International Journal of Psychophysiology, 12 (1992) 71-79 0 1992 Elsevier Science Publishers B.V. All rights reserved 0167~8760/92/$05.00
71
INTPSY 00354
atterns an
ifferences i
Francisco Rom5n, Eduvigis Carrillo and Francisco A. Garcia-Sgnchez Departamento de Ciencias Morfol~gicas, Anatomiu Patokigica y Psicobiologia, Facuitad de Psicologia, Unirersidad de Mwcia, Murcic (Spain)
(Accepted IS April f991)
Key words: Electrodermal asymmetry: Responsiveness pattern; Handedness; Skin conductance; Habituation
The aim of the present study was to examine the relationship between subjects’ handedness and electrodermal asymmetry when subjects were divided according to their electrodermal Responsiveness Patterns, obtained during a stimulus-free period. Skin conductance was bilaterally recorded in a sample of 48 subjects (24 dextrals and 24 sinistrals) durirrg the performance of two tasks (watching a film and listening to neutral tones). Subjects were divided into four groups of 12 subjects each (dextral right-hand responders, dextral left-hand responders, sinistral right-hand responders and sinistral left-hand responders). No differences were found between dextral and sinistral subjects. Orientation of electrodermal asymmetry in each Responsiveness Patterns remained constant throughout the two conditions: right-hand responders showed higher electrodermal activity on their right hand, while left-hand responders showed higher electrodermal activity on their left hand. In addition, left-hand responders showed a slower habituation rate on their left hand than on their right hand, whereas right-hand responders showed no significant differences between hands.
INTRODUCTION Most works on bilateral electrodermal activity (EDA) have assumed that differences in orientation of electrodermal asymmetry could be related to differential hemispheric activation and their conclusions have focussed on this (see Hugdahl, 1984; Freixa i Baque et al., 1984a). Because differences in hemispheric organization can be related to subjects’ handedness (see e.g., Curtis, 19851, it has been considered as an important variable in electrodermal asymmetry researches. However, the majority of works about electrodermal asymmetry either do not indicate the sub-
Currespundenw
F. Roman, Area de Psicobiologia. Facultad de Psicologia, Universidad de Murcia. 30071 Murcia, Spain.
jects’ hand-preference (e.g. Obrist, 1963; Patterson and Venables, 1978; Freixa i Baque and De Bonis, 1983; Freixa i Baque et al., 1984b) or only dextral subjects are selected attempting to control possible influences of handedness in electroderma1 asymmetry (e.g. Erwin et al., 1980; Gross and Stern, 1980; Ketterer and Smith, 1977; Lacroix and Comper, 1979; Smith and Ketterer, 1982; Boyd and Maltzman, 1983, 1984; Martinez-Selva et al., 1987; Roman et al., 1987; Roman et al., 1989). rfowever, a few researchers have studied differences between dextral and sinistrai subjects in bilateral EDA (Wyatt and Turskey, 1969; Robinson and Zahn, 1981; Smith, Ketterer and Concannon, 1981; Gruzelier et al., 1981). Wyatt and Turskey (1969), Gruzelier et al. (1981) and Smith et al. (1981) did not find significant differences in the orientation of electrodermal asymmetry between dextral and sinistral sub-
72
jects. Lacroix and Comper (1979) found different patterns of bilateral electrodermal responses between dextral females when cognitive tasks (verbal and spatial) were employed, but did not find differences in sinistral females who showed always a higher EDA on their left hand independently of the tasks employed. This result was confirmed by Robinson and Zahn (1981) who observed that sinistral subjects showed higher nonspecific and specific skin conductance responses (NSR and SCR, respectively) and skin conductance levels @CL) on their left hand for all conditions presented, while dextral subjects showed higher EDA on their right hand for the most conditions presented. In a previous paper (Roman et al. 1989), the authors considered the possibility that bilateral EDA could show a specific latcralization independently of the tasks being emp!oyed for activating a single hemisphere. In that work, the authors observed that electrodermal lateralization remained constant through different experimental conditions. In addition, gender, which is considered an important variable in the studies of hemispheric specialization, did not appear important in the orientation of EDA when electrodermal responsiveness patterns (RP) and the percentage of subjects with a specific RP were taken into account in each experimental group. Then, the authors indicated the possibility that these RPs could help to explain the results obtained by different authors (e.g., Erwin et al. 1980; De aque, 1980; Gross and Stern, 1980; Fedora and Schopflocher, 1984; MartinezSelva et al. 1987; RornGn et al. 1987), who have reported that the direction of the asymmetry in the EDA remained constant through different experimental conditions. In the same way that the authors had considered the possibility that sex differences in the orientation of electrodermal asymmetry could be due to different percentages of subjects with a right-hand or left-hand dominance in EDA in the male and female groups (RomBn et al. 1989), in the present study we consider the possibility that subjects’ handedness does not affect the orientation of electrodermal asymmetry when subjects arc divided into right-hand and left-hand respon-
ders according to their Laterality Coefficient (LC) during a stimulus-free period. The purpose of the present paper was to study the orientation of bilateral electrodermal asymmetry between dextral and sinistral subjects when the RPs are considered and there are the same proportions of left-hand and right-hand responders in each experimental group. Differences hetween dextral and sinistral subjects in SCRs, NSRs, and SCLs, will be analyzed. Simple tasks, such as presentation of neutral tones and a cartoon film sequence were employed: the first to obtain SCRs and the second NSRs. The employment of a series of neutral tones allowed the measurement of habituation and the analysis of the relationship between habituation and electrodermal asymmetry in right-hand and left-hand responders. According to previous data (Romin et al., 1987, 1989), it was expected each electrodermal RP should remain constant during the two conditions: presentation of neutral tones (SCRs) and cartoon film sequence WSRs). In addition, it was predicted that electrodermal asymmetry observed during the stimulus-free period would indicate orientation of the electrodermal asymmetry of both SCRs and NSRs and, besides, that no differences between dextral and sinistral subjects would be found within each electrodermal RP. In relation to Habituation, and according to previous results (Martinez-Se:va et al. 19871, it was finally predicted that habituation would be faster on the left hand in the right-hand responders group and on the right hand in the left-hand responders group.
METHOD Subjects
A total of 48 first-year psychology undergi-aduates at the University of Murcia (24 dextral and 24 sinistral subjects) volunteered for the experiment. The age range was 17 to 29 years (average age for dextral subjects = 22.6 years, S.D. = 2.60; average age for sinistral subjects = 21.4, S.D. = 1.81). Subjects were divided into right-hand and left-hand responders depending on their LC, ob-
73 TABLE 1 Vulaes of the Laterality Coefficient in the stimulus-free period Right-hand responders Left-hand responders mean
Dextrals Sinistrals
0.86 - 0.95
S.D.
mean
S.D.
