INVESTIGATIVE PSYCHOLOGY

Introduction

The domain of investigative psychology covers all aspects of psychology that are relevant to the conduct of criminal or civil investigations. Its focus is on the ways in which criminal activities may be examined and understood in order for the detection of crime to be effective and legal proceedings to be appropriate. As such, investigative psychology is concerned with psychological input to the full range of issues that relate to the management, investigation and prosecution of crime.
As Canter made clear, when he first labeled and introduced the term ‘investigative psychology’, its constituents can be derived from consideration of the sequence of activities that constitute the investigative process, which runs from the point at which a crime is committed through to the bringing of a case to court. This makes it apparent that detectives and others involved in investigations are decision-makers.
They have to identify the possibilities for action on the basis of the information they can obtain. For example, when a burglary is committed they may seek to match fingerprints found at the crime scene with those of known suspects. This is a relatively straightforward process of making inferences about the likely culprit from the information drawn from the fingerprint. The action of arresting and questioning the suspect follows from this inference.
However, in many cases the investigative process is not so straightforward. Detectives may not have such clear-cut information but, for example, suspect that the style of the burglary is typical of one of a number of people they have arrested in the past. Or, in an even more complex example, such as a murder, they may infer from the disorder at the crime scene that the offender was a burglar disturbed in the act. These inferences will either lead them on to seek other information or to select from a possible range of actions, including the arrest and charging of a likely suspect.
Investigative decision-making thus involves the identification and selection of options, such as possible suspects or possible lines of inquiry, that will lead to the eventual narrowing down of the search process.
In order to generate possibilities and select from them, detectives and other investigators must draw on some understanding of the actions of the offenders) involved in the offence they are investigating. They must have some idea of typical ways in which offenders behave that will enable them to make sense of the information obtained. Throughout this process they must amass the appropriate evidence to identify the perpetrator and prove their case in court.
It follows that three processes are always present in any investigation that can be improved by psychological study. First, the collection and evaluation of information derived from accounts of the crime. These accounts may include photographs or other recordings derived from the crime scene. There may also be records of other transactions, such as bills paid or telephone calls made. Often there will be witnesses to the crime or there will be results of the crime available for examination. There will be transcripts of interviews or reports from various experts. Further, there will be information in police and other records that may be drawn upon to provide indications for action. Once suspects are elicited there is further potential information about them, either directly from interviews with them, or indirectly through reports from others. In addition there may be information from various experts that has to be understood and may lead to actions. The major task of a police investigation is, therefore, typically to collect, assess and utilize a great variety of sources of information that provide accounts of crime. This is a task that can benefit considerably from the scientific study of human memory processes and other psychological studies of the reliability and validity of reports and their assessment.
The second set of tasks is the making of decisions and the related actions that will move towards the arrest and conviction of the perpetrator. There is remarkably little study of exactly what decisions are made during an investigation, or how those decisions are made. Yet there is clearly a limited range of actions available to police officers, constrained by the legal system within which they operate. From many studies of human decision-making in other contexts, it is also apparent that there are likely to be many heuristic biases and other inefficiencies in the decision-making process. Awareness of these can lead to effective ways of overcoming them.
In order for decisions to be derived from the information available, inferences have to be made about the import of that information. The third set of tasks therefore derives from developing a basis for those inferences at the heart of police investigations. These inferences derive from an understanding of criminal behavior. For appropriate conclusions to be drawn from the accounts available of the crime, it is necessary to have, at least implicitly, models of how various offenders act. These models allow the accounts of crime to be processed in order to generate possibilities for action. This process of model-building and testing is, in effect, a scientific, psychological development of the informal, anecdote-based process often referred to as ‘offender profiling’ or ‘criminal profiling’.
A simple framework for these three sets of tasks that gives rise to the field of Investigative Psychology is shown in Fig. 1. More formally then, investigative psychology is the systematic, scientific study of:
• investigative information, its retrieval, evaluation and utilization;
• police actions and decisions, their improvement and support;
• the inferences that can be made about criminal activity, its development, differentiation and prediction;with the objective of improving criminal and civil investigations. More detailed information about each of these three strands of investigative psychology now follow.


