- Email: [email protected]
Focusing versus distraction and the response to clinical electric shocks
Journal of Behavior Therapy and Experimental Psychiatry 30 (1999) 199}204
Focusing versus distraction and the response to clinical electric shocks Pieter C. Duker*, John van den Bercken, Marie-Anne Foekens Psychology Lab. A05.04, University of Nijmegen, P.O. Box 6104, 6500 HE Nijmegen, The Netherlands
Abstract This study pertains to assessing the e!ects of electric shocks that are used in the treatment of severe self-injurious behavior. With pain sensation and startle response as the dependent variables and focusing versus distraction of recipient's attention to the electric shocks as the independent variable, these stimuli were administered to 60 paid volunteers. Using ANOVA, no signi"cant e!ect of the independent variable was found on either measure. However, repeated administration of the electric shock produced a signi"cant increment of pain sensation, with a concomitant signi"cant decrease of magnitude of the startle response. No interaction e!ect was found. ( 1999 Elsevier Science Ltd. All rights reserved.
Although contingent shock (CS) can be applied as treatment with a variety of severe behavior problems, it is currently most often used in the treatment of refractory and health-endangering self-injurious behavior with individuals with mental retardation and/or autism. A meta-analysis on over 50 predominantly single-subject studies using CS reveals an e!ectiveness score that largely exceeds that of other treatment procedures of self-injurious behavior (SIB) (see Didden, Duker & Korzilius, 1997). Despite the controversy regarding the use of aversive stimuli with individuals who are unable to give their consent, refraining them from the most e!ective treatment poses an even more serious ethical problem, given the often damaging e!ects of this behavior (i.e., blindness, brain injury). It should be acknowledged that studies exploring the variables controlling the sensation of clinical electric shocks are virtually non-existent. This issue is of major clinical importance, given the fact that if suppression of SIB fails to occur, this is
* Corresponding author. Tel.: #31-24-3612822; fax: #31-24-3612776. 0005-7916/99/$ - see front matter ( 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 0 5 - 7 9 1 6 ( 9 9 ) 0 0 0 2 5 - 7
200
P.C. Duker et al. / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 199}204
attributed to the characteristics of the aversive stimuli being used (von Heyn, Israel & Worsham, 1993; Williams, Kirkpatrick-Sanchez & Iwata, 1993). Duker, Reurslag and van den Bercken (1998) found that the personality factor introversion/extroversion accounted for statistically signi"cant di!erences in pain sensation ratings of clinical electric shocks, with extroverts producing lower scores. The objective of research in this area is to reveal procedural information that may enhance the e!ectiveness of CS. A relevant study in this respect was conducted by Arntz, Dreessen and Merckelbach (1991). They found that focusing the recipient's attention towards the place of delivery of the shock (i.e., focusing) led to a signi"cant stronger sensation than to instruct the recipient to disregard the stimulus as much as possible (i.e., distraction). Not less important was their "nding that, across repeated administrations of the shock, focusing preserved the aversive sensation of the stimulus, whereas distraction led to habituation. If focusing to the stimulus would block habituation, faster conditioning and thereby faster suppression of the response is to be expected and fewer electric shocks are needed to suppress the target response (e.g., the self-injurious response). The purpose of the present study is to investigate the di!erential e!ect of focusing versus distraction on subjects' pain sensation and startle response to the administration of electric shocks.