0.21
0.92 - 0.99
0.20 0.03
0.13
tained from an initial stimulus-free period (see Response identification and quantification section below for definition), giving 12 subjects in each group that is, dextrals right-hand responders, dextrals left-hand responders, sinistrals right-hand responders and sinistrals left-hand responders. Once completed, the experimental groups subjects were assigned to other investigations. LC means and standard deviations of these groups in the stimulus-free period are shown in Table 1. Lateral dominance was assessed using the Spanish adaptation of Harris’ tests of lateral dominance (Harris, 1974). The criterion for identifying dextra! subjects was that they had to obtain 100% of responses indicating dextra! dominance and the criterion for identifying sinistral subjects was that they had to obtain at least 90% of responses indicating sinistra! dominance, a!ways employing their left hand for writing and the left foot for kicking a ball. This criterion was established due to the difficulty of finding large samples of sinistra! subjects in the Spanish population. Ambidextrous subjects (n = 7) were excluded from statistical analyses, along with four other subjects who did not give any electroderma! response in the initial stimulus-free period. Two additional subjects were dropped because of a technical problem in the recording. Subjects were replaced to complete the sample of 48 subjects who participated in the experiment. Subjects received course credits for participation. All par ticipants were treated in accordance with the ‘Ethical Principles of Psychologists’ (American Psychological Association, 198 I). Apparatus
Auditory stimu!! (neutral tones) were produced by a sine wave generator, Letica Stimula-
tor 150 (rise and decay time < 1 ms). Stimuli were delivered via Audiotechnica ATH-2,8 ohms, stereoheadphones, calibrated at 80 dB in the output. A tape recorder, TEAC-A-3440 (TEAC Co., Tokyo, Japan), was used to record and to deliver the instructions. The film sequence was recorded and presented by a videocassette rccorder JVC (JVC Victor Company of Japan Limited), VHS system, on a television set, placed 1.5 m in front of the subjects. Bilateral skin conductance was recorded on a four-channel Net San Ei oscillograph (Net San-Ei Instruments, Ltd. Tokyo, Japan), through two amplifiers with Coulbourn S71-22 skin conductance couplers which provided 0.5 V (DC) constant voltage excitation (Coulbourn Instruments, Lehight Valley, PA, USA). Because of their size (7 mm diameter Ag/AgCl, Couiboum Tll-01) electrodes were attached by double faced adhesive collars to the thenar (C6) and hypothenar (03) eminences of both hands. A bioge! (LA 76621, 0.068 M NaC!) was used as the contact medium. The skin was cleansed with soap and water. The electrodes, amplifiers and pens were alternated between hands from subject to subject to eliminate the possibility of systematic influences. The chart speed was 5 mm/s. Procedure
Subjects were seated in a comfortable armchair in a darkened, sound-attenuated chamber (2.5 X 2 X 2.40 m). The recording apparatus was located in an adjoining room. An intercom rystem allowed communication with the subject, who could be seen ;hrough a small glass window in one of the chamber walls. The temperature in this chamber was held constant 235°C (k 1.5”C). After electrodes and headphones had been attached by one female experimenter, general instmctions were given through a loudspeaker placed in the chamber, 1.5 m in front of the subjects and a 5 min adaptation period was observed before the stimulus-free period which lasted 3 min. Subjects were told that they would hear a series of harmless tones and watch a film sequence and that the experimenter was interested in recording their physiological responses to those conditions. They were also told to remain
74
quiet and to avoid movements. Immediately before each condition, subjects were informed of its beginning. After the stimulus-free period the subjects received the neutral tones and the cartoon film sequence which were counterbalanced in the order of presentation for all the subjects. An interval of 2 min was observed between the conditions. The duration of the film condition was 2 min and consisted of a sequence of a cartoon movie for children. Auditory stimuli consisted of 12 tones of 2 s (80 dB and 1000 Hz). Interstimulus intentals varied randomly between 15 and 45 s, with an average duration of 30 s. The recording session had a total duration of 22 min, approximately. Immediately after the recording session, subjects completed the handedness inventory. Response identification and quantification Polygraph recordings were scored by hand. Compound responses were scored from an inflection point to peak, as recommended by Edelberg (1967). Any artifact-free change in skin conductance higher than 0.05 pmhos was considered an electrodermal response. SCLs were taken immediately before and after the stimulus-free period and the film condition, and seven-times during the neutral tones presentation: before the first tone, and after stimuli number 2, 4, 6, 8, 10 and 12. A SCR was the first response taking piace l-5 s after stimulus onset. The criterion for habituation was three consecutive trials with no s were scored during the initial 3 min stimulus-free period and during the 2 min film condition. NSRs dur’ng the initial stimulus-free period were only used to determine the RP of each subject. The LC proposed by Freixa i BaquC and De Bonis (19831 was employed to obtain electrodermal lateraiity: %R>L-%L>R %R>Lc%L>R %R > L reflected the percentage of cases in which the right hand showed an electrodermal response of a higher magnitude than the accompanying response on the left hand, and %L > R reflected the percentage of cases in which the
contrary occurred. Differences in magnitude between simultaneous electrodermal responses in each hand below 0.05 pmhos (differentiai response criterion) were not considered and R =zL decision was made in these cases. The observation of a response on one hand but not on the other hanll resulted in an amplitude of 0 pmhos being scored for the latter. The value of LC was zero when %R > L = %L > R, positive when the asymmetrical electrodermal responses were of higher magnitude on the right hand (R > LJ, and negative when the asymmetrical electrodermal responses were of higher magnitude on the left hand (L > R). LC range from - 1 to + 1. RP in each subject was obtained according to their LC scores: When the value was positive then the subject was considered a right-hand responder; when negative, a left-hand responder. The denominator of this formula was employed to obtain the asymmetry proportion (API, which was defined as the total sum of all the percentages of asymmetrical electroderma1 responses in both hands (%cR > f. + %L > R). The response frequency was the number of NSRs per minute. In the response magnitude analysis, all instances in which a phasic response of measurable magnitud: (define’d as 0.05 pmhos or higher) occurred in either hand were included. In this analysis, both asymmetrical responses (R > L and L > RI and symmetrical responses (R = L) were included. Again, the observation of a response from one hand but not from the other resulted in an amplitude of 0 pmhos being scored for the latter. Data analysis Four independent two-way 2 x 2 ANOVAS were employed to analyse AP and LC data, during neural tones presentation and film conditions, rGJpectively, with Handedness and Responsiveness as between-subjects factors. Magnitude and frequency of NSRs during the film condition, the SCR magnitude data of the initial orienting response (l@R) to the first trial of the neutral tones presentation, and the habituation criterion (number of trials to habituation) were analyzed by four mixed analyses of variance
7s TABLE
II
Means and standard deviations fin parentheses) of the Responsivenesx Hand interaction an the variables analyzed
Right-hand responders
Left-hand resporrders
RH
LH
RH
LH
0.198 (0.075) 5.938 (2.585) 0.599 (0.375)
0.136 to.0791 5.458 (2.734) 0.423 (0.306)
0.092 (0.070) 4.854 (3.140) 0.385 (0.272)
0.184 (0.086) 7.208 (3.501) 0.636 (0.438)
0.933 (0.586) 0.345
0.676 (0.528) 0.219
0.464 CO.5861 C.275
0.957 (I ,002) 0.553
(0.441)
(0 291)
(0.427)
(0.764)
7.542
6.750
6.917
8,417
(4.880)
(4.413)
Film condition
Fdanspecificresponsesmagnitude tpmhos) Nonspecific responses frequency (resp/min) Skin conductance levels tpmhos) Tones condition Initial orienting response (r.rmhast Specific responsesmagnitude (pmhos) Criterion
for habituation
(trials to criterion) Skin conductance
(3.149) levels
(pmhos)
(4.71
I)
0.480
0.311
0.362
0.565
(0.395)
co.im,
(0.282)
(0.350)
RH. right hand: LH. left hand.