Information Retrieval

There are many psychological questions raised in relation to the retrieval of information during a police investigation, and various studies have led to the development of procedures to improve the information collected during an investigation.

Detail

One of the most important aspects of the information obtained during an investigation is that it should have as much relevant detail as possible. Psychologists have therefore helped to develop processes, especially for police interviews, that maximize the information obtained. In doing this, the perspective is taken that there are two issues that need to be as effective as possible. One is based on the assumption that the respondent in an interview is essentially trying to remember what occurred. Therefore anything that can help the memory process should be of value.
Investigation cycle giving rise to the field of investigative psychology.
Figure 1 Investigation cycle giving rise to the field of investigative psychology.
The second issue is the relationship between the interviewer and the interviewee. If this relationship can be as supportive and helpful as possible, then more effective information is likely to be obtained.
Out of these considerations, guidelines for interviews have been developed. The best known of these is referred to as the ‘cognitive interview’ developed by Fisher and Geiselman. This is based on the assumption that memory is an active reconstructive process rather than a relatively passive act of recall. It draws on the well-established finding that recognition of information is much easier than its recall. Therefore any procedure that can help the interviewee to recreate the events in his or her own mind will be of value. This includes encouraging respondents to describe the events as they remember them rather than in strict response to particular questions in a given sequence. Reinstating the circumstances of the offence whenever possible, by returning to the scene or exploring details like sounds and smells, also accords with an understanding of the psychological processes by which memories are reconstructed. Attempts to consider the events from a variety of different perspectives are also considered valuable.
Investigative hypnosis has also been used to improve recall of information. In many respects hypnosis can be seen as a more intensive form of cognitive interview in which the respondent is helped to relax and concentrate. There are certainly many anecdotal accounts of its effectiveness. However, the possibilities of leading the respondent to offer information that may be suggested by the interviewing hypnotist are considered much greater than for the interviewer in a cognitive interview. Many jurisdictions therefore have very close controls over the ways in which hypnotic interviews can be conducted.

Accuracy

A number of studies have shown that the cognitive interview generates significantly more detailed information than conventional police interviews. Some studies show that the information obtained is more accurate, and also that more relevant information is obtained; but it is remarkably difficult to measure relevance or accuracy precisely, so the full value of the cognitive interview is likely to vary considerably between situations.
Attempts have also been made to use similar psychological processes to improve the recollection of faces and other details. This has proved less successful, in part because human recall of faces is so poor. Psychologists have therefore been involved in a variety of studies of how faces are reconstructed from memory and the procedures that can facilitate this. This has led to developments beyond the traditional ‘photo-fit’ approach. But the training involved in the use of these new systems, and their heavy reliance on effective interviewing, has meant that they have not had the uptake that would have been expected from the scientific findings.
Psychological research has also contributed considerably to the improvement in the validity of the traditional ‘identity parade’. Various procedures have been introduced by many police forces around the world to ensure that the recognition task set for the witness is appropriate and not open to bias. In particular, these take account of the need to protect the suspect against the possibility of the witness’s memory being modified by experiences subsequent to the crime, such as meeting the suspect in other circumstances.

Vulnerable interviewees

A number of witnesses may be regarded as vulnerable because of their age, emotional state or intellectual ability. Such witnesses may be particularly open to suggestion or may be made especially anxious or confused by the interview process. Special interview procedures have therefore been developed for interviewing such people. The procedures pay particular attention to the relationship established between the interviewee and the interviewer and the need to phrase questions and facilitate answers in ways that make sense to the respondent.

False confessions

Psychologists, in particular Gudjonsson and Mac-Keith, have drawn attention to the possibility that some individuals may confess to crimes they have not committed. These ‘false confessions’ may be a consequence of characteristics similar to those that make witnesses vulnerable, such as emotional state and intellectual ability, making the suspect more willing to accept suggestions from the interviewer. Gudjons-sen has developed a measure of a person’s ‘suggestibility’ that has been drawn on by the courts to support claims of a false confession. These may also be a product of cultural processes rather than aspects of personality, in which, for example, groups from certain ethnic minorities may deem it essential to agree with whatever a person in authority, such as a police officer, says to them. Investigative psychologists have also considered the ways in which false confessions may be produced in response to various forms of psychological or physical coercion.

