Voice stress analysis (VSA) and computer voice stress analysis (CVSA) are collectively a pseudoscientific technology that aims to infer deception from stress measured in the voice. The CVSA records the human voice using a microphone, and the technology is based on the tenet that the non-verbal, low-frequency content of the voice conveys information about the physiological and psychological state of the speaker. Typically utilized in investigative settings, the technology aims to differentiate between stressed and non-stressed outputs in response to stimuli (e.g., questions posed), with high stress seen as an indication of deception.
The use of voice stress analysis (VSA) for the detection of deception is controversial. Discussions about the application of VSA have focused on whether this technology can indeed reliably detect stress, and, if so, whether deception can be inferred from this stress. Critics have argued that—even if stress could reliably be measured from the voice—this would be highly similar to measuring stress with the polygraph, for example, and that all critiques centered on polygraph testing apply to VSA as well. A 2002 review of the state of the art conducted for the United States Department of Justice found several technical challenges to the technology, including the same problem of determining deception. When reviewing the literature on the effectiveness of VSA in 2003, the National Research Council concluded, “Overall, this research and the few controlled tests conducted over the past decade offer little or no scientific basis for the use of the computer voice stress analyzer or similar voice measurement instruments”.1:168 A 2013 paper published in Proceedings of Meetings on Acoustics reviewed the "scientific implausibility" of its principles and "ungrounded claims of the aggressive propaganda from sellers of voice stress analysis gadgets".
Confession made following a voice stress examination was allowed to be used as evidence in a case in Wisconsin in 2014. In the case of the murder of 12-year-old Stephanie, Crowe confessions were made while three suspects were undergoing VSA which were later found to be false by a judge; the manufacturer of the VSA equipment later settled a lawsuit that alleged that it was liable for the harm the three suspects suffered. In a similar case, Donovan Allen falsely confessed to killing his mother after failing a VSA test. He was acquitted 15 years later based on exonerating DNA evidence. George Zimmerman was given a VSA after he fatally shot Florida teenager Trayvon Martin in 2012.
The Mark II Voice Stress Analyzer
The Mark II Voice Stress Analyzer is the most advanced, accurate and simple instrument of its kind. It allows the user to automatically detect, measure and analyze the exact degree of psychological stress in a word or phrase spoken by anyone. It can be used in person, through recorded messages or by telephone, or in radio communications.
When a subject tries to cheat, or has emotional difficulty with certain questions, he will experience psychological distress. Any misrepresentation or psychological stress induced by stimuli causes a lack of synchronization between the control of the voice of the brain and the mechanism of the voice in the throat. This stress connection produces an inaudible roughness in the voice called tremolo. Tremolos related to stress occurs uniformly in all human allocations, independent of language or sex.
With this in mind, the Mark II applications are incredible. Traders or entrepreneurs concerned with critical negotiations or personal evaluations, law enforcement agencies and intelligence organizations will find the Mark II very useful. It has new applications for trainings where stress and emotional reactions are a priority, and can be used for medical-psychiatric diagnosis.
The unit is equipped with digital or printed display, which helps in the interpretation and analysis of the final results. The Mark II is totally portable, contained in an attractive briefcase.