(2 X 2 X 2) Handedness X Responsiveness X Hand, with repeated measures in the last factor. SCLs during the film condition were analyzed by a mixed analysis of variance (2 x 2 x 2 X 2) Handedness x Responsiveness x Recording time X Hand, with repeated measures in the two last factors. There were two levels of recording time: pre and post film presentation. A 2 x 2 x 12 x 2 univariate analysis of variance (ANOVA), Handedness X Responsiveness x Trials x Hand, with Trials and Hand as within-subject factors, was performed for magnitude of SCRs to the 12 tones of the neutral tones presentation. On the other hand, SCLs data during this-condition were analyzed by a mixed anaiysis of variance (2 X 2 x 7 >: 2) Handedness X Responsiveness x Recording time x Hand, with repeated measures in the two last factors. In these two analyses a multivariate analysis of variance (MANOVA) on the within-subjects effects were run (O’Brien and Kaiser, 1985). A variable of the within-subjects portion of these two anaiyses was only considered significant when the Piilais trace multivariate statistic showed it (Wiikinson, 1988).
In ail first-order interactions simple main effects analyses were run, b:ing the mea13 qxare error corrected for all simple effects of withinsubjects factors following Kirk’s computational procedures (Kirk, 1982). The 0.05 rejection was adopted for ail statistical analyses.
RESULTS Film condition In the analysis of NSR magnitude data, only the Responsiveness x Hand interaction was significant, F(l,44) = 73.233, MSc = 0.002. Simple main effects analysis demonstrated that righthand responders showed higher NSR magnitudes on tneir right hand than on their left one, while left-hand responders showed the oppositt result, F(1,44) = 23.06, MSe = 0.002 and F(1,44) = 50.78, MSe = 0.002. respectively (means and standard deviations are shown in Table II). In addition, simple main effects showed that on the right hand right-hand responders showed significantly higher NSR magnitudes than left-hand responders (F(1,44) = 22.5, MSe = 0.0061, while signifi-
Dewtrals Rtght-nand
Dextrols
Slnl5trt7lS
Left
responders
Fig.1. Mean laterality coefficient for dentrnl and sinistral
right-h;md
-hand
Smlstrals responders
and left-hand respc)ndcrsduring film condition and IWPS
prrsrntation.
cant diffcrenccs bctwecn groups wcrc not observed on the left hand. The analysis of NSR frequency data showed significant effects in both the Hand factor and the Responsiveness x Hand interaction. Hand effect, F(1.44) = 7.674. MSc = 2.74’1. was due to higher frequrncy of NSRs on the left hand (mean = 6.333 rcsp,‘min. S.D. = 3.231) than on the right hanlT (mean = 5.396 rcsp/min. SD. = 2.809). In regard to the Rcsponsivcncss x Hand interaction. F( 1.44) = 17.534, MSe = 2.749, simple main cffccts analysis pointed out that left-hand responders showed more freqt.,:ncy of NSRs on their left hand than on their right hand, F(1.44) = 24.189. MSe = 2.74’1.while right-hand responders did not show significant differences between hands (means and standard deviations are shown in Table II). In the analysis of SCL data, only the Responsiveness x Hand interaction was significant. F( 1,44)= 25.21 I. MSe = 0.087. Simple main cffacts analysis pointed out that left-hand responders showed higher SCLs on their left hand than on their right hand, while right-hand responders did not show significant differences hctwecn hands, F(1.44) = 8.69, MSe = 0.087 (means and standard deviations arc shown in Table II).
The analysis of AP data did not show any significant effect. AP rnpans were 58.30% (S.D. = 27.87) and K0.05% (SD. = 19.89) for right-hand dextral and sinistral responders. respectively, and 73.26% (SD. = 30.57) and 77.84% (S.D. = 28.06) for left-hand dcxtral and sink&al responders. respectively. The analysis of LC data yielded a significant effect for the Responsiveness factor. F(1.44) = 68.607. MSc = 0.358. Means were 0.652 (S.D. = 0.6411) for right-hand responders and - 0.779 (S.D. = 0.577) for left-hand responders. Means of each RP for dcxtral and sinistral subjects in each condition are shown in Fig. 1. No other effects wefe significant.
The analysis of 10R magnitude to the first trial of neutral tones presentation only found a significant Responsiveness x Hand interaction. F( 1.44) = 32.454. MSe = 0.104. Simple main effects analysis showed that IOR was higher in right-hand responders on their right hand than on their left hand (F( 1,44)= 7.625, MSe = 0.104); while left-hand responders showed the opposite result (F(1.44) = 28.05. MSe = 0.104) (means and standard deviations are shown in Table II). In
75
addition, simple main effects showed a significant difference between RPs on the right hand (F(1,88) = 5.397, MSe = 0.489) due to higher IOR magnitudes in the group of right-hand responders, while significant differences were not obeen right-hand and left-hand responIn the analysis of SCR magnitude data both the Trials and the Hand factors were significant. In the Trials effect, F(11,484) = 12.635, MSe = 0.21, a trend analysis showed significant linear (F(1.44) = 38.014, MSe = 0.533), quadratic (F(1.44) = 19.730, MSe = 0.180). and cubic trends (F(1.44) = 13.824. MSe = 0.251). Higher order trends did not reach the significant level. The Hand effect (F( 1.44) = 4.116, MSe = 0.40s). was due to higher response magnitude on the left hand (mean = 3.863 pmhos, S.D. = 0.603) than on the right hand (mean = 3.099 pmhos, S.D. = 0.445). The other main effects, Handedness and Responsiveness, did not approach the significance level. Among the first-order interactions of this analysis, only the Responsiveness X Hand interaction was significant, F( 1,44) = 28.772, MSe = 0.408. Simple main effects analysis pointed out that right-hand responders showed higher magnitude of SCRs on their right hand than on their left hand (F(1.44) = 5.515. MSe = 0.408), while lefthand responders showed the opposite result (F( 1.44) = 27.277. MSe = 0.408) (means and standard deviations are shown in Table 11). Moreover, simple main effects analysis pointed out higher magnitude of SCRs on the left hand in left-hand responders than in right-hand responders (F(1.88) = 7.996, MSe = 1.997), while on the right hand differences between RPs were not significant (F < !). Among the second-order interactions, only, the Responsiveness x Trials x Hand interaction was significant, F(11,484) = 4.271, MSe = 0.033, P < 0.001. A multiple comparison Scheffe test pointed out that on the left hand left-hand responders showed higher SCR magnitudes than right-hand responders in all trials, except in the 2nd trial in which no significant differences were found. On their right hand right-hand responders only
showed significantly higher SCR magnitudes than left-hand responders in the three first trials on the right hand. Furthermore, left-hand responders showed significant higher SCR magnitudes on their left hand than on their right hand in a!! trials, while right-hand responders showed significantly higher SCR magnitudes on their right hand than on their left hand only in the three first trials. In the analysis of Criterion for Habituation (number of trials to habituation), only the Responsiveness X Hand interaction was significant, F(1,44) = 10.179, MSe = 3.096. Simple main effects analysis pointed out that left-hand responders habituated more quickly on their right hand than on their left hand (F(l,44) = 8.721, MSe = 3.096) while right-hand responders showed no significant differences between hands (means and standard deviations are shown in Table II). In the analysis of SCL only the Responsiveness X Hand interaction was significant, F(1,44) = 31.743, MSe = 0.183. Simple main effects analysis pointed out that right-hand responders showed higher SCLs on their right hand than on their left hand (F(l,44) = 13.11, MSe = 0.183); while lefthand responders showed the opposite result (F(1,44) = 18.916, MSe = 0.183) (means and standard deviations are shown in Table JI). In addition, left-hand responders showed higher SCLs than right-hand responders on the left hand (F( 1,88) = 10.841, MSe = 0.499). while significant differences were not observed on the right hand. The analysis of AP data did not S~IOWany significant effect. AP means were 78.70% (S.D. = 19.70) and 77.27% (S.D. = 27.83) for right-hand dextra! and sinistral responders respectively and 81.04% (S.D. = 21.89) and 55% (S.D. = 27.53) for left-hand dextral and sinistral responders, respectively. The analysis of 1-C data only yielded a significant effect for the Responsiveness factor, F(1.44) = 211.673, MSe = 0.169. Means were 0.835 (S.D. = 0.418) for right-hand responders and - 0.894 (S.D. = 0.411) for left-hand responders. Means of each RP for dextra! and sinistra! subjects in each condition are shown in Fig. 1. No other effects were significant.
