THE PSYCHOLOGY OF INVESTIGATION - THE COGNITIVE INTERVIEW
Interviews are one of the most common ways of gathering information across a range of settings for a variety of reasons (Memon & Bull, 1999). In the context of crime investigation, there will be interviews with witnesses, suspects and victims, all conducted with various aims, including gathering evidence, cross-checking information and eliciting confessions (Milne & Bull, 1999). Interviewing children has become something of a speciality in its own right (Lamb et al., 1999). The less salubrious aspects of police interviewing have been highlighted by investigators of false confessions, but there are other, more constructive, aspects of the interview process to consider. A technique known as the cognitive interview [cognitive interview method of questioning witnesses, devised for use by the police, based on principles taken from memory research] illustrates the application of psychology to facilitate investigative interviewing. A great deal of the research on eyewitness testimony points to the frailties of memo ry and questions the reliability of eyewitness evidence. The cognitive interview is an attempt to find a constructive solution to these problems and improve the accuracy of eyewitness recall.
Fisher, McCauley and Geiselman (1994) describe how the original cognitive interview protocol, used by police officers, incorporated four techniques to enhance memory retrieval:
1. Context reinstatement – the witness is encouraged to recollect aspects of the situational context (such as sights and sounds at the time of the event and relevant personal factors, such as how they felt and what they were thinking at the time of the incident).
2. Report everything – the witness engages in perfectly free recall, unconstrained by focused (and potentially leading) questioning, or self-censoring of what is reported. The theory underpinning these two techniques lies in the contextual similarity between encoding and retrieval. So if the process of retrieval from memory can take place in a similar psychological context to that in which the information was encoded, the witness should have facilitated access to stored memories, improving the accuracy and completeness of recall (Fisher et al., 1994).
3. Reverse order – the witness is encouraged to begin their description of an event from different starting points (such as a mid-point), or to start at the end and work backwards to the beginning.
4. Change perspective – witnesses are encouraged try to give an account of the event from the point of view of another person, such as another witness or the victim.
Techniques 3 and 4 are intended to encourage witnesses to try to use many different paths to retrieve information from memory. If memories are stored as networks of associations, increasing the number of retrieval points should lead to more complete recall of the original event (Fisher at al., 1994). As the research and practice base developed, so the protocols for the cognitive interview expanded to include, for example, a broader range of specific questioning techniques and the use of guided imagery (Fisher & Geiselman, 1992). A body of evaluation studies, conducted in both laboratory and field settings, has accumulated since 1984. According to Milne and Bull (1999), the weight of evidence shows that the cognitive interview elicits more correct (that is, truthful) information than other types of interview. While there are some reservations, the technique is generally well received by police officers and has become widely used. Furthermore, recent research suggests that it is a reliable and helpful technique with child witnesses (Milne & Bull, 2003).
DETECTING LIES AND DECEIT
How easy is it to tell when someone is telling lies and seeking to deceive? Kassin (1997) cites several examples taken from police training manuals that suggest suspects’ verbal and nonverbal cues can be read to determine if they are lying. For example, it has been suggested that guilty suspects do not make eye contact, while innocent suspects give clear, concise answers.
It is possible that these general rules are useful, but the empirical evidence suggests that even skilled questioners are not good at detecting deceit simply on the basis of a suspect’s verbal and non-verbal cues (Ekman & O’Sullivan, 1991). Vrij (2000) suggested that most liars are caught because it becomes too difficult to continue to lie, and they have not made sufficient preparation to avoid detection. Vrij lists seven qualities that make a good liar:
1. having a well prepared story;
2. being original in what is said;
3. thinking quickly when the need arises;
4. eloquence in storytelling;
5. having a good memory for what has been said previously;
6. not experiencing emotions such as fear or guilt while lying; and
7. good acting ability.
If verbal and non-verbal cues are hard to read, how does an investigator catch out an individual who possesses all the attributes listed above? One approach is a highly structured analysis of verbal content, known as Statement Validity Assessment (SVA). Originally developed as a clinical tool for analysing children’s statements in cases of sexual abuse (Undeutsch, 1982), SVA consists of three elements:
1. A statement is taken in a structured interview.
2. The content of the statement is judged by the forensic psychologist in a criterion-based content analysis (CBCA). These content criteria are concerned with the general characteristics of the statement (such as whether it has a logical structure), the specific contents of the statement (such as descriptions of events and people), motivation-related content (such as admission of a lack of memory) and offencespecific elements (concerning the fine details of the offence).