Validity

In many circumstances investigators wish to assess the validity of information from witnesses because they consider allegations may be false. If there are no objective criteria for doing this they may use one of a number of validity assessment techniques. Most of these techniques are based on the assumption that honest accounts have identifiable characteristics that are different from fabricated accounts. The most frequently used approach to statement validation is that developed by Undeutsch, known as Statement Validity Assessment, which draws upon detailed analysis of the content of a statement, referred to as Criteria Based Content Analysis. This procedure has been widely used to evaluate allegations of abuse made by children, especially in Germany, where it originated. Attempts have also been made to extend its application to statements from other groups of witnesses, with less success.

Authorship

A subset of validity questions relate to whether the words attributed to a particular author are actually the words of that person or not. This may occur, for example, when a suspect denies that he or she made the statement attributed, or in other cases of forgery or fraud. To deal with these questioned utterances, there have been a variety of attempts to use techniques based on the quantitative examination of language. These approaches are sometimes put under the general heading of ‘stylistics’, or forensic linguistics, or more generally forensic psycholinguistics. Yet although much is claimed for these procedures by their protagonists, the systematic research into them rarely finds any evidence to support even the mildest claims. Advances in computing techniques may change this.
These procedures are not to be confused with ‘graphology’, which claims to be able to provide accounts of the personality of an author from the style of his or her handwriting. There is no consistent scientific evidence for these claims.

Detecting deception

When the suspect is the source of the information, additional factors are also important beyond those of memory retrieval. These often relate to the need to determine if a person is attempting to deceive the interviewer. Thus, although there are many objective, conventional police strategies for detecting deception, most obviously determining if the known facts contradict the suspect’s claims, there are a number of situations in which some knowledge of behavioral and psycholinguistic cues to deception would be very helpful. A number of researchers, most notably Ekman, have claimed that such cues are available, but others are more skeptical as to the possibility of any generally available indexes of deception from the actions or words of the suspect during a police interview.
There is much more evidence to indicate that, for many people, there are psychophysiological responses that may be indicators of false statements. The procedure for examining these responses is often referred to as a polygraph or ‘lie detector’. In essence, this procedure records changes in the autonomic arousal system, i.e. emotional response. Such responses occur whenever a person perceives an emotionally significant stimulus. The most well-established indicator is when the respondent is asked to consider information that only the perpetrator would be aware of, known as the ‘guilty knowledge’ test. A more controversial procedure is to ask ‘control questions’ that many people would find emotionally significant, in order to determine if the questions elicit responses that can be distinguished from those relating directly to the crime. However, in both these applications of psychophysiological measures the most important element is the very careful interview procedure before measurements are made and during the process. In general, the technique is more productive in supporting a claim of innocence than in providing proof of guilt. For this reason many jurisdictions do not allow ‘lie detector’ results to be presented as evidence in court.

False allegations

In recent years there has been growing concern about the various conditions under which people will falsely allege they have suffered at the hands of others. Often, but not always, this is an allegation of sexual abuse or harassment. The various procedures for detecting deception may be relevant in these cases but, because the complainant is not a suspect, the more intrusive processes of lie detection are rarely, if ever, utilized. Instead, there have been attempts to indicate the circumstances under which such false allegations are made and use those as guidelines for more intensive examination of the circumstances; however, the validity of these procedures is still highly questionable.

Investigative Decision-Making

The main challenge to investigators is to make important decisions in often ambiguous and sometimes dangerous circumstances. The events surrounding the decisions are likely to carry a great emotional charge and there may be other political and organizational stresses that also make objective judgments very difficult. A lot of information, much of which may be of unknown reliability, needs to be amassed and digested. These are the conditions under which various biases in investigators’ thought processes are likely to occur, with consequent inadequacies in the decisions made and the subsequent actions. Recognition of the potential for these problems can lead to the development of procedures to reduce their likelihood. The challenges of police and other investigations may also be reduced by the development of decision support tools that reduce the complexity of the information that needs to be understood and assists in the derivation of appropriate inferences from the material that is available to the police.
The decision support tools that are emerging for use by police investigators each draw on particular perspectives on the nature of the problem.