DISCUSSION fn relation to differences between dextral and sinistral subjects, these results are in accordance with those obtained by Wyatt and Turskey ( 19691, Gruzelier et al. (19811 and Smith et al. 11981). who did not find significant differences in the orientation of electrodermal asymmetry between dextral and sinistral subjects. In the present experiment no differences were found between these two groups of subjects in all dependent variables studied (LC, AP, SCLs, magnitude of IOR, SCR magnitude, NSR magnitude and NSR frequency). However, because the present expertment did not employ hemisphere-specific tasks. it could be possible that ditferenccs between dextral and sinistral subjects can appear when appropriate tasks are employed, as demonstrated by Lacroix and Comper (1979) or Robinson and Zahn (1981). As expected, the orientation of the elcctrodermal asymmetry in each RP obtained during a stimulus-free period, remained constant throughout the film condition and the neutral tones presentation, independently of subjects’ hand-dominance. The significant Responsiveness effect observed in the analyses of LC data supports this result. This constancy was also observed in all dependent variables studied in which the Responsiveness X Hand interaction was significant. Our results are in accordance with those obtained by different authors who have reported that electrodcrmal Imteralization remains consiant independently of both the task and the dependent variables being employed (De Bonis and Freixa i Baqu6. 19X0;Erwin et al. 1980; Gross and Stern, 1980; Fedora and Schopflochcr, 1984;Freixa i Baqut et al. 1984b and previous results of Roman et al. 1987, 1989). In this way, the orientation of NSRs observed during a stimulus-free period predicts the further electrooermal orientation of both NSRs and SCRs. Although the Responsiveness x Hand interaction was significant in all dependent variables studied, simple main effects analyses showed a more marked difference between hands in lefthand responders than in right-hand responders. At the moment, this result cannot bc explained
although a possible explanation of it could be related to uncontrolled variables. for example cigarette smoking or peripheral factors such as hydration or skin temperature in each hand. The prediction for habituation was only supported partially. Left-hand responders showed slower habituation rate on their left hand than on their right hand, while right-hand responders showed no significant differences between hands although they showed slower rates on their right hand. At the moment, implications of this result are not clear, so further research is recommended in two directions: (a) to explore the difterences between right-hand responders and left-hand responders in habituation of electroderma1 activity; and (b) to explore the relation between RPs and habituation of electrodermal activity when unilateral recording is employed together with other variables (i.e.. gender, personality). In general. the results of the present cxperiment, together with the authors’ previous work (Rom&r et al., 1989) could be interpreted in three ways; (a) orientation of electrodermal asymmetry appears as a subject variable independent on sex and handedness. This subject variable can be labeled RP. It is therefore necessary to know clectrodermal RPs before any other experimental manipulation. In this regard, the selection of experimental samples according to their right or left RP could help to a better knowledge of the relation between electrodermal and hemispheric asymmetry when specific stimu!ation is employed to attempt the activation of a single hcmispherc; (b) LC obtained from a stimulation-free period appear as an useful index to predict further electrodermal orientation or RP; and (cl RP remains constant for each subject and is independent of the tasks when hemisphere specific task or hemisphere input-techniques were not employed, In relation to the constancy of RPs, these experiments showed that the RP remains constant in each subject throughout the experimental session of 22 or 25 min. At the moment, we do not know if this constancy is maintained over 25 min. The theoretical implications of these results are not clear. In our opinion, when electrodermal activity is employed in the study of individual
79
differences RPs appear as an important subject variable which might reflect differences in peripheral mechanisms of electrodermal system. In this regard, the results obtained in both unilateral and bilateral EDA could be contaminated by different proportions of right-hand or left-hand respo ers in any specific experimental sample.
ACKNOWLEDGEMENTS We wish to thank Pilar Martinez-Pelegrin her review of the English translation.
for
REFEYIENCES American Psychoiogicai Association f 1981) Ethica! principles of psjichoiogists. Am. Fsychologist. 36: 633-638. Boyd. G.M. and Maitzman. 1. (1983) Bilateral asymmetry of skin conductance responses during auditory and visual tasks. Psychophysiology, 20: 196-203. Boyd, G.M. and Maitzman, 1. (1984) Effects of cigarette smoking on bilateral skin conductance. Psychophysiology, 21: 334-341.
Curtis, S. (198.5) The development of human cerebral iateraiization. In Benson, F. and Zaidei, E. (Eds.), The &al brain. Hemispheric spe&lization in humans. New York: The Guiiford Press, pp. 97-i 16. De Bonis. M. and Freixa i Baqui, E. (1980) Stress, verbal cognitive activity and bilateral eiectrodermai responses. Neuropsychobiology, 6: 249-259. Edeiberg, R. (1967) Eiectrtcai propertlrs of skin. In C.C. Brown (Ed.). Methods in Psychophysiology. Baltimore: Willians and Wilkins. pp. l-35. Erwin, R.J., McCianahan, B.A., and Kieinman, K.M. (1980) Effects of level of arousal and type of tasks on bilateral skin conductance asymmetry and conjugate lateral eye movements. Par. J. of Biol. Sci., 15: 292-298. Fedora, 0. and Schopflocher. D. (1984) Bilateral eiectrodermai activity during differential cognitive hemispheric activation. Psychophysiology. 21: 307-311. Freixa i Baque, E., Catteau, MC., Miossec, Y. and Roy, J.C. fi984aJ Asymmetry of eiectrodermai activity: A review. Biol. Psychol., 18: 219-239. Freixa i Baque, E. and De Bonis, M. (1983) Eiectrodermai asymmetry during human sleep. Biol. Psycho!, 17: 14.5-151. Freixa i Baque, E.. Somogyi. M. and De Bonis, M. (1984b3) Etude de i’asimetrie des responses iiectroderrnaies d’orientation et de defense. fsychologie MGdicale, 16: 4Y5487.
