Gustafson, L. A., & Orne, M. T. Effects of heightened motivation on the detection of deception. Journal of Applied Psychology, 1963, 47, 408-411.



Massachusetts Mental Health Center and Harvard Medical School, Boston

1 of 5 cards was selected by each S and 2 minutes association to this card was required. GSR response to the selected card was compared to the responses for nonselected cards in 2 groups of Ss. 1 group was motivated to "deceive the operator and withhold responses." The other group was given no special instruction. The hypothesis that Ss who are motivated to deceive will more frequently produce disproportionately large skin resistance responses to critical items as opposed to noncritical items than will Ss who have not been so motivated was upheld. Ss who were motivated to deceive were more successfully detected. In addition detection took place at a much greater than chance level in the motivated group, while in the other group it occurred only at chance levels. The degree of autonomic response to significant stimuli appears to be a function of the motivational state of the S.

The apparent causal relationship between certain classes of verbal stimuli and physiological responses is the basis for the detection of deception by means of a polygraph. While variables which may increase or decrease the number of successful detections are often mentioned, these variables have not been manipulated experimentally.

It has been postulated repeatedly that the factor which produces the physiological response is not lying or guilt per se but rather something relating to the consequences of being detected (Burtt, 1921; Chappell, 1929; Marston, 1917). This has been formulated as the punishment, or threat of punishment, theory (Davis, 1961). According to this theory, the greater the consequences of being detected, the greater the physiological response will be to the critical items, and therefore the greater the chance of detection. However, the consequences of detection have not been treated as an independent variable. This would be one of the critical tests for the theory: If no differences were found when the consequences were varied, the theory would not be valid.

On the basis of the previous observations, it is hypothesized: Subjects who are motivated to deceive will more frequently produce disproportionately large skin resistance responses to critical items as opposed to noncritical items than will subjects who have not been so motivated. The frequency of detection among the motivated group is therefore likely to be greater.



Thirty-six male subjects between 18 and 25 years of age were recruited for a "paid psychological experiment" from the employment offices of four colleges in the Boston area. None of the subjects had previously participated in a study in deception. All subjects were randomly assigned to one of two experimental groups -- 18 to each group.


Subjects were run individually. Upon reporting, all subjects were given a very ambiguous idea of the nature of the experiment. They were told that the purpose of the study was to find out how normal subjects reacted physiologically to a series of numbers and letters, that it would be necessary to attach a number of recording electrodes to them, but that none of the electrodes would carry current to them. Electrodes for recording skin resistance (Wenger, Engle, & Clemens, 1957) and five other variables were then attached. (The remaining variables will be discussed in a later paper.) At this point a tape recording was played to half the subjects according to a random schedule, previously determined. This group was referred to as the tape group, while the remaining subjects were in the no-tape group. This recording contained the following information: (a) the experi-

1 The research in this study was supported in part by the Institute for Experimental Psychiatry and by the Mental Health Research Training Program, Harvard Medical School.

The authors wish to express their appreciation to Emily Carota Orne for her critical comments in the preparation of this manuscript.




ment was designed to see how well the subject could keep information away from the experimenter; (b) that this was extremely difficult to do, and that only people of superior intelligence and great emotional control were able to do this; (c) they were to try as hard as they could to beat the experimenter and the equipment; and (d) if they were successful, they would be paid an extra dollar.

All subjects then picked a card from a deck of five cards. All the cards in a deck were either all number or all letter cards, with a single character on each. Half the time the letter deck was used, and half the time the number deck. The numbers were between two and nine and the letters between B and I.

The cards were arranged so the subject could not see the face of the card until he had drawn. The experimenter could not see the face at any time. After the subject had memorized the card, he placed it face down on a stand beside him. In order to make the selected card more significant to the subject, he was instructed to write down on a piece of paper, in the case of a letter card, all the words he could think of beginning with that letter, or, in the case of a number card, all the expressions and titles he could think of containing that number. He was given 2 minutes for the task. The subject was then told to lie down and relax as much as possible, and that after about 5 minutes he would hear a series of numbers (or letters), including the number (or letter) he had removed from the deck. He was not to respond verbally to any of these. The experimenter left the room and began recording the physiological measures on an Offner Type R dynograph located in an adjacent room. At the end of 5 minutes he turned on a tape recording and one item of information was presented every 15 seconds. The first item presented was a dummy foil, while the next five were the same as the characters on the five cards. After all six had been presented, they were presented in a different order. This was repeated until each character had been presented five times.