1. Method 1.1. Participants Sixty university students (51 women and 9 men, mean age 19.3 yr, age range: 17}24) volunteered and were paid D# 15 (approx. $5) for their participation. They were informed that they would be administered electric shocks at an intensity that is being used in the treatment of severe SIB with individuals with a mental handicap and/or autism. Participants remained, however, naive with respect to the speci"c experimental purpose of the study (i.e., focusing versus distraction). They had signed a consent form. Participants were free to quit the experiment, but were then denied from payment. 1.2. Device Shocks were delivered remotely with the HSP 3012, a device that is used in the clinical practice (see Duker & Seys, 1996). The characteristics of the electric stimuli (i.e., AC; 40 mA; 30 Hz; impedance"1 k); duration of shock "0.5 s) are similar to those that are administered to clients with severe SIB. The units of the device are: (a) a remote control unit with a battery, which, when pressed, transmits a coded signal to activate the receiver; (b) a receiver (8 cm]8 cm]5 cm; 425 g) consisting of an inductorium, a radio frequency receiver, and a 5.5 V battery. Administration of shocks occurred through two circular electrodes (surface size 1 cm, set 40 mm apart) attached to participant's forearm; and (c) a charger, equipped with control lights indicating the
P.C. Duker et al. / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 199}204
201
condition of the batteries of the above units. The device has been approved by the Belgian, Canadian, and Dutch health authorities. 1.3. Procedure The participants were examined individually and randomly assigned to either the focusing condition (n"30) or the distraction condition (n"30). They were informed that 5 electric shocks would be administered to them and that they had to rate the intensity of the sensation of the shock following each single administration with a number ranging from 0 to 10 (see Recording). They were also informed that their behavior was videotaped during the experimental session. Duration of the session was approximately 10 min for each participant. Focusing. While the electrical device was visible (i.e., located on the table) for the participant, each single administration of the shock was preceded by the experimenter pointing to the electrodes on participant's forearm and saying: `attention, the shock comes there and it comes nowa. Simultaneously, the experimenter pressed the button on top of the sender, visible for the participant, and the shock was delivered to the body. Interstimulus intervals were approximately 2 min. Distraction. While the electrical device remained out of sight, the participant was requested to read loudly a text on publishing scienti"c reports. While reading, the experimenter unpredictably administered each of the "ve shocks. Interstimulus intervals varied, therefore, in duration. 1.4. Recording There were two dependent variables. Pain sensation. Following each single administration of the shock the participant had to verbally rate the pain sensation on a scale ranging from 0 (not painful at all) to 10 (extremely painful, intolerable). Such a scale has proven accurate to assess subjective pain intensity (Jensen, Karoly & Braver, 1986). Startle response. Video recordings were made of the participant's face, hands/arms, and upper trunk immediately before, during, and following each single administration of the shock. These recordings were rated by the experimenter and a second rater, who remained naive as to the purpose of the study, using a 4-point scale. Scale values were: no movement of face, hands/arms, and trunk (0); some movement of face (including mouth and eyes), hands/arms, and trunk (1); clear movement of face, participant raises arm to which the electrodes are attached rapidly; raises body 5}20 cm from chair (2); twitches face, raises arm with electrodes rapidly and pulls it back; raises body from chair over 20 cm, throws body backward (3). 1.5. Reliability Interobserver agreement for recording startle responses was obtained by comparing the ratings of the two observers, who independently and simultaneously rated these responses for 50% of the participants in both experimental conditions. By dividing the
202
P.C. Duker et al. / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 199}204
number of agreements by the number of disagreements plus the number of agreements, times one hundred, an agreement of 97% was found. 2. Results 2.1. Pain sensation Fig. 1 shows the mean values of the pain sensation ratings for the conditions of focusing and distraction for each of the "ve successive administrations of electric shock. Using a 2 (conditions) ]5 (successive administration of shocks) ANOVA with repeated measures we failed to "nd a statistically signi"cant main e!ect for the conditions of focusing versus distraction, F(1, 58)"0.30, p"0.59. Successive administration of shocks produced a signi"cant main e!ect, F(4, 55)"9.58, p(0.05, indicating that, across administrations of electric shock, higher pain sensation ratings were obtained in either condition. There was no interaction e!ect between conditions and successive administrations of shock, F(4, 55)"1, p"0.42. Further testing of di!erences between adjacent administrations of electric shock across conditions revealed a signi"cant increase in pain sensation ratings between the "rst and second administration of shock, t"2.39, p(0.05, 2-tailed, and the third and fourth administration of shock, t"3.62, p(0.05, 2-tailed. 2.2. Startle response Fig. 2 shows the mean values of the startle responses for the conditions of focusing and distraction for each of the successive administrations of electric shock. Using
Fig. 1. Mean pain sensation ratings at successive administrations of electric shock.
P.C. Duker et al. / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 199}204
203
Fig. 2. Mean startle responses at successive administrations of electric shock.