3. The CBCA is necessarily subjective, and needs to be evaluated against a standard set of questions set in the ‘validity checklist’ (Raskin & Esplin, 1991). This checklist raises questions about the conclusions drawn from the analysis.
In other words, the content analysis itself is put to the test by systematic consideration of interviewee characteristics.
The interviewee’s psychological and motivational characteristics, the characteristics of the interview and a ‘reality check’ against other forensic evidence are all examined. It is clear that SVA represents an attempt to bring order and rigour to the essentially subjective matter of judging the veracity and reliability of an interviewee’s statement. However, in a review of the substantial evaluative literature with regard to SVA, Vrij (2000) has expressed several reservations about the technique and highlighted areas where questions remain. He concludes that ‘SVA evaluators appear to be able to detect truths and lies more accurately than would be expected by chance’ (p. 153). In other words, while not a perfect technique, SVA does help improve accuracy beyond guesswork and inaccurate beliefs about how to judge accuracy.
OFFENDER PROFILING
If ever a topic generated a great deal of heat and rather less light, offender profiling [offender profiling constructing a picture of an offender’s characteristics from their modus operandi together with the clues left at the crime scene] would be high on the list of most forensic psychologists.
But as our knowledge base increases, it is likely that the technique will become increasingly sophisticated (Ainsworth, 2001; Jackson & Bekerian, 1997). Wrightsman (2001) distinguishes between profiling historical and political figures, profiling likely criminals from crime scene characteristics, and profiling the common characteristics of known offenders. Turvey (2000) draws the distinction between inductive and deductive methods of profiling. Inductive methods rely on the expert skills and knowledge of the profiler – a method often referred to as ‘clinical’ in style. By contrast, deductive methods rely on forensic evidence, such as crime scene characteristics and offencerelated empirical data – an approach often referred to as ‘statistical’.
Profiling historical and political figures Attempts have been made to construct psychological profiles of historical figures (from Jack the Ripper to Adolf Hitler) by systematically gathering and organizing information in an effort to understand their motives and behaviour. Experts will undoubtedly have constructed psychological profiles of Saddam Hussain in order to try to predict his behaviour during the 2003 conflict in Iraq. These types of profile typically rely on specialist knowledge (e.g. military, historical).
Profiling criminals from the crime scene Way back in the late 1880s, forensic pathologists were trying to link series of crimes by the similarity of crime scene characteristics, such as the nature of a victim’s wounds. More recently, the American Federal Bureau of Investigation (FBI) pioneered an investigative system based on central features (such as the details of a crime scene and forensic evidence) in order to construct a profile of the psychological and behavioural characteristics of the criminal (Douglas et al., 1986). While forensic evidence can yield many clues, the starting point for the FBI was to use the crime scene to construct a picture of the type of person who committed the offence. This approach yielded various classifications of types of offender associated with their psychological characteristics.
For example, a much used distinction (mainly concerned with serious offenders such as murderers or rapists), incorporated within the FBI framework, is that between ‘organized’ and ‘disorganized’ offenders (Ressler, Burgess & Douglas, 1988). An organized offender will plan the offence, be careful not to leave evidence, and target the victim. The disorganized offender will seemingly offend at random, use a weapon that is discarded near to the scene of the crime, and make few attempts to hide evidence or potential clues. In terms of psychological characteristics, the organized offender is seen as intelligent and socially adjusted, although this apparent normality can mask a psychopathic personality. According to this framework, the disorganized offender is said to be less intelligent and socially isolated, may have mental health problems, and is likely to offend when in a state of panic. The obvious criticisms of such distinctions (and the FBI approach more generally) is that they are inductive, highly subjective and lacking in robust empirical validation. Profiling common characteristics of known offenders.
The third approach to profiling is to look to empirical data, rather than an expert’s opinion, to construct profiles. This approach emphasizes the rigorous gathering of data about the crime from multiple sources (such as geographical location and victim statements), the application of complex statistical analyses to databases of crime scene details (and other forensic evidence), and attempts to build a profile of the offender with theoretical integrity. Adopting this approach, Canter and Heritage (1990) analysed data from over 60 cases of sexual assault and were able to identify over 30 offence characteristics, such as level of violence, use of a weapon, type of assault and use of threats. Statistical analyses were used to search for relationships and patterns between the factors, and to build up characteristic profiles of types of sexual assault. This and other similar studies provide preliminary support for the central premises of offender profiling based on the common characteristics of known offenders.