The nonverbal content of speech carries information about the physiological and psychological condition of the speaker. Psychological stress is a pathological element of this condition, of which the cause is accepted to be “workload”. Objective, quantifiable correlates of stress are searched for by means of measuring the acoustic modifications of the voice brought about by workload. Different voice features from the speech signal to be influenced by stress are loudness, fundamental frequency, jitter, zero-crossing rate, speech rate and high-energy frequency ratio. To examine the effect of workload on speech production an experiment was designed. 108 native speakers of Dutch were recruited to participate in a stress test (Stroop test). The experiment and the analysis of the test results will be reported in this paper. 1 Introduction Although speech is a vocal activity of which much is verbal, there are a number of human vocalizations that are essentially non-linguistic. Nonverbal aspects of speech are intonation, voice quality, prosody, rhythm and pausing. These phenomena stand for a non-verbal signaling system, which intertwines with the verbal or linguistic system. The non-verbal content of the voice carries, among other things, information about the physiological and psychological state of the speaker. Human beings are able to identify different emotional states because these are characterized by clearly perceptible (non-verbal) behavior. Part of this non-verbal communication takes place via other modalities like body movements and facial expressions. The question that remains is how much of this information can be recovered from non-verbal vocalizations only. One of the most interesting research areas concerning non-verbal communication in relation to a person’s psychological state is the search for objective, quantifiable correlates of stress. In the past, this search focused primarily on physiological measures, but over the last years, a broader range of behaviors has been examined especially non-verbal behavior. The advantage being that stress indexes from non-verbal vocalizations can be obtained nonintrusively. From a practical point of view, this is critical in a situation in which co-operation for physiological measurement is precluded, for example in the case of negotiating with terrorists. However, even when co-operation is possible, the presence of monitoring devices needed for physiological measurement can be stressful and anxiety-arousing or simply not practical. Psychological stress is a pathological element of the physiological and psychological condition of the speaker, of which the cause is accepted to be “workload.” Objective, Petr Sojka, Ivan Kopecek, and Karel Pala (Eds.): TSD 2004, LNAI 3206, pp. 449–456, 2004. ˇ Springer-Verlag Berlin Heidelberg 2004 450 Leon J.M. Rothkrantz et al. quantifiable correlates of stress are searched for by means of measuring the acoustic modifications of the voice brought about by workload. These changes in the acoustic speech signal due to stress are mainly caused by the physiological changes that accompany the stress reaction. These changes also affect the organs of speech, such as the respiration and muscle tension (vocal cords) and therefore the speech signal. Hence, it should be possible to establish whether a person is stressed just by analyzing his voice. 2 Related Work Much work on stress analysis in real life situations concentrates on air-ground communication in aviation and space flight under dangerous conditions. In many of these studies, an increase of the fundamental frequency (F0) of the voice in situations of increasing danger is reported. Williams and Stevens also reported an increase in F0 range and abrupt fluctuations of F0 contour, with increasing stress. In a Russian study, the voices of astronauts are examined and changes in spectral energy distribution (spectral centroid moving to higher frequency) are reported. An increase of the energy of high-frequency components has also been reported by in a study involving pilot communication. Scherer et al. found depressive patients speak with higher F0 and a larger proportion of high-frequency components, just before the admission at a psychiatric hospital. Jones found increases in fundamental frequency and statistically significant decreases of the vocal jitter in recordings obtained from pilots training in a simulated AWACS environment. In many laboratory studies, stress is brought about by showing unpleasant or disgusting slides or films, or by placing the subject in situations that produce unpleasant emotions, such as stage fright. The degree of stress perceived will vary from person to person depending on the person's experience and arousability. Apart from these individual differences, some studies show an increase in intensity, increased fundamental frequency, a stronger concentration of energy above 500 Hz and an increase in speech rate. More recently, many experiments were conducted in which cognitive or achievement tasks were used to induce stress on a subject. When persons were subjected to a psychomotor task, the speaking fundamental frequency showed an increase when the task became more difficult. In addition, word duration increased during the task, but decreased again when the task became more complex. Brenner also found an increase in average amplitude when subjects were performing a tracking task. Table 1 summarizes the parameters that have been shown to be indicators of the vocal expression of emotion, emotional disturbance or stress. 3 Experimental Design To study the correspondence between human stress levels and speech production and to assess the relevance of the features listed in Table 1, an exploring experiment has been conducted. 108 native speakers of Dutch were subjected to several tasks that have been designed to place a cognitive workload on the subject. The cognitive workload is defined as the information processing load placed on the human operator while performing a particular task. This information processing load is considered to be correlated with the amount of attention that must be directed to a task. It is assumed that cognitive workload increases with the difficulty.