ANOVA no signi"cant di!erence was found in participants' startle response to electric shock between the conditions of focusing and distraction, F(1, 58)"0.53, p"0.47. Successive administrations of shock appeared to produce a signi"cant e!ect in startle response, F(4, 55)"8.28, p(0.05. This indicates that, across administrations of shock, lower magnitudes of the startle response were obtained in either condition. There was no signi"cant interaction e!ect, F(4, 55)"2.42, p"0.06. Testing di!erences between adjacent administrations of shock across conditions revealed a signi"cant decrease in startle response between the "rst and second administration of shock only, t"4.39, p(0.05, 2-tailed. Analysis revealed that a negative relationship between the two dependent variables was approached, r"!0.832, p"0.08.
3. Discussion The main conclusion of this study is that, given the electric stimulus being used, the conditions of focusing and distraction fail to di!er in their e!ect on participants' pain sensation and startle responses. These results di!er from those obtained by Arntz et al. (1991). The di!erence may be due to the intensity of the stimulus, in that Arntz et al. employed a relatively mild electric shock, whereas we used a clinically relevant shock that, in most cases, succeeds to suppress severe SIB (Duker & Seys, 1996). Also, the number of stimuli given di!ered between both studies. The increased pain sensation across "ve administrations of the electric shock is a "nding that is useful for the clinical practice in that habituation to the stimulus is not very likely to occur, provided that not too many shocks are given during a relatively short period of time (i.e., 10 min). This observation is congruent with our own
204
P.C. Duker et al. / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 199}204
experience of CS with clients with mental retardation and severe SIB. When mechanical restraints are used, a high rate of SIB usually occurs when these restraints are removed, forcing the behavior therapist to administer a large number of contingent shocks to suppress SIB at the early stages of the treatment. Our "nding is that this practice often results in a failure to suppress SIB, a "nding that may be related to a process of habituation to the shock. If this occurs conditioning fails to occur. The present data may indicate that repeated administration of electric shocks need not result in habituation, provided that shocks are given in an expanded period of time. Future research should focus on the di!erential e!ect of shock on pain sensation when these shocks are given within di!erent time windows. A second clinically relevant "nding of this study is that the startle responses were unrelated to the pain sensation ratings. As e!ectiveness of the clinical electric shock is more likely inferred from client's overt responses to the shock, rather than on the basis of the suppression rate of the deviant behavior, this may lead to false conclusions regarding the e!ectiveness of the treatment. This study also demonstrates that 60 (healthy) individuals, who were administered a total of 300 painful electric shocks failed to experience any negative e!ects, such as fainting or exhibiting aggressive reactions. Only one participant quit the experiment. The present results must be interpreted cautiously, because in this study electric shocks are not given following SIB and are not supposed to suppress an operant behavior. We dealt with healthy subjects rather than individuals who had mental retardation and other debilitating medical conditions. Nevertheless, as controlled research in this speci"c area with clients is ethically and practically impossible, obtaining information in an analogue framework is considered `second besta. References Arntz, A., Dreessen, L., & Merckelbach, H. (1991). Attention, not anxiety, in#uences pain. Behaviour Research and Therapy, 29, 41}50. Didden, R., Duker, P. C., & Korzilius, H. (1997). Meta-analytic study on treatment e!ectiveness for problem behaviors with individuals who have mental retardation. American Journal on Mental Retardation, 101, 387}399. Duker, P. C., Reurslag, I., & van den Bercken, J. (1998). Variables in#uencing the sensation of clinical electric shock. Unpublished manuscript. Duker, P. C., & Seys, D. M. (1996). Long-term use of electrical aversion treatment with self-injurious behavior. Research in Developmental Disabilities, 17, 293}301. Jensen, M. P., Karoly, P., & Braver, S. (1986). The measurement of clinical pain intensity: A comparison of six methods. Pain, 27, 117}126. von Heyn, R. E., Israel, M. L., & Worsham, R. W. (1993). A comparison of the long-term decelerative ewectiveness of two intensities of contingent electric shock on aggressive and health dangerous behavior with individuals with severe behavioral disorders (Publication no. 93-2) Behavior Research Institute, Providence, Rhode Island. Williams, D. E., Kirkpatrick-Sanchez, S., & Iwata, B. A. (1993). A comparison of shock intensity in the treatment of longstanding and severe self-injurious behavior. Research in Developmental Disabilities, 14, 207}219.
