<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>In human intelligence, a verbal statement from a source is seldom 100% true or false, and not very often is the source a total liar or a truth teller. From this standing point, a simple dichotomy of a liar or a truth teller might not offer an adequate diagnostic value for the purposes of human intelligence. A more diagnostic approach would be to assess which parts of the predominantly truthful verbal statement are likely to be true and which parts are assessed to be doubtful. In addition, the use of two parallel methods to detect deceit should improve the diagnostic value of the results. A pilot study in laboratory conditions (<italic>n</italic> = 8, yielding 190 assessment points) utilising an applied mock crime scenario was conducted. Correlation calculations showed that a dual-method approach slightly improved the within-statement truth accuracy, and it was achieved mainly by decreasing the number of false positives. As the truth accuracy was increased, the lie accuracy within the test group slightly decreased. The results confirmed that by applying parallel orienting response (EDA) and cognitive load (speech-related indices)-based assessment methods, it is possible to detect embedded lies successfully in an information-gathering interview setup.</p>
</abstract>

In human intelligence, a verbal statement from a source is seldom 100% true or false, and not very often is the source a total liar or a truth teller. From this standing point, a simple dichotomy of a liar or a truth teller might not offer an adequate diagnostic value for the purposes of human intelligence. A more diagnostic approach would be to assess which parts of the predominantly truthful verbal statement are likely to be true and which parts are assessed to be doubtful. In addition, the use of two parallel methods to detect deceit should improve the diagnostic value of the results. A pilot study in laboratory conditions (n = 8, yielding 190 assessment points) utilising an applied mock crime scenario was conducted. Correlation calculations showed that a dual-method approach slightly improved the within-statement truth accuracy, and it was achieved mainly by decreasing the number of false positives. As the truth accuracy was increased, the lie accuracy within the test group slightly decreased. The results confirmed that by applying parallel orienting response (EDA) and cognitive load (speech-related indices)-based assessment methods, it is possible to detect embedded lies successfully in an information-gathering interview setup.

Detecting deceit within a predominantly true statement using two parallel assessment methods: A pilot study

Journal of Military Studies

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

In human intelligence, a verbal statement from a source is seldom 100% true or false, and not very often is the source a total liar or a truth teller. From...

	EDA			Speech			EDA & Speech
	rs	p*	N	rs	p*	N	rs	p*	N
Truthtellers
Test Subject 3	0.27	0.131	33	0.21	0.250	33	0.48	0.005	33
Test Subject 5	1.00	0.000	13	-1.00	0.000	13	1.00	0.000	13
Test Subject 7	0.34	0.163	18	0.24	0.330	18	0.54	0.020	18
Test Subject 9	0.29	0.096	34	0.13	0.469	34	0.31	0.070	34
Test Subject 10	0.67	0.016	12	0.21	0.516	12	0.67	0.016	12
	rs	p*	N	rs	p*	N	rs	p*	N
Deceivers
Test Subject 4	0.56	0.006	23	0.17	0.435	23	0.46	0.026	23
Test Subject 6	0.49	0.002	37	0.49	0.002	37	0.40	0.014	37
Test Subject 8	0.62	0.004	20	0.15	0.518	20	0.49	0.027	20

	Truth			Deceit			Overall
	EDA	Speech	EDA & Speech	EDA	Speech	EDA & Speech	EDA	Speech	EDA & Speech
Truthtellers
Test Subject 3	70%	58%	88%	na	na	na	na	na	na
Test Subject 5	100%	0%	100%	na	na	na	na	na	na
Test Subject 7	67%	50%	83%	na	na	na	na	na	na
Test Subject 9	92%	33%	92%	na	na	na	na	na	na
Test Subject 10	74%	35%	76%	na	na	na	na	na	na
	EDA	Speech	EDA & Speech	EDA	Speech	EDA & Speech	EDA	Speech	EDA & Speech
Deceivers
Test Subject 4	92%	25%	92%	73%	91%	64%	82%	58%	78%
Test Subject 6	88%	76%	92%	58%	75%	42%	73%	76%	67%
Test Subject 8	73%	36%	82%	89%	78%	67%	81%	57%	74%

Detecting deceit within a predominantly true statement using two parallel assessment methods: A pilot study

Article Category: Original Study

Publicado en línea: 11 mar 2024

Páginas: 61 - 74

Recibido: 05 ene 2023

Aceptado: 14 sept 2023

DOI: https://doi.org/10.2478/jms-2023-0005

Palabras clave
human intelligence, deception detection, embedded lies, within-statement lie detection, parallel methods, information-gathering interview

© 2023 Marko Uotinen et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Fig. 1:

Fig. 2:

Correlation Between the Assessement and the Actual Status of the Statement Using EDA Indicators, Speech Indicators and Combined EDA & Speech Indicators.

Accuracy Rates for Deception and Truth Using EDA Indicators, Speech Indicators and Combined EDA & Speech Indicators.

Detecting deceit within a predominantly true statement using two parallel assessment methods: A pilot study

Article Category: Original Study

Publicado en línea: 11 mar 2024

Páginas: 61 - 74

Recibido: 05 ene 2023

Aceptado: 14 sept 2023

DOI: https://doi.org/10.2478/jms-2023-0005

Palabras clavehuman intelligence, deception detection, embedded lies, within-statement lie detection, parallel methods, information-gathering interview

© 2023 Marko Uotinen et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Fig. 1:

Fig. 2:

Correlation Between the Assessement and the Actual Status of the Statement Using EDA Indicators, Speech Indicators and Combined EDA & Speech Indicators.

Accuracy Rates for Deception and Truth Using EDA Indicators, Speech Indicators and Combined EDA & Speech Indicators.

Palabras clave
human intelligence, deception detection, embedded lies, within-statement lie detection, parallel methods, information-gathering interview