Gross, J.S. and Stern, J.A. (1980) An investigation of biiaterai
asymmetries in eiectrodermai activity. Par. 1. of Biol.sci., 15: 74-81. Gruzeiier, J., Eves. F. and Contohy, J. (1981) Reciprocal hemispheric influences on response habituation in the eiectrodermai system. Physiol. Psychol., 9: 313-317. Harris, A.J. (1974) Harris Tests of Lateral Dominance. New York: The Psychological Corporation. Spanish adaptation by Ticnicos Especiaiistas Asociados, S.A. (TEA), 1978. Hugdahi, K. (1984) Hemispheric asymmetry and bilateral electrodermal recording: A review of the evidence. prychophysiology, 2 I : 37 l-393. Ketterer, M.W. and Smith, B.D. (1977) Bilateral eiectrodcrmai activity, iateraiized cerebral processing and sex. &chophysioiogy , 14: 5 13-5 16. Kirk, R.E. (1982) Experimental design: Procedures for the Beharioral SLiences. Belmont: Wadsworth, Inc. Lacroix, J.M. and Comper, P. (1979) Lateraiization in the eiectrodermai system as a function of cognitive/ hemispheric manipulation. Psychophysiology, 16: 116-129. Martinez-Seiva. J.M.. Roman, F., Garcia-Sanchez. F.A. and Gomez-Amor, J. (1987) Sex differences and the asymmetry of specific and nonspecific eiectrodermai responses. ht. J. of Psychophysiol.. 5: 155-160.
O’Brien, R.G. and Kaiser, M.K. (1985) MANOVA method for analyzing repeated measures designs: An extensive primer. Psychological Bull., 97: 316-333. Obrist. P.A. (1963) Skin resistance levels and galvanic skin response: Unilateral differences. Science, 3: 252-258. Patterson, J. and Venabies, P.H. (1978) Bilateral skin conductance and skin potential in schizophrenic and normal subjects: The identification of the fast habituator group of schizophrenics. Psychophysiology. 15: 556-560. Robinson, T.N. and Zahn, T.P. (1981) Bilateral EDA, hand preference and hand position (Abstract). Psychophysiology, 18: 137-138.
Romin. F., Martinez-Seiva, J.M., Garcia-Sanchez, F.A. and Gomez-Amor, J. (1987) Sex differences, activation level, and bilateral eiectrodermai activity. Par. J. of Biof. Sci., 22: 113-l 17. Roman, F.. Garcia-Sanchez, F.A., Martinez-Seiva, J.M., Gomez-Amor. J. and Carriiio. E. (1989) Sex differences and bilateral eiectrodermai activity: A replication. Par-. J. of Biol. Sci., 24: 150-155. Smith, B.D. and Ketterer. M.W. (1982) Lateraiized cortical/ cognitive processing and biiarerai clectrodermai activity: Effects of sensory mode and sex. Biol. Psychol., 15: 191201. Smith, L.D., Ketterer, M.W. and Concannon, M. (1981) Biiaterai eiectrodermai activity as a function of hemisphere specific stimulation, hand preference, sex and familial handedness. Biol. Psychol., 12: i-11. Wilkinson. L. (1988) Systat: The system for statistics. Evanston, IL: Systat. Wyatt, R. and Turskey, B. (1969) Skin potential levels in right and left-handed males. Psychophysiology, 6: 133-137.
71
INTPSY 00354
atterns an
ifferences i
Francisco Rom5n, Eduvigis Carrillo and Francisco A. Garcia-Sgnchez Departamento de Ciencias Morfol~gicas, Anatomiu Patokigica y Psicobiologia, Facuitad de Psicologia, Unirersidad de Mwcia, Murcic (Spain)
(Accepted IS April f991)
Key words: Electrodermal asymmetry: Responsiveness pattern; Handedness; Skin conductance; Habituation
The aim of the present study was to examine the relationship between subjects’ handedness and electrodermal asymmetry when subjects were divided according to their electrodermal Responsiveness Patterns, obtained during a stimulus-free period. Skin conductance was bilaterally recorded in a sample of 48 subjects (24 dextrals and 24 sinistrals) durirrg the performance of two tasks (watching a film and listening to neutral tones). Subjects were divided into four groups of 12 subjects each (dextral right-hand responders, dextral left-hand responders, sinistral right-hand responders and sinistral left-hand responders). No differences were found between dextral and sinistral subjects. Orientation of electrodermal asymmetry in each Responsiveness Patterns remained constant throughout the two conditions: right-hand responders showed higher electrodermal activity on their right hand, while left-hand responders showed higher electrodermal activity on their left hand. In addition, left-hand responders showed a slower habituation rate on their left hand than on their right hand, whereas right-hand responders showed no significant differences between hands.
INTRODUCTION Most works on bilateral electrodermal activity (EDA) have assumed that differences in orientation of electrodermal asymmetry could be related to differential hemispheric activation and their conclusions have focussed on this (see Hugdahl, 1984; Freixa i Baque et al., 1984a). Because differences in hemispheric organization can be related to subjects’ handedness (see e.g., Curtis, 19851, it has been considered as an important variable in electrodermal asymmetry researches. However, the majority of works about electrodermal asymmetry either do not indicate the sub-
