should also seek to analyse true confessions to a variety of crimes, which would provide a ‘control’ condition to compare false confessions to. Advanced statistical methods, such as optimal matching (Abbott & Tsay, 2000; Rosenbaum, 1989), may then be used to analyse the differences between typical true and false confession sequences. Furthermore, researchers could use more complex behaviour sequence analysis approaches, such as T-system analysis (TSA; Magnusson, 2000, 2017), to map the interaction of multiple concurrent and sequential behaviours. T-system analysis, for instance, could be used to analyse the role of verbal and nonverbal behaviours together. TSA may bring even more clarity to the techniques that interrogators use to influence or persuade suspects (such as reassuring touching, proximity and gaze).

Conclusions

Overall, the present research highlights a novel approach to mapping interrogation dynamics, and showing how psychological techniques in the Reid Technique can be used to make a suspect provide a (false) confession. This is the first step towards a major new approach in understanding interrogation procedures, and forms the basis of perhaps being able to classify false confession sequences from true confession sequences. While this area of research is very new, it holds potential for enhanced understanding of a complex and extremely important issue: how do we tell if someone has made a false confession, and how do we simplify this complex process into an easy outcome that a jury can understand and accept?

Notes

1.

There are, of course, more complex analyses that can be conducted. These higher order sequence analyses allow for longer chains to be made (i.e. AB→C). However, higher order sequence analysis is typically not conducted for a variety of reasons (see Keatley, Barsky, & Clarke, 2017). Interested readers can access the data chains in the current research to perform their own higher order analyses.

2.

This is quite normal in American legal systems, and is an issue facing all research into police interrogations. The analyses are based on recorded transcripts; however, the effects of pre-recorded sessions can never fully be known. This limitation should be taken into account on all research into interrogation procedures, and for more information, readers are directed to Shuy (2007).

Ethical standards

Declaration of conflicts of interest

David A. Keatley has declared no conflicts of interest

Abbie Marono has declared no conflicts of interest

David D. Clarke has declared no conflicts of interest

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.