As each character was reproduced by the tape recorder, the signal pen on the polygraph was activated and the letter or number was written on the record. At the conclusion of the tape, the experimenter returned to the subjects' room and did one of three things, according to a previously arranged, randomized order. To one third of the subjects in each group he told which card they had picked, to one third he deliberately misinformed them as to which card they had picked, and to the remaining third he said nothing concerning the card they had picked. (The reasons for this design will be discussed in a separate paper.)

The subject then picked a card from a second deck. If the first deck had been numbers, the second was letters and vice versa. The remainder of the trial was exactly the same as the first trial.

The difference in skin resistance between the level immediately prior to the stimulus and the lowest level reached within 4 seconds was used as the response measure for each stimulus. Readings were made to the nearest 500 ohms. The readings were all made by a person who did not know in which group the record belonged and did not know which was the chosen letter or number.

The largest mean response was used as the predictor of the card that the subject had chosen. The mean responses for each character were determined and these means were then ranked, the largest response receiving a rank of 1. The rank of the character chosen by the subject was then determined. If this rank was 1, it was considered a correct detection, while if it was more than 1, it was considered as not successful.


The ranks of the selected card for Trials I and II for all subjects are shown in Table 1 for the tape and no-tape groups, along with the number of correct predictions.

A comparison was made between the ranks of the selected character for the tape and no-tape condition on Trials I and II. The Mann-Whitney U for Trial I was 90.0 and for Trial II was 88.5. Both of these are significant at the .05 level (two-tailed).

While primary concern of this study as put forth in the introduction was to determine



whether difference in subject motivation would affect the magnitude of the response to a chosen card in relationship to the magnitude of the response to other cards, it is also of interest to see how successful detection itself was in the two conditions.

Fisher exact probability tests for Trials I and II for the two conditions indicated that there was a significant difference in the number of correct detections between the tape and no-tape conditions (see Table 1). It was decided to see if detection was occurring at a greater than chance frequency in both groups. A binomial test indicated that for both Trials I and II the tape group was detected at a significantly greater than chance frequency (p < .001 on both trials) while for the no-tape condition this was not the case (p > .10 on both trials).

The records of those subjects who were in the tape group appeared to show both larger and more frequent responses, not only to the correct character but to all the characters. These differences were significant at the .05 level (Mann-Whitney U; two-tailed).


The significant difference between the ranks of the selected character in the tape and no­tape conditions (for the first trial 1.44 and 2.36, and for the second trial 1.81 and 2.83, respectively) with corresponding differences in the relative response size to critical and noncritical items for both trials is supportive of the hypothesis put forward in the Introduction of this paper and the punishment theory of detection of deception. According to this theory the "person will give a large physiological response during lying because he anticipates serious conseqences if he fails to deceive [Davis, 1961, p. 163]." In the present experiment, while the subject is paid an extra dollar if not detected, probably the greatest consequence of being detected would be a loss of self-esteem. In the tape it is mentioned that the only persons who are able to deceive are those with superior intelligence and great emotional control, two qualities which most undergraduate students cherish. (The experimenter was careful to assure the subjects who had been detected that it had been difficult and that they had put up a good struggle.) Marston (1917) suggested that the factors which make detection possible are not directly due to the response of lying, but rather are due to an emotional reaction, probably of fear, surrounding the verbal response of lying. Burtt (1921) found that having other people present during the detection procedure increased the likelihood of successful detection. In a study by Chappell (1929), it was found that simply having the subject lie without any possibility of detection or punishment did not produce any marked responses. Further, Larson (1922) had noted that after a confession, the critical items no longer produced responses. Here again the stimulus no longer produces a response after the consequences of deception have been eliminated.