Focusing versus distraction and the response to clinical electric shocks Pieter C. Duker*, John van den Bercken, Marie-Anne Foekens Psychology Lab. A05.04, University of Nijmegen, P.O. Box 6104, 6500 HE Nijmegen, The Netherlands
Abstract This study pertains to assessing the e!ects of electric shocks that are used in the treatment of severe self-injurious behavior. With pain sensation and startle response as the dependent variables and focusing versus distraction of recipient's attention to the electric shocks as the independent variable, these stimuli were administered to 60 paid volunteers. Using ANOVA, no signi"cant e!ect of the independent variable was found on either measure. However, repeated administration of the electric shock produced a signi"cant increment of pain sensation, with a concomitant signi"cant decrease of magnitude of the startle response. No interaction e!ect was found. ( 1999 Elsevier Science Ltd. All rights reserved.
Although contingent shock (CS) can be applied as treatment with a variety of severe behavior problems, it is currently most often used in the treatment of refractory and health-endangering self-injurious behavior with individuals with mental retardation and/or autism. A meta-analysis on over 50 predominantly single-subject studies using CS reveals an e!ectiveness score that largely exceeds that of other treatment procedures of self-injurious behavior (SIB) (see Didden, Duker & Korzilius, 1997). Despite the controversy regarding the use of aversive stimuli with individuals who are unable to give their consent, refraining them from the most e!ective treatment poses an even more serious ethical problem, given the often damaging e!ects of this behavior (i.e., blindness, brain injury). It should be acknowledged that studies exploring the variables controlling the sensation of clinical electric shocks are virtually non-existent. This issue is of major clinical importance, given the fact that if suppression of SIB fails to occur, this is
* Corresponding author. Tel.: #31-24-3612822; fax: #31-24-3612776. 0005-7916/99/$ - see front matter ( 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 0 5 - 7 9 1 6 ( 9 9 ) 0 0 0 2 5 - 7
200
P.C. Duker et al. / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 199}204
attributed to the characteristics of the aversive stimuli being used (von Heyn, Israel & Worsham, 1993; Williams, Kirkpatrick-Sanchez & Iwata, 1993). Duker, Reurslag and van den Bercken (1998) found that the personality factor introversion/extroversion accounted for statistically signi"cant di!erences in pain sensation ratings of clinical electric shocks, with extroverts producing lower scores. The objective of research in this area is to reveal procedural information that may enhance the e!ectiveness of CS. A relevant study in this respect was conducted by Arntz, Dreessen and Merckelbach (1991). They found that focusing the recipient's attention towards the place of delivery of the shock (i.e., focusing) led to a signi"cant stronger sensation than to instruct the recipient to disregard the stimulus as much as possible (i.e., distraction). Not less important was their "nding that, across repeated administrations of the shock, focusing preserved the aversive sensation of the stimulus, whereas distraction led to habituation. If focusing to the stimulus would block habituation, faster conditioning and thereby faster suppression of the response is to be expected and fewer electric shocks are needed to suppress the target response (e.g., the self-injurious response). The purpose of the present study is to investigate the di!erential e!ect of focusing versus distraction on subjects' pain sensation and startle response to the administration of electric shocks.