Currespundenw
F. Roman, Area de Psicobiologia. Facultad de Psicologia, Universidad de Murcia. 30071 Murcia, Spain.
jects’ hand-preference (e.g. Obrist, 1963; Patterson and Venables, 1978; Freixa i Baque and De Bonis, 1983; Freixa i Baque et al., 1984b) or only dextral subjects are selected attempting to control possible influences of handedness in electroderma1 asymmetry (e.g. Erwin et al., 1980; Gross and Stern, 1980; Ketterer and Smith, 1977; Lacroix and Comper, 1979; Smith and Ketterer, 1982; Boyd and Maltzman, 1983, 1984; Martinez-Selva et al., 1987; Roman et al., 1987; Roman et al., 1989). rfowever, a few researchers have studied differences between dextral and sinistrai subjects in bilateral EDA (Wyatt and Turskey, 1969; Robinson and Zahn, 1981; Smith, Ketterer and Concannon, 1981; Gruzelier et al., 1981). Wyatt and Turskey (1969), Gruzelier et al. (1981) and Smith et al. (1981) did not find significant differences in the orientation of electrodermal asymmetry between dextral and sinistral sub-
72
jects. Lacroix and Comper (1979) found different patterns of bilateral electrodermal responses between dextral females when cognitive tasks (verbal and spatial) were employed, but did not find differences in sinistral females who showed always a higher EDA on their left hand independently of the tasks employed. This result was confirmed by Robinson and Zahn (1981) who observed that sinistral subjects showed higher nonspecific and specific skin conductance responses (NSR and SCR, respectively) and skin conductance levels @CL) on their left hand for all conditions presented, while dextral subjects showed higher EDA on their right hand for the most conditions presented. In a previous paper (Roman et al. 1989), the authors considered the possibility that bilateral EDA could show a specific latcralization independently of the tasks being emp!oyed for activating a single hemisphere. In that work, the authors observed that electrodermal lateralization remained constant through different experimental conditions. In addition, gender, which is considered an important variable in the studies of hemispheric specialization, did not appear important in the orientation of EDA when electrodermal responsiveness patterns (RP) and the percentage of subjects with a specific RP were taken into account in each experimental group. Then, the authors indicated the possibility that these RPs could help to explain the results obtained by different authors (e.g., Erwin et al. 1980; De aque, 1980; Gross and Stern, 1980; Fedora and Schopflocher, 1984; MartinezSelva et al. 1987; RornGn et al. 1987), who have reported that the direction of the asymmetry in the EDA remained constant through different experimental conditions. In the same way that the authors had considered the possibility that sex differences in the orientation of electrodermal asymmetry could be due to different percentages of subjects with a right-hand or left-hand dominance in EDA in the male and female groups (RomBn et al. 1989), in the present study we consider the possibility that subjects’ handedness does not affect the orientation of electrodermal asymmetry when subjects arc divided into right-hand and left-hand respon-
ders according to their Laterality Coefficient (LC) during a stimulus-free period. The purpose of the present paper was to study the orientation of bilateral electrodermal asymmetry between dextral and sinistral subjects when the RPs are considered and there are the same proportions of left-hand and right-hand responders in each experimental group. Differences hetween dextral and sinistral subjects in SCRs, NSRs, and SCLs, will be analyzed. Simple tasks, such as presentation of neutral tones and a cartoon film sequence were employed: the first to obtain SCRs and the second NSRs. The employment of a series of neutral tones allowed the measurement of habituation and the analysis of the relationship between habituation and electrodermal asymmetry in right-hand and left-hand responders. According to previous data (Romin et al., 1987, 1989), it was expected each electrodermal RP should remain constant during the two conditions: presentation of neutral tones (SCRs) and cartoon film sequence WSRs). In addition, it was predicted that electrodermal asymmetry observed during the stimulus-free period would indicate orientation of the electrodermal asymmetry of both SCRs and NSRs and, besides, that no differences between dextral and sinistral subjects would be found within each electrodermal RP. In relation to Habituation, and according to previous results (Martinez-Se:va et al. 19871, it was finally predicted that habituation would be faster on the left hand in the right-hand responders group and on the right hand in the left-hand responders group.
METHOD Subjects
A total of 48 first-year psychology undergi-aduates at the University of Murcia (24 dextral and 24 sinistral subjects) volunteered for the experiment. The age range was 17 to 29 years (average age for dextral subjects = 22.6 years, S.D. = 2.60; average age for sinistral subjects = 21.4, S.D. = 1.81). Subjects were divided into right-hand and left-hand responders depending on their LC, ob-
73 TABLE 1 Vulaes of the Laterality Coefficient in the stimulus-free period Right-hand responders Left-hand responders mean
Dextrals Sinistrals
0.86 - 0.95
S.D.
mean
S.D.
0.21
0.92 - 0.99
0.20 0.03
0.13
tained from an initial stimulus-free period (see Response identification and quantification section below for definition), giving 12 subjects in each group that is, dextrals right-hand responders, dextrals left-hand responders, sinistrals right-hand responders and sinistrals left-hand responders. Once completed, the experimental groups subjects were assigned to other investigations. LC means and standard deviations of these groups in the stimulus-free period are shown in Table 1. Lateral dominance was assessed using the Spanish adaptation of Harris’ tests of lateral dominance (Harris, 1974). The criterion for identifying dextra! subjects was that they had to obtain 100% of responses indicating dextra! dominance and the criterion for identifying sinistral subjects was that they had to obtain at least 90% of responses indicating sinistra! dominance, a!ways employing their left hand for writing and the left foot for kicking a ball. This criterion was established due to the difficulty of finding large samples of sinistra! subjects in the Spanish population. Ambidextrous subjects (n = 7) were excluded from statistical analyses, along with four other subjects who did not give any electroderma! response in the initial stimulus-free period. Two additional subjects were dropped because of a technical problem in the recording. Subjects were replaced to complete the sample of 48 subjects who participated in the experiment. Subjects received course credits for participation. All par ticipants were treated in accordance with the ‘Ethical Principles of Psychologists’ (American Psychological Association, 198 I). Apparatus
Auditory stimu!! (neutral tones) were produced by a sine wave generator, Letica Stimula-
tor 150 (rise and decay time < 1 ms). Stimuli were delivered via Audiotechnica ATH-2,8 ohms, stereoheadphones, calibrated at 80 dB in the output. A tape recorder, TEAC-A-3440 (TEAC Co., Tokyo, Japan), was used to record and to deliver the instructions. The film sequence was recorded and presented by a videocassette rccorder JVC (JVC Victor Company of Japan Limited), VHS system, on a television set, placed 1.5 m in front of the subjects. Bilateral skin conductance was recorded on a four-channel Net San Ei oscillograph (Net San-Ei Instruments, Ltd. Tokyo, Japan), through two amplifiers with Coulbourn S71-22 skin conductance couplers which provided 0.5 V (DC) constant voltage excitation (Coulbourn Instruments, Lehight Valley, PA, USA). Because of their size (7 mm diameter Ag/AgCl, Couiboum Tll-01) electrodes were attached by double faced adhesive collars to the thenar (C6) and hypothenar (03) eminences of both hands. A bioge! (LA 76621, 0.068 M NaC!) was used as the contact medium. The skin was cleansed with soap and water. The electrodes, amplifiers and pens were alternated between hands from subject to subject to eliminate the possibility of systematic influences. The chart speed was 5 mm/s. Procedure
Subjects were seated in a comfortable armchair in a darkened, sound-attenuated chamber (2.5 X 2 X 2.40 m). The recording apparatus was located in an adjoining room. An intercom rystem allowed communication with the subject, who could be seen ;hrough a small glass window in one of the chamber walls. The temperature in this chamber was held constant 235°C (k 1.5”C). After electrodes and headphones had been attached by one female experimenter, general instmctions were given through a loudspeaker placed in the chamber, 1.5 m in front of the subjects and a 5 min adaptation period was observed before the stimulus-free period which lasted 3 min. Subjects were told that they would hear a series of harmless tones and watch a film sequence and that the experimenter was interested in recording their physiological responses to those conditions. They were also told to remain