Our findings by no means eliminate alternate explanations of the events underlying the detection of deception. In this experiment the consequences of being detected are quite different for the motivated group not only during the actual recording session, but also during the period when they are memorizing the card and associating to it. In terms of a conditioned response theory, it would be assumed that during the period of making associations to the card, a greater response is probably produced for the tape group than for the no-tape group because of their increased involvement and this becomes conditioned to the selected character and produces a larger response during the test situation. A future experiment which would resolve this issue would be to change the consequences of deceiving during different parts of the experiment. Certain trials could be highly rewarded for deception while other unmotivated trials could be run for an innocuous reason, such as to "check the equipment." Any differences in the relative sizes of the responses would clearly be due to differences in the consequences of being detected and not due to differences in the responses conditioned to the selected card.

The significant differences between the number of successful detections (that is the number of times a critical item was assigned a rank of 1) for the tape and no-tape conditions uphold our hypothesis that the number of successful detections is increased as



motivation is increased. It also lends support to the consequences theory mentioned earlier.

It is of interest to see what ranks were assigned to the critical items in cases where the subject was not successfully detected. In the tape condition on Trial I, of the 6 individuals who were not detected, 4 were assigned a rank of 2 on the critical number. Even in cases where detection did not occur, the critical item produced an abnormally large response and the assignment of ranks was not random. However, in the no-tape group, of the 12 who were not assigned a rank of 1 on the critical item, only 3 were assigned a rank of 2 on the critical item.

Individuals in the tape group were not able to suppress the response to the critical item, though they were able to enhance their response to one or more noncritical items.

The fact that motivated subjects were detected far more readily than chance, supports the claims made for lie detection in actual life contexts where motivation would be maximal. On the other hand, the finding that without special motivation detection in the laboratory is difficult explains some of the skepticism toward laboratory studies of deception (Berrien, 1939). Clearly the situational variables play a crucial role in the responses of the autonomic nervous system.

As mentioned in the preceding section, the tape and no-tape groups appeared to be different, not only in the number of responses made to the selected card, but to all cards. While autonomic responses are usually considered to be more or less out of the area of experimental control, except by the manipulation of certain characteristics of the stimulus, such as the intensity or duration, here we find that by manipulating the role of the subject we have greatly altered his responsiveness to a stimulus which objectively remains unchanged. This relationship of the demands of the experiment to autonomic nervous system (ANS) activity is a factor that has not been considered in the discussions of response specificity and stimulus specificity. One can only speculate concerning the effect that different expectations of experimenters have on the nature of their subjects' responses. This relationship between the subject's role and ANS activity could be important for a theory of the etiology and treatment of psychosomatic disorders.


BERRIEN, F. K. A note on laboratory studies of deception. J. exp. Psychol., 1939, 24, 524-546.

BURTT, H. E. The inspiration-expiration ratio during truth and falsehood. J. exp. Psychol., 1921, 4, 1-23.

CHAPPELL, M. N. Blood pressure changes in deception. Arch. Psychol., 1929, 17, 5-39.

DAVIS, R. C. Physiological responses as a means of evaluating information. In A. D. Biderman & H. Zimmer (Eds.), The manipulation of human behavior. New York: Wiley, 1961. Pp. 142-168.

LARSON, J. A. The cardio-pneumo-psychogram and its use in the study of the emotions, with practical applications. J. exp. Psychol., 1922, 5, 323­328.

MARSTON, W. M. Systolic blood pressure symptoms of deception. J. exp. Psychol., 1917, 2, 117-163.

WENGER, M. A., ENGLE, B. T., & CLEMENS, T. K. Studies of autonomic response patterns: Rationale and methods. Behav. Sci., 1957, 2, 216-221.

(Received December 4, 1962)

The preceding paper is a reproduction of the following article (Gustafson, L. A., & Orne, M. T. Effects of heightened motivation on the detection of deception. Journal of Applied Psychology, 1963, 47, 408-411.). It is reproduced here with the kind permission of the American Psychological Association © 1963. No further reproduction or distribution of this article is permitted without written permission of the publisher.