Interviews are one of the most common ways of gathering information across a range of settings for a variety of reasons (Memon & Bull, 1999). In the context of crime investigation, there will be interviews with witnesses, suspects and victims, all conducted with various aims, including gathering evidence, cross-checking information and eliciting confessions (Milne & Bull, 1999). Interviewing children has become something of a speciality in its own right (Lamb et al., 1999). The less salubrious aspects of police interviewing have been highlighted by investigators of false confessions, but there are other, more constructive, aspects of the interview process to consider. A technique known as the cognitive interview [cognitive interview method of questioning witnesses, devised for use by the police, based on principles taken from memory research] illustrates the application of psychology to facilitate investigative interviewing. A great deal of the research on eyewitness testimony points to the frailties of memo ry and questions the reliability of eyewitness evidence. The cognitive interview is an attempt to find a constructive solution to these problems and improve the accuracy of eyewitness recall.
Fisher, McCauley and Geiselman (1994) describe how the original cognitive interview protocol, used by police officers, incorporated four techniques to enhance memory retrieval:
1. Context reinstatement – the witness is encouraged to recollect aspects of the situational context (such as sights and sounds at the time of the event and relevant personal factors, such as how they felt and what they were thinking at the time of the incident).
2. Report everything – the witness engages in perfectly free recall, unconstrained by focused (and potentially leading) questioning, or self-censoring of what is reported. The theory underpinning these two techniques lies in the contextual similarity between encoding and retrieval. So if the process of retrieval from memory can take place in a similar psychological context to that in which the information was encoded, the witness should have facilitated access to stored memories, improving the accuracy and completeness of recall (Fisher et al., 1994).
3. Reverse order – the witness is encouraged to begin their description of an event from different starting points (such as a mid-point), or to start at the end and work backwards to the beginning.
4. Change perspective – witnesses are encouraged try to give an account of the event from the point of view of another person, such as another witness or the victim.
Techniques 3 and 4 are intended to encourage witnesses to try to use many different paths to retrieve information from memory. If memories are stored as networks of associations, increasing the number of retrieval points should lead to more complete recall of the original event (Fisher at al., 1994). As the research and practice base developed, so the protocols for the cognitive interview expanded to include, for example, a broader range of specific questioning techniques and the use of guided imagery (Fisher & Geiselman, 1992). A body of evaluation studies, conducted in both laboratory and field settings, has accumulated since 1984. According to Milne and Bull (1999), the weight of evidence shows that the cognitive interview elicits more correct (that is, truthful) information than other types of interview. While there are some reservations, the technique is generally well received by police officers and has become widely used. Furthermore, recent research suggests that it is a reliable and helpful technique with child witnesses (Milne & Bull, 2003).
DETECTING LIES AND DECEIT
How easy is it to tell when someone is telling lies and seeking to deceive? Kassin (1997) cites several examples taken from police training manuals that suggest suspects’ verbal and nonverbal cues can be read to determine if they are lying. For example, it has been suggested that guilty suspects do not make eye contact, while innocent suspects give clear, concise answers.
It is possible that these general rules are useful, but the empirical evidence suggests that even skilled questioners are not good at detecting deceit simply on the basis of a suspect’s verbal and non-verbal cues (Ekman & O’Sullivan, 1991). Vrij (2000) suggested that most liars are caught because it becomes too difficult to continue to lie, and they have not made sufficient preparation to avoid detection. Vrij lists seven qualities that make a good liar:
1. having a well prepared story;
2. being original in what is said;
3. thinking quickly when the need arises;
4. eloquence in storytelling;
5. having a good memory for what has been said previously;
6. not experiencing emotions such as fear or guilt while lying; and
7. good acting ability.
If verbal and non-verbal cues are hard to read, how does an investigator catch out an individual who possesses all the attributes listed above? One approach is a highly structured analysis of verbal content, known as Statement Validity Assessment (SVA). Originally developed as a clinical tool for analysing children’s statements in cases of sexual abuse (Undeutsch, 1982), SVA consists of three elements:
1. A statement is taken in a structured interview.
2. The content of the statement is judged by the forensic psychologist in a criterion-based content analysis (CBCA). These content criteria are concerned with the general characteristics of the statement (such as whether it has a logical structure), the specific contents of the statement (such as descriptions of events and people), motivation-related content (such as admission of a lack of memory) and offencespecific elements (concerning the fine details of the offence).
3. The CBCA is necessarily subjective, and needs to be evaluated against a standard set of questions set in the ‘validity checklist’ (Raskin & Esplin, 1991). This checklist raises questions about the conclusions drawn from the analysis.
In other words, the content analysis itself is put to the test by systematic consideration of interviewee characteristics.
The interviewee’s psychological and motivational characteristics, the characteristics of the interview and a ‘reality check’ against other forensic evidence are all examined. It is clear that SVA represents an attempt to bring order and rigour to the essentially subjective matter of judging the veracity and reliability of an interviewee’s statement. However, in a review of the substantial evaluative literature with regard to SVA, Vrij (2000) has expressed several reservations about the technique and highlighted areas where questions remain. He concludes that ‘SVA evaluators appear to be able to detect truths and lies more accurately than would be expected by chance’ (p. 153). In other words, while not a perfect technique, SVA does help improve accuracy beyond guesswork and inaccurate beliefs about how to judge accuracy.
OFFENDER PROFILING
If ever a topic generated a great deal of heat and rather less light, offender profiling [offender profiling constructing a picture of an offender’s characteristics from their modus operandi together with the clues left at the crime scene] would be high on the list of most forensic psychologists.
But as our knowledge base increases, it is likely that the technique will become increasingly sophisticated (Ainsworth, 2001; Jackson & Bekerian, 1997). Wrightsman (2001) distinguishes between profiling historical and political figures, profiling likely criminals from crime scene characteristics, and profiling the common characteristics of known offenders. Turvey (2000) draws the distinction between inductive and deductive methods of profiling. Inductive methods rely on the expert skills and knowledge of the profiler – a method often referred to as ‘clinical’ in style. By contrast, deductive methods rely on forensic evidence, such as crime scene characteristics and offencerelated empirical data – an approach often referred to as ‘statistical’.
Profiling historical and political figures Attempts have been made to construct psychological profiles of historical figures (from Jack the Ripper to Adolf Hitler) by systematically gathering and organizing information in an effort to understand their motives and behaviour. Experts will undoubtedly have constructed psychological profiles of Saddam Hussain in order to try to predict his behaviour during the 2003 conflict in Iraq. These types of profile typically rely on specialist knowledge (e.g. military, historical).
Profiling criminals from the crime scene Way back in the late 1880s, forensic pathologists were trying to link series of crimes by the similarity of crime scene characteristics, such as the nature of a victim’s wounds. More recently, the American Federal Bureau of Investigation (FBI) pioneered an investigative system based on central features (such as the details of a crime scene and forensic evidence) in order to construct a profile of the psychological and behavioural characteristics of the criminal (Douglas et al., 1986). While forensic evidence can yield many clues, the starting point for the FBI was to use the crime scene to construct a picture of the type of person who committed the offence. This approach yielded various classifications of types of offender associated with their psychological characteristics.
For example, a much used distinction (mainly concerned with serious offenders such as murderers or rapists), incorporated within the FBI framework, is that between ‘organized’ and ‘disorganized’ offenders (Ressler, Burgess & Douglas, 1988). An organized offender will plan the offence, be careful not to leave evidence, and target the victim. The disorganized offender will seemingly offend at random, use a weapon that is discarded near to the scene of the crime, and make few attempts to hide evidence or potential clues. In terms of psychological characteristics, the organized offender is seen as intelligent and socially adjusted, although this apparent normality can mask a psychopathic personality. According to this framework, the disorganized offender is said to be less intelligent and socially isolated, may have mental health problems, and is likely to offend when in a state of panic. The obvious criticisms of such distinctions (and the FBI approach more generally) is that they are inductive, highly subjective and lacking in robust empirical validation. Profiling common characteristics of known offenders.
The third approach to profiling is to look to empirical data, rather than an expert’s opinion, to construct profiles. This approach emphasizes the rigorous gathering of data about the crime from multiple sources (such as geographical location and victim statements), the application of complex statistical analyses to databases of crime scene details (and other forensic evidence), and attempts to build a profile of the offender with theoretical integrity. Adopting this approach, Canter and Heritage (1990) analysed data from over 60 cases of sexual assault and were able to identify over 30 offence characteristics, such as level of violence, use of a weapon, type of assault and use of threats. Statistical analyses were used to search for relationships and patterns between the factors, and to build up characteristic profiles of types of sexual assault. This and other similar studies provide preliminary support for the central premises of offender profiling based on the common characteristics of known offenders.
No comments:
Post a Comment