74
quiet and to avoid movements. Immediately before each condition, subjects were informed of its beginning. After the stimulus-free period the subjects received the neutral tones and the cartoon film sequence which were counterbalanced in the order of presentation for all the subjects. An interval of 2 min was observed between the conditions. The duration of the film condition was 2 min and consisted of a sequence of a cartoon movie for children. Auditory stimuli consisted of 12 tones of 2 s (80 dB and 1000 Hz). Interstimulus intentals varied randomly between 15 and 45 s, with an average duration of 30 s. The recording session had a total duration of 22 min, approximately. Immediately after the recording session, subjects completed the handedness inventory. Response identification and quantification Polygraph recordings were scored by hand. Compound responses were scored from an inflection point to peak, as recommended by Edelberg (1967). Any artifact-free change in skin conductance higher than 0.05 pmhos was considered an electrodermal response. SCLs were taken immediately before and after the stimulus-free period and the film condition, and seven-times during the neutral tones presentation: before the first tone, and after stimuli number 2, 4, 6, 8, 10 and 12. A SCR was the first response taking piace l-5 s after stimulus onset. The criterion for habituation was three consecutive trials with no s were scored during the initial 3 min stimulus-free period and during the 2 min film condition. NSRs dur’ng the initial stimulus-free period were only used to determine the RP of each subject. The LC proposed by Freixa i BaquC and De Bonis (19831 was employed to obtain electrodermal lateraiity: %R>L-%L>R %R>Lc%L>R %R > L reflected the percentage of cases in which the right hand showed an electrodermal response of a higher magnitude than the accompanying response on the left hand, and %L > R reflected the percentage of cases in which the
contrary occurred. Differences in magnitude between simultaneous electrodermal responses in each hand below 0.05 pmhos (differentiai response criterion) were not considered and R =zL decision was made in these cases. The observation of a response on one hand but not on the other hanll resulted in an amplitude of 0 pmhos being scored for the latter. The value of LC was zero when %R > L = %L > R, positive when the asymmetrical electrodermal responses were of higher magnitude on the right hand (R > LJ, and negative when the asymmetrical electrodermal responses were of higher magnitude on the left hand (L > R). LC range from - 1 to + 1. RP in each subject was obtained according to their LC scores: When the value was positive then the subject was considered a right-hand responder; when negative, a left-hand responder. The denominator of this formula was employed to obtain the asymmetry proportion (API, which was defined as the total sum of all the percentages of asymmetrical electroderma1 responses in both hands (%cR > f. + %L > R). The response frequency was the number of NSRs per minute. In the response magnitude analysis, all instances in which a phasic response of measurable magnitud: (define’d as 0.05 pmhos or higher) occurred in either hand were included. In this analysis, both asymmetrical responses (R > L and L > RI and symmetrical responses (R = L) were included. Again, the observation of a response from one hand but not from the other resulted in an amplitude of 0 pmhos being scored for the latter. Data analysis Four independent two-way 2 x 2 ANOVAS were employed to analyse AP and LC data, during neural tones presentation and film conditions, rGJpectively, with Handedness and Responsiveness as between-subjects factors. Magnitude and frequency of NSRs during the film condition, the SCR magnitude data of the initial orienting response (l@R) to the first trial of the neutral tones presentation, and the habituation criterion (number of trials to habituation) were analyzed by four mixed analyses of variance
7s TABLE
II
Means and standard deviations fin parentheses) of the Responsivenesx Hand interaction an the variables analyzed
Right-hand responders
Left-hand resporrders
RH
LH
RH
LH
0.198 (0.075) 5.938 (2.585) 0.599 (0.375)
0.136 to.0791 5.458 (2.734) 0.423 (0.306)
0.092 (0.070) 4.854 (3.140) 0.385 (0.272)
0.184 (0.086) 7.208 (3.501) 0.636 (0.438)
0.933 (0.586) 0.345
0.676 (0.528) 0.219
0.464 CO.5861 C.275
0.957 (I ,002) 0.553
(0.441)
(0 291)
(0.427)
(0.764)
7.542
6.750
6.917
8,417
(4.880)
(4.413)
Film condition
Fdanspecificresponsesmagnitude tpmhos) Nonspecific responses frequency (resp/min) Skin conductance levels tpmhos) Tones condition Initial orienting response (r.rmhast Specific responsesmagnitude (pmhos) Criterion
for habituation
(trials to criterion) Skin conductance
(3.149) levels
(pmhos)
(4.71
I)
0.480
0.311
0.362
0.565
(0.395)
co.im,
(0.282)
(0.350)
RH. right hand: LH. left hand.
(2 X 2 X 2) Handedness X Responsiveness X Hand, with repeated measures in the last factor. SCLs during the film condition were analyzed by a mixed analysis of variance (2 x 2 x 2 X 2) Handedness x Responsiveness x Recording time X Hand, with repeated measures in the two last factors. There were two levels of recording time: pre and post film presentation. A 2 x 2 x 12 x 2 univariate analysis of variance (ANOVA), Handedness X Responsiveness x Trials x Hand, with Trials and Hand as within-subject factors, was performed for magnitude of SCRs to the 12 tones of the neutral tones presentation. On the other hand, SCLs data during this-condition were analyzed by a mixed anaiysis of variance (2 X 2 x 7 >: 2) Handedness X Responsiveness x Recording time x Hand, with repeated measures in the two last factors. In these two analyses a multivariate analysis of variance (MANOVA) on the within-subjects effects were run (O’Brien and Kaiser, 1985). A variable of the within-subjects portion of these two anaiyses was only considered significant when the Piilais trace multivariate statistic showed it (Wiikinson, 1988).
In ail first-order interactions simple main effects analyses were run, b:ing the mea13 qxare error corrected for all simple effects of withinsubjects factors following Kirk’s computational procedures (Kirk, 1982). The 0.05 rejection was adopted for ail statistical analyses.
RESULTS Film condition In the analysis of NSR magnitude data, only the Responsiveness x Hand interaction was significant, F(l,44) = 73.233, MSc = 0.002. Simple main effects analysis demonstrated that righthand responders showed higher NSR magnitudes on tneir right hand than on their left one, while left-hand responders showed the oppositt result, F(1,44) = 23.06, MSe = 0.002 and F(1,44) = 50.78, MSe = 0.002. respectively (means and standard deviations are shown in Table II). In addition, simple main effects showed that on the right hand right-hand responders showed significantly higher NSR magnitudes than left-hand responders (F(1,44) = 22.5, MSe = 0.0061, while signifi-
Dewtrals Rtght-nand
Dextrols
Slnl5trt7lS
Left
responders
Fig.1. Mean laterality coefficient for dentrnl and sinistral
right-h;md
-hand
Smlstrals responders
and left-hand respc)ndcrsduring film condition and IWPS
prrsrntation.
cant diffcrenccs bctwecn groups wcrc not observed on the left hand. The analysis of NSR frequency data showed significant effects in both the Hand factor and the Responsiveness x Hand interaction. Hand effect, F(1.44) = 7.674. MSc = 2.74’1. was due to higher frequrncy of NSRs on the left hand (mean = 6.333 rcsp,‘min. S.D. = 3.231) than on the right hanlT (mean = 5.396 rcsp/min. SD. = 2.809). In regard to the Rcsponsivcncss x Hand interaction. F( 1.44) = 17.534, MSe = 2.749, simple main cffccts analysis pointed out that left-hand responders showed more freqt.,:ncy of NSRs on their left hand than on their right hand, F(1.44) = 24.189. MSe = 2.74’1.while right-hand responders did not show significant differences between hands (means and standard deviations are shown in Table II). In the analysis of SCL data, only the Responsiveness x Hand interaction was significant. F( 1,44)= 25.21 I. MSe = 0.087. Simple main cffacts analysis pointed out that left-hand responders showed higher SCLs on their left hand than on their right hand, while right-hand responders did not show significant differences hctwecn hands, F(1.44) = 8.69, MSe = 0.087 (means and standard deviations arc shown in Table II).
The analysis of AP data did not show any significant effect. AP rnpans were 58.30% (S.D. = 27.87) and K0.05% (SD. = 19.89) for right-hand dextral and sinistral responders. respectively, and 73.26% (SD. = 30.57) and 77.84% (S.D. = 28.06) for left-hand dcxtral and sink&al responders. respectively. The analysis of LC data yielded a significant effect for the Responsiveness factor. F(1.44) = 68.607. MSc = 0.358. Means were 0.652 (S.D. = 0.6411) for right-hand responders and - 0.779 (S.D. = 0.577) for left-hand responders. Means of each RP for dcxtral and sinistral subjects in each condition are shown in Fig. 1. No other effects wefe significant.