References

  1. Abbott A., & Tsay A. (2000). Sequence analysis and optimal matching methods in sociology. Sociological Methods & Research, 29(1), 3–33. Retrieved from https://doi.org/0803973233 [Google Scholar]
  2. Adams S. (1996). Statement analysis: What do suspects’ words really reveal? FBI Law Enforcement Bulletin, 65, 12–20. [Google Scholar]
  3. Bakeman R., & Gottman J. M. (1986). Observing behavior: An introduction to sequential analysis (p. 207). Cambridge: Cambridge University; Retrieved from 10.1017/CBO9780511527685 [DOI] [Google Scholar]
  4. Bakeman R., Gottman J. M., & Osofsky J. D. (1987). Applying observational methods: A systematic view. Handbook of infant development (2nd ed). Retrieved from http://search.ebscohost.com/login.aspx?direct=true&db=psyh&AN=1987-97963-016&site=ehost-live [Google Scholar]
  5. Drake K. (2011). Why might innocents make false confessions? Psychologist, 24(10), 752–755. [Google Scholar]
  6. Ellis H. E., Clarke D. D., & Keatley D. A. (2017). Perceptions of behaviours in stranger rape cases: A sequence analysis approach. Journal of Sexual Aggression. Retrieved from 10.1080/13552600.2017.1361618 [DOI] [Google Scholar]
  7. Fossi J. J., Clarke D. D., & Lawrence C. (2005). Bedroom rape: Sequences of sexual behavior in stranger assaults. Journal of Interpersonal Violence, 20(11), 1444–1466. Retrieved from 10.1177/0886260505278716 [DOI] [PubMed] [Google Scholar]
  8. Gnisci A., & Bakeman R. (2007). Sequential accommodation of turn taking and turn length. Journal of Language and Social Psychology, 26(3), 234–259. Retrieved from [DOI] [Google Scholar]
  9. Gudjonsson G. H., & Pearse J. (2011). Suspect interviews and false confessions. Current Directions in Psychological Science, 20(1), 33–37. Retrieved from 10.1177/0963721410396824 [DOI] [Google Scholar]
  10. Gudjonsson G. H., Sigurdsson J. F., & Sigfusdottir I. D. (2010). Interrogation and false confessions among adolescents: Differences between bullies and victims. Journal of Psychiatry & Law, 38(1–2), 57–77. [Google Scholar]
  11. Henkel L. A., Coffman K. A. J., & Dailey E. M. (2008). A survey of people's attitudes and beliefs about false confessions. Behavioral Sciences and the Law, 26(5), 555–584. Retrieved from 10.1002/bsl.826 [DOI] [PubMed] [Google Scholar]
  12. Inbau F., Reid J., Buckley J., & Brian J. (2001). Criminal Interrogation and confessions (4th ed.). Gaithersburg, MD: Aspen. [Google Scholar]
  13. Innocence Project (2006). Retrieved from www.innocenceproject.org
  14. Ivanouw J. (2007). Sequence analysis as a method for psychological research. Nordic Psychology, 59(3), 251–267. Retrieved from 10.1027/1901-2276.59.3.251 [DOI] [Google Scholar]
  15. Kassin S. M. (2008). False confessions: Causes, consequences, and implications for reform. Current Directions in Psychological Science, 17(4), 249–253. Retrieved from 10.1111/j.1467-8721.2008.00584.x [DOI] [Google Scholar]
  16. Kassin S. M. (2017). False confessions. Wiley Interdisciplinary Reviews: Cognitive Science. Retrieved from 10.1002/wcs.1439 [DOI] [PubMed] [Google Scholar]
  17. Kassin S. M., & Gudjonsson G. H. (2004). The psychology of confessions: A review of the literature and issues. Psychological Science in the Public Interest, Supplement. Retrieved from 10.1111/j.1529-1006.2004.00016.x [DOI] [PubMed] [Google Scholar]
  18. Kassin S. M., & Kiechel K. L. (1996). The social psychology of false confessions: Compliance, internalization, and confabulation. Psychological Science, 7(3), 125–128. Retrieved from 10.1111/j.1467-9280.1996.tb00344.x [DOI] [Google Scholar]
  19. Keatley D. A., Barsky A. D., & Clarke D. D. (2017). Driving under the influence of alcohol: A sequence analysis approach. Psychology, Crime and Law, 23(2). Retrieved from 10.1080/1068316X.2016.1228933 [DOI] [Google Scholar]
  20. Keatley D. A., Golightly H., Shephard R., Yaksic E., & Reid S. (2018). Using behaviour sequence analysis to map serial killers' life histories. Journal of Interpersonal Violence, advance online publication. [DOI] [PubMed] [Google Scholar]
  21. Lawrence C., Fossi J., & Clarke D. (2010). A sequential examination of offenders’ verbal strategies during stranger rapes: The influence of location. Psychology, Crime & Law, 16(5), 381–400. Retrieved from 10.1080/10683160902754964 [DOI] [Google Scholar]