1. Method 1.1. Participants Sixty university students (51 women and 9 men, mean age 19.3 yr, age range: 17}24) volunteered and were paid D# 15 (approx. $5) for their participation. They were informed that they would be administered electric shocks at an intensity that is being used in the treatment of severe SIB with individuals with a mental handicap and/or autism. Participants remained, however, naive with respect to the speci"c experimental purpose of the study (i.e., focusing versus distraction). They had signed a consent form. Participants were free to quit the experiment, but were then denied from payment. 1.2. Device Shocks were delivered remotely with the HSP 3012, a device that is used in the clinical practice (see Duker & Seys, 1996). The characteristics of the electric stimuli (i.e., AC; 40 mA; 30 Hz; impedance"1 k); duration of shock "0.5 s) are similar to those that are administered to clients with severe SIB. The units of the device are: (a) a remote control unit with a battery, which, when pressed, transmits a coded signal to activate the receiver; (b) a receiver (8 cm]8 cm]5 cm; 425 g) consisting of an inductorium, a radio frequency receiver, and a 5.5 V battery. Administration of shocks occurred through two circular electrodes (surface size 1 cm, set 40 mm apart) attached to participant's forearm; and (c) a charger, equipped with control lights indicating the
P.C. Duker et al. / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 199}204
201
condition of the batteries of the above units. The device has been approved by the Belgian, Canadian, and Dutch health authorities. 1.3. Procedure The participants were examined individually and randomly assigned to either the focusing condition (n"30) or the distraction condition (n"30). They were informed that 5 electric shocks would be administered to them and that they had to rate the intensity of the sensation of the shock following each single administration with a number ranging from 0 to 10 (see Recording). They were also informed that their behavior was videotaped during the experimental session. Duration of the session was approximately 10 min for each participant. Focusing. While the electrical device was visible (i.e., located on the table) for the participant, each single administration of the shock was preceded by the experimenter pointing to the electrodes on participant's forearm and saying: `attention, the shock comes there and it comes nowa. Simultaneously, the experimenter pressed the button on top of the sender, visible for the participant, and the shock was delivered to the body. Interstimulus intervals were approximately 2 min. Distraction. While the electrical device remained out of sight, the participant was requested to read loudly a text on publishing scienti"c reports. While reading, the experimenter unpredictably administered each of the "ve shocks. Interstimulus intervals varied, therefore, in duration. 1.4. Recording There were two dependent variables. Pain sensation. Following each single administration of the shock the participant had to verbally rate the pain sensation on a scale ranging from 0 (not painful at all) to 10 (extremely painful, intolerable). Such a scale has proven accurate to assess subjective pain intensity (Jensen, Karoly & Braver, 1986). Startle response. Video recordings were made of the participant's face, hands/arms, and upper trunk immediately before, during, and following each single administration of the shock. These recordings were rated by the experimenter and a second rater, who remained naive as to the purpose of the study, using a 4-point scale. Scale values were: no movement of face, hands/arms, and trunk (0); some movement of face (including mouth and eyes), hands/arms, and trunk (1); clear movement of face, participant raises arm to which the electrodes are attached rapidly; raises body 5}20 cm from chair (2); twitches face, raises arm with electrodes rapidly and pulls it back; raises body from chair over 20 cm, throws body backward (3). 1.5. Reliability Interobserver agreement for recording startle responses was obtained by comparing the ratings of the two observers, who independently and simultaneously rated these responses for 50% of the participants in both experimental conditions. By dividing the
202
P.C. Duker et al. / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 199}204
number of agreements by the number of disagreements plus the number of agreements, times one hundred, an agreement of 97% was found. 2. Results 2.1. Pain sensation Fig. 1 shows the mean values of the pain sensation ratings for the conditions of focusing and distraction for each of the "ve successive administrations of electric shock. Using a 2 (conditions) ]5 (successive administration of shocks) ANOVA with repeated measures we failed to "nd a statistically signi"cant main e!ect for the conditions of focusing versus distraction, F(1, 58)"0.30, p"0.59. Successive administration of shocks produced a signi"cant main e!ect, F(4, 55)"9.58, p(0.05, indicating that, across administrations of electric shock, higher pain sensation ratings were obtained in either condition. There was no interaction e!ect between conditions and successive administrations of shock, F(4, 55)"1, p"0.42. Further testing of di!erences between adjacent administrations of electric shock across conditions revealed a signi"cant increase in pain sensation ratings between the "rst and second administration of shock, t"2.39, p(0.05, 2-tailed, and the third and fourth administration of shock, t"3.62, p(0.05, 2-tailed. 2.2. Startle response Fig. 2 shows the mean values of the startle responses for the conditions of focusing and distraction for each of the successive administrations of electric shock. Using
Fig. 1. Mean pain sensation ratings at successive administrations of electric shock.
P.C. Duker et al. / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 199}204
203
Fig. 2. Mean startle responses at successive administrations of electric shock.