Visualization

Some support tools are based on the fact that human beings can often see patterns between associations, and within activities, if they can be presented in some form of visual summary. Bar charts of frequencies are one common example of this, but commercially available software will chart networks of contacts and other sequences of associations or actions.
While these tools can be productive in summarizing a great deal of information and, in association with databases, can improve the search for and access to crucial information, they are very dependent on the skills of the particular user, often referred to in police forces as a crime analyst. In the wrong hands these systems can imply a behavioral pattern through the strong visual impact that a diagram produced, when in fact the diagram is a biased emphasis of some peripheral aspect of the criminal behavior being investigated.

Description

A further level of support to decisions can be made by identifying the salient characteristics of the offences and offenders and by producing summary accounts of them. One widespread application of this use is in the production of maps that indicate the locations where there are high frequencies of crimes, sometimes called criminal hot-spots. In these cases the salient characteristics are simply where the crimes have occurred, and the description consists of some summary or averaging of the crimes over an area in order to indicate where its geographical focus might be. All description requires some selection, distillation or averaging of information, and when that is done validly the description is helpful.

Analysis

A further level of assistance to police decision-makers can be given by carrying out some form of analysis on the crime material, typically looking for patterns of co-occurrences or discriminating nonoccurrence. An example of the former would be the recognition that certain acts of vandalism occur shortly after the end of the school day near to schools. Knowledge from descriptive analyses of the age and backgrounds of offenders prosecuted for vandalism and the geographical hot-spot information could be combined to target possible culprits and introduce other forms of crime reduction. More advanced analysis of the cooccurrence of criminal behaviors could also be used for classifying offenders and generating different investigative strategies for the different forms of offender.

Inference

When clear relationships can be established between different aspects of crimes that are of investigative interest, inferences can be made from one to the other. For example, an understanding of the relationships between where an offender offends and where he or she lives can be used to infer residential location from knowledge of offence location. The use of inference for decision support activities is at the core of the area popularly known as ‘offender profiling’, which will be dealt with in more detail below.

Appropriate Inferences

The traditional approach police investigators take to making inferences is the one that has always been characterized in crime fiction as ‘deduction’. This is the process of reasoning from commonly known principles. For example, if a walking stick is found with strong, large teeth marks in it, then it may be reasoned that these were most likely caused by a large dog that carried the stick (as Sherlock Holmes reasons in The Hound of the Baskervilles). A more subtle piece of reasoning may come from the knowledge that an offender had long nails on the right hand but short ones on the left. This is a pattern favored by some guitarists, and so it may be assumed that the offender was a serious guitar player.
However, as attractive as such deductions are in fiction, they are a very poor basis for developing robust inferences in real-life crime. They are vulnerable to the knowledge and reasoning ability of the deducers and the particular features that they notice. Even more importantly, they may be worthless. It turns out that many trades give rise to people having longer nails on one hand than the other, so the inference of a guitarist could be very misleading. A dog may have bitten a walking stick in situations other than when the stick was being carried, and so the dog may not be directly associated with the owner of the walking stick.
In order to determine what the salient aspects of an offence are, and how they may be related validly to useful investigative inferences, it is necessary to collect information across a range of cases and to test hypotheses about the actual co-occurrence of various features. This is the process of ‘inductive reasoning’, which is at the heart of empirical science. Investigative psychologists have consequently been active in conducting a wide range of empirical studies aimed at providing objective bases for investigative inferences. These are studies that have been characterized by Canter as attempts to solve the set of equations that link the Actions that occur during the offence, including when and where it happens and to whom, to the Characteristics of the offender, including the offender’s criminal history, background and relationships to others. These have become known as the A — C equations, or the ‘profiling equations’.
Studies of these equations have given rise to the identification of a number of aspects of criminal behavior that are crucial to any models of inference for effective use in investigations.
One important aspect of these models is that the variables on which they can draw are limited to those of use to police investigations. This implies that the A variables are restricted to those known prior to any suspect being identified. The C variables are limited to those on which the police can act. So, offender’s personality characteristics, detailed measures of intelligence, attitudes and fantasies are all of less utility than information about where the person might be living, and his or her criminal history or domestic circumstances.