  22. Leo R. A. (1992). From coercion to deception: The changing nature of police interrogation in America. Crime, Law and Social Change, 18(1–2), 35–59. Retrieved from 10.1007/BF00230624 [DOI] [Google Scholar]
  23. Leo R. A. (2008). Police interrogation and American justice. Cambridge: Harvard University Press. [Google Scholar]
  24. Leo R. A., & Davis D. (2010). From false confession to wrongful conviction : Seven psychological processes. Journal of Psychiatry & Law, 38, 415–422. Retrieved from https://doi.org/Article [Google Scholar]
  25. Leo R. A., & Liu B. (2009). What do potential jurors know about police interrogation techniques and false confessions? Behavioral Sciences and the Law, 27(3), 381–399. Retrieved from 10.1002/bsl.872 [DOI] [PubMed] [Google Scholar]
  26. Leo R. A., & Ofshe R. J. (1998). The consequences of false confessions: Deprivations of liberty and miscarriages of justice in the age of psychological interrogation. Journal of Criminal Law and Criminology, 88(2), 429–496. Retrieved from 10.2307/1144288 [DOI] [Google Scholar]
  27. Magnusson M. S. (2000). Discovering hidden time patterns in behavior: T-patterns and their detection. Behavior Research Methods, Instruments, & Computers : A Journal of the Psychonomic Society, Inc, 32(1), 93–110. Retrieved from 10.3758/Bf03200792 [DOI] [PubMed] [Google Scholar]
  28. Magnusson M. S. (2017). Why search for hidden repeated temporal behavior patterns: T-pattern analysis with theme. International Journal of Clinical Pharmacology & Pharmacotherapy, 1(128), p 1–7. Retrieved from 10.15344/2017/2456-3501/128 [DOI] [Google Scholar]
  29. Marono A., Clarke D. D., Navarro J., & Keatley D. A. (2017a). A behaviour sequence analysis of nonverbal communication and deceit in different personality clusters. Psychiatry, Psychology and Law. Retrieved from 10.1080/13218719.2017.1308783 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Marono A., Clarke D. D., Navarro J., & Keatley D. A. (2017b). A sequence analysis of nonverbal behavior and deception. Journal of Police and Criminal Psychology. Retrieved from https://doi.org/10.1007%2Fs11896-017-9238-9 [Google Scholar]
  31. Meissner C. A., & Kassin S. M. (2002). “He's guilty!”: Investigator bias in judgments of truth and deception. Law and Human Behavior, 26, 469–480. Retrieved from 10.1023/A:1020278620751 [DOI] [PubMed] [Google Scholar]
  32. Miller G., & Stiff J. (1993). Deceptive communication. Thousand Oaks: Sage Publications. [Google Scholar]
  33. R Core Team (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienne, Austria. ISBN 3-900051-07-0, http://www.R-project.org/ [Google Scholar]
  34. Redlich A. D., Summers A., & Hoover S. (2010). Self-reported false confessions and false guilty pleas among offenders with mental illness. Law and Human Behavior, 34(1), 79–90. Retrieved from 10.1007/s10979-009-9194-8 [DOI] [PubMed] [Google Scholar]
  35. Rosenbaum P. R. (1989). Optimal matching for observational studies. Journal of the American Statistical Association, 84(408), 1024–1032. Retrieved from 10.1080/01621459.1989.10478868 [DOI] [Google Scholar]
  36. Shuy R. W. (1993). Language crimes : The use and abuse of language evidence in the courtroom. The language library. Cambridge, MA: Blackwell. [Google Scholar]
  37. Shuy R. W. (2007). Creating language crimes: How law enforcement uses (and misuses) language. Oxford: Oxford University Press. [Google Scholar]
  38. Taylor O., Keatley D. A., & Clarke D. D. (2017). A behavior sequence analysis of perceptions of alcohol-related violence surrounding drinking establishments. Journal of Interpersonal Violence. Retrieved from 10.1177/0886260517702490 [DOI] [PubMed] [Google Scholar]
  39. Vrij A., Mann S., & Fisher R. P. (2006). An empirical test of the behaviour analysis interview. Law and Human Behavior, 30(3), 329–345. Retrieved from 10.1007/s10979-006-9014-3 [DOI] [PubMed] [Google Scholar]
  40. Woestehoff S. A., & Meissner C. A. (2016). Juror sensitivity to false confession risk factors: Dispositional vs. situational attributions for a confession. Law and Human Behavior, 40(5), 564–579. Retrieved from 10.1037/lhb0000201 [DOI] [PubMed] [Google Scholar]

Articles from Psychiatry, Psychology, and Law are provided here courtesy of Taylor & Francis

RESOURCES