ANOVA no signi"cant di!erence was found in participants' startle response to electric shock between the conditions of focusing and distraction, F(1, 58)"0.53, p"0.47. Successive administrations of shock appeared to produce a signi"cant e!ect in startle response, F(4, 55)"8.28, p(0.05. This indicates that, across administrations of shock, lower magnitudes of the startle response were obtained in either condition. There was no signi"cant interaction e!ect, F(4, 55)"2.42, p"0.06. Testing di!erences between adjacent administrations of shock across conditions revealed a signi"cant decrease in startle response between the "rst and second administration of shock only, t"4.39, p(0.05, 2-tailed. Analysis revealed that a negative relationship between the two dependent variables was approached, r"!0.832, p"0.08.
3. Discussion The main conclusion of this study is that, given the electric stimulus being used, the conditions of focusing and distraction fail to di!er in their e!ect on participants' pain sensation and startle responses. These results di!er from those obtained by Arntz et al. (1991). The di!erence may be due to the intensity of the stimulus, in that Arntz et al. employed a relatively mild electric shock, whereas we used a clinically relevant shock that, in most cases, succeeds to suppress severe SIB (Duker & Seys, 1996). Also, the number of stimuli given di!ered between both studies. The increased pain sensation across "ve administrations of the electric shock is a "nding that is useful for the clinical practice in that habituation to the stimulus is not very likely to occur, provided that not too many shocks are given during a relatively short period of time (i.e., 10 min). This observation is congruent with our own
204
P.C. Duker et al. / J. Behav. Ther. & Exp. Psychiat. 30 (1999) 199}204
experience of CS with clients with mental retardation and severe SIB. When mechanical restraints are used, a high rate of SIB usually occurs when these restraints are removed, forcing the behavior therapist to administer a large number of contingent shocks to suppress SIB at the early stages of the treatment. Our "nding is that this practice often results in a failure to suppress SIB, a "nding that may be related to a process of habituation to the shock. If this occurs conditioning fails to occur. The present data may indicate that repeated administration of electric shocks need not result in habituation, provided that shocks are given in an expanded period of time. Future research should focus on the di!erential e!ect of shock on pain sensation when these shocks are given within di!erent time windows. A second clinically relevant "nding of this study is that the startle responses were unrelated to the pain sensation ratings. As e!ectiveness of the clinical electric shock is more likely inferred from client's overt responses to the shock, rather than on the basis of the suppression rate of the deviant behavior, this may lead to false conclusions regarding the e!ectiveness of the treatment. This study also demonstrates that 60 (healthy) individuals, who were administered a total of 300 painful electric shocks failed to experience any negative e!ects, such as fainting or exhibiting aggressive reactions. Only one participant quit the experiment. The present results must be interpreted cautiously, because in this study electric shocks are not given following SIB and are not supposed to suppress an operant behavior. We dealt with healthy subjects rather than individuals who had mental retardation and other debilitating medical conditions. Nevertheless, as controlled research in this speci"c area with clients is ethically and practically impossible, obtaining information in an analogue framework is considered `second besta. References Arntz, A., Dreessen, L., & Merckelbach, H. (1991). Attention, not anxiety, in#uences pain. Behaviour Research and Therapy, 29, 41}50. Didden, R., Duker, P. C., & Korzilius, H. (1997). Meta-analytic study on treatment e!ectiveness for problem behaviors with individuals who have mental retardation. American Journal on Mental Retardation, 101, 387}399. Duker, P. C., Reurslag, I., & van den Bercken, J. (1998). Variables in#uencing the sensation of clinical electric shock. Unpublished manuscript. Duker, P. C., & Seys, D. M. (1996). Long-term use of electrical aversion treatment with self-injurious behavior. Research in Developmental Disabilities, 17, 293}301. Jensen, M. P., Karoly, P., & Braver, S. (1986). The measurement of clinical pain intensity: A comparison of six methods. Pain, 27, 117}126. von Heyn, R. E., Israel, M. L., & Worsham, R. W. (1993). A comparison of the long-term decelerative ewectiveness of two intensities of contingent electric shock on aggressive and health dangerous behavior with individuals with severe behavioral disorders (Publication no. 93-2) Behavior Research Institute, Providence, Rhode Island. Williams, D. E., Kirkpatrick-Sanchez, S., & Iwata, B. A. (1993). A comparison of shock intensity in the treatment of longstanding and severe self-injurious behavior. Research in Developmental Disabilities, 14, 207}219.