Consistency

In order to generate some form of A – C equation it is essential that the two sides of the equation are stable enough for a relationship to be established. Therefore much investigative psychology research is devoted to establishing what the salient features of an offender’s crimes are, and what it is within those features that is consistent enough to form the basis of their characteristics.
It is from these studies that classification schemes are emerging considering, for example, relevant variations between serial killers and between stalkers. What is emerging from these studies is that styles of interpersonal transaction may well be consistent enough for some inferential models to be built. A distinct subset of offenders has also been identified: they have consistent relationships between their residence and where they commit their crimes, also allowing geographical inference models to be developed.

Differentiation

Although an offender’s consistency is one of the starting points for empirically based models of investigative inference, in order to use these models operationally it is also necessary to have some indication of how offenders can be distinguished from each other. If every offender were consistent in the same way, then the A – C equations would provide characteristics that were the same for every offender. In part, this question reflects a debate within criminology about whether offenders are typically specialist or versatile in their patterns of offending. Research tackling this problem has tended to support the contention that the majority of chronic criminals will commit a wide range of crimes and thus cannot be considered specialist, thereby making differentiating inferences extremely difficult. However, current research is suggesting that it is possible to model offender’s behavior in terms of both those aspects that they share with most other criminals and those aspects that are more characteristic of them. It is these rarer, distinguishing features that may provide a productive basis for differentiating inferences.

Development and change

A further complication to establishing the A – C equations is that the way a person commits a crime, and indeed the characteristics of a person, will change over time, even if there is a background of consistencies. However, if the basis of these changes can be understood, they can be used to enhance the inference process. In essence, the following five forms of change have been identified.
Responsiveness One important reason for differences between a criminal’s actions on two different occasions may be his or her reaction to the different circumstances faced. By an understanding of these circumstances and how the offender has responded to them, some inferences about his or her interpersonal style or situational responsiveness may be made, which can have investigative implications.
Maturation This is the, essentially biological, process of change in a person’s physiology with age. Knowledge of what is typical of people at certain ages, such as sexual activity, can thus be used to form a view as to the maturity of the person committing the crimes and to the basis for longer term variations in an individual’s criminal activity.
Development The unfolding psychological mechanisms that come with age provide a basis for change in cognitive and emotional processes. One reflection of this is an increase in expertise in doing a particular task. Evidence of such expertise in a crime can thus be used to help make inferences about the stages in a criminal’s development that he or she has reached, and indeed to indicate the way the crimes might change in the future.
Learning Most offenders will learn from their experiences. They will therefore be expected to alter their actions in the light of the consequences of previous actions. An inferential implication of this is that it may be possible to link crimes to a common offender by understanding the logic of how behavior has changed from one offence to the next.
Careers The most general forms of change that may be expected from criminals is one that may be seen as having an analogy to a legitimate career. This would imply stages such as apprenticeship, middle management, leadership and retirement. Unfortunately, the criminology literature often uses the term ‘criminal career’ simply to mean the sequence of crimes a person has committed. It is also sometimes confused with the idea of a ‘career criminal’ – someone who makes a living entirely out of crime. As a consequence, much less is understood about the utility of the career analogy for criminals than might be expected. There are some indications that the more serious crimes are committed by people who have a history of less serious crimes and that, as a consequence, the more serious a crime, the older an offender is likely to be. But commonly held assumptions – such that serious sexual offences are presaged by less serious ones – do not have much supporting empirical evidence.
5. Which suspects are most likely to have committed the crime in question?
6. What sense can be made of the offense that will help to organize the legal case?
The complexity of deriving inferences to answer these operational questions is considerable. It requires managing issues of consistency and differentiation, together with development and change, across a range of aspects of crimes that will vary in degree of specialism. Therefore a number of investigative psychologists have collaborated with police forces around the world to develop computer-based decision support systems drawing on the ideas indicated above. These inductively developed systems are likely to replace rapidly the outmoded methods of police deduction.

The Future

Investigative psychology provides a holistic perspective on the investigation of crime, showing that all aspects of the detective’s work is open to scientific psychological examination. It is helping police forces to recognize that they need to build psychological expertise into their modern computing capability, rather than just bringing an expert in when an investigation has reached a particularly difficult stage. They are learning to answer the question, ‘At which point in an investigation should a psychologist be brought in?’, with ‘Before the crime is committed!’.

Next post:

Previous post: