Detection of Deception

There are three ways of interpreting behaviour, guesstimating, approximating, and enquiring.

Guesstimating

Also referred to as mind reading, guesstimating assumes what behaviour means based on observation alone. Growing insecurity in western democracies has spawned a massive amount of research into detection of deception, seeking to identify tell tale macro and micro behaviours without input from the subjects under scrutiny. ?

?Although research has been helpful in identifying erroneous biases about behaviour, stand alone, non verbal behavioural research has only marginally improved our mind reading skills, even with the development of a multitude of remote sensing devices.?

The reasons are obvious. We cannot detect cognition, only the external manifestations of visceral activity associated with thinking. (Even if we connect to a brain we are only measuring blood flow to regions of the brain taxed with indistinguishable thoughts.)?

?By observing behaviour we are not trying to unmask deception - our stoic masks and purposeful composure are the deception, a measure of our civility and need for autonomy. What we are attempting to unmask are concealed emotions and there is no way of distinguishing stress from hurrying to catch a plane from stress caused by bad intentions.? Which perhaps explains why we are terrible at identifying liars based on non verbal behaviour. Our composure is the deception, and since composure is a universal feature - our social skin, the behaviour of liars is not dissimilar to truth tellers.

Approximating refers to the timing of non verbal behaviour in response to directed attention to words, pictures, sounds.? The meaning of expression can more confidently approximated only if the response clearly relates to the context.? For example in response to the question “Did you do it?” increased heart rate, shortened breaths, paling skin, higher pitch and lower volume with speech does not indicate guilt. It does indicate anxiety about the question and would lead one to assume the person may be withholding information relating to the question. Early research into incongruence or mixed messages between what is said and behaviour suggest that ‘leaks’, like polygraph, provide avenues for further enquiry? but are not infallible and present a host of problems controlling variables, which when manipulated allow for a nearly infinite number of outcomes. For example increased time with a subject improves accuracy in detection. As well the research does not take into account subjective perception of the meaning of the question, and how the risk is perceived by the subject which directly affects the robustness of the response.? Nevertheless, as flawed as this pseudo scientific research is, many machines requiring cooperation from the subject as well as remote sensors that do not require cooperation are proliferating with exaggerated claims of effectiveness.? As an example the technology, dubbed Future Attribute Screening Technology, or FAST, deploys a range of "innovative physiological and behavioural technologies" to pick up "indications of malintent or the intent or desire to cause harm", according to the DHS.

The technology, developed by the Human Factors division of Homeland Security's directorate for Science and Technology, would be used at border checkpoints, airports and special events that require security screening.

The system, called MALINTENT, uses a raft of "non-invasive" sensors and imagers to detect such factors remotely - subjects are not hooked up to anything. It also evaluates a person's facial expression to help to gauge whether they could be planning to commit an attack or crime.? Of course there is no way of detecting malintent because there are no distinctive behaviours of intention. FAST relies on questioning subjects where responses both verbal and non-verbal can be evaluated. The problem with this technology is the volume of traffic through airports - literally hundreds of millions of travellers means that Homelands Security continues to cherry pick based on profiling, with the technology rendered virtually useless.? As I have mentioned before a terrorist operative entering the US would have no knowledge of their mission until after they landed so there would be no anxiety due to ignorance.

Which brings me to enquiring - using emotional intelligence, and rapport, to encourage subjects to articulate the meaning of apparent stress. Let me share an example:?

An OH&S investigator is interviewing a worker at a job site following a fatal fall. Initially the worker is very nervous, highly stressed and seems hesitant to respond to questions. The investigator addresses the issue - “You seem very unsettled - is there something that concerns you that I might be able to help you with?” The worker explains that he is worried about losing his job if he gives a statement. As the interview moves along, the interviewer, who has now calibrated what stress looks like with this worker notices a recurrence of similar behaviour. The investigator again pauses to enquire about what is troubling the worker. This time the issue is not wanting to be a ‘rat’. The concern is acknowledged, the interview moves on until anxiety surfaces again. At this point the worker suggests that he shares responsibility for the fatality by an act of negligence. He was not setting an example by wearing a harness.

Note episodic stress in this case has multiple layers of meaning yet by appearance each concern has a similar face. If we try to interpret the meaning of expression at the outset,(mind reading), we would very likely be wrong and diminish rapport with misunderstanding and misinterpretation. We might be able to gauge how stress appears, we can’t always know what it means, even at times if it is a response to a critical issue question.

Tapping into subjective perception makes one vulnerable to being misled by the interviewee although this a minor risk compared to mind reading and approximating, and has the advantage of being a very compelling rapport maneuver.?