The analysis of 10R magnitude to the first trial of neutral tones presentation only found a significant Responsiveness x Hand interaction. F( 1.44) = 32.454. MSe = 0.104. Simple main effects analysis showed that IOR was higher in right-hand responders on their right hand than on their left hand (F( 1,44)= 7.625, MSe = 0.104); while left-hand responders showed the opposite result (F(1.44) = 28.05. MSe = 0.104) (means and standard deviations are shown in Table II). In
75
addition, simple main effects showed a significant difference between RPs on the right hand (F(1,88) = 5.397, MSe = 0.489) due to higher IOR magnitudes in the group of right-hand responders, while significant differences were not obeen right-hand and left-hand responIn the analysis of SCR magnitude data both the Trials and the Hand factors were significant. In the Trials effect, F(11,484) = 12.635, MSe = 0.21, a trend analysis showed significant linear (F(1.44) = 38.014, MSe = 0.533), quadratic (F(1.44) = 19.730, MSe = 0.180). and cubic trends (F(1.44) = 13.824. MSe = 0.251). Higher order trends did not reach the significant level. The Hand effect (F( 1.44) = 4.116, MSe = 0.40s). was due to higher response magnitude on the left hand (mean = 3.863 pmhos, S.D. = 0.603) than on the right hand (mean = 3.099 pmhos, S.D. = 0.445). The other main effects, Handedness and Responsiveness, did not approach the significance level. Among the first-order interactions of this analysis, only the Responsiveness X Hand interaction was significant, F( 1,44) = 28.772, MSe = 0.408. Simple main effects analysis pointed out that right-hand responders showed higher magnitude of SCRs on their right hand than on their left hand (F(1.44) = 5.515. MSe = 0.408), while lefthand responders showed the opposite result (F( 1.44) = 27.277. MSe = 0.408) (means and standard deviations are shown in Table 11). Moreover, simple main effects analysis pointed out higher magnitude of SCRs on the left hand in left-hand responders than in right-hand responders (F(1.88) = 7.996, MSe = 1.997), while on the right hand differences between RPs were not significant (F < !). Among the second-order interactions, only, the Responsiveness x Trials x Hand interaction was significant, F(11,484) = 4.271, MSe = 0.033, P < 0.001. A multiple comparison Scheffe test pointed out that on the left hand left-hand responders showed higher SCR magnitudes than right-hand responders in all trials, except in the 2nd trial in which no significant differences were found. On their right hand right-hand responders only
showed significantly higher SCR magnitudes than left-hand responders in the three first trials on the right hand. Furthermore, left-hand responders showed significant higher SCR magnitudes on their left hand than on their right hand in a!! trials, while right-hand responders showed significantly higher SCR magnitudes on their right hand than on their left hand only in the three first trials. In the analysis of Criterion for Habituation (number of trials to habituation), only the Responsiveness X Hand interaction was significant, F(1,44) = 10.179, MSe = 3.096. Simple main effects analysis pointed out that left-hand responders habituated more quickly on their right hand than on their left hand (F(l,44) = 8.721, MSe = 3.096) while right-hand responders showed no significant differences between hands (means and standard deviations are shown in Table II). In the analysis of SCL only the Responsiveness X Hand interaction was significant, F(1,44) = 31.743, MSe = 0.183. Simple main effects analysis pointed out that right-hand responders showed higher SCLs on their right hand than on their left hand (F(l,44) = 13.11, MSe = 0.183); while lefthand responders showed the opposite result (F(1,44) = 18.916, MSe = 0.183) (means and standard deviations are shown in Table JI). In addition, left-hand responders showed higher SCLs than right-hand responders on the left hand (F( 1,88) = 10.841, MSe = 0.499). while significant differences were not observed on the right hand. The analysis of AP data did not S~IOWany significant effect. AP means were 78.70% (S.D. = 19.70) and 77.27% (S.D. = 27.83) for right-hand dextra! and sinistral responders respectively and 81.04% (S.D. = 21.89) and 55% (S.D. = 27.53) for left-hand dextral and sinistral responders, respectively. The analysis of 1-C data only yielded a significant effect for the Responsiveness factor, F(1.44) = 211.673, MSe = 0.169. Means were 0.835 (S.D. = 0.418) for right-hand responders and - 0.894 (S.D. = 0.411) for left-hand responders. Means of each RP for dextra! and sinistra! subjects in each condition are shown in Fig. 1. No other effects were significant.
DISCUSSION fn relation to differences between dextral and sinistral subjects, these results are in accordance with those obtained by Wyatt and Turskey ( 19691, Gruzelier et al. (19811 and Smith et al. 11981). who did not find significant differences in the orientation of electrodermal asymmetry between dextral and sinistral subjects. In the present experiment no differences were found between these two groups of subjects in all dependent variables studied (LC, AP, SCLs, magnitude of IOR, SCR magnitude, NSR magnitude and NSR frequency). However, because the present expertment did not employ hemisphere-specific tasks. it could be possible that ditferenccs between dextral and sinistral subjects can appear when appropriate tasks are employed, as demonstrated by Lacroix and Comper (1979) or Robinson and Zahn (1981). As expected, the orientation of the elcctrodermal asymmetry in each RP obtained during a stimulus-free period, remained constant throughout the film condition and the neutral tones presentation, independently of subjects’ hand-dominance. The significant Responsiveness effect observed in the analyses of LC data supports this result. This constancy was also observed in all dependent variables studied in which the Responsiveness X Hand interaction was significant. Our results are in accordance with those obtained by different authors who have reported that electrodcrmal Imteralization remains consiant independently of both the task and the dependent variables being employed (De Bonis and Freixa i Baqu6. 19X0;Erwin et al. 1980; Gross and Stern, 1980; Fedora and Schopflochcr, 1984;Freixa i Baqut et al. 1984b and previous results of Roman et al. 1987, 1989). In this way, the orientation of NSRs observed during a stimulus-free period predicts the further electrooermal orientation of both NSRs and SCRs. Although the Responsiveness x Hand interaction was significant in all dependent variables studied, simple main effects analyses showed a more marked difference between hands in lefthand responders than in right-hand responders. At the moment, this result cannot bc explained
although a possible explanation of it could be related to uncontrolled variables. for example cigarette smoking or peripheral factors such as hydration or skin temperature in each hand. The prediction for habituation was only supported partially. Left-hand responders showed slower habituation rate on their left hand than on their right hand, while right-hand responders showed no significant differences between hands although they showed slower rates on their right hand. At the moment, implications of this result are not clear, so further research is recommended in two directions: (a) to explore the difterences between right-hand responders and left-hand responders in habituation of electroderma1 activity; and (b) to explore the relation between RPs and habituation of electrodermal activity when unilateral recording is employed together with other variables (i.e.. gender, personality). In general. the results of the present cxperiment, together with the authors’ previous work (Rom&r et al., 1989) could be interpreted in three ways; (a) orientation of electrodermal asymmetry appears as a subject variable independent on sex and handedness. This subject variable can be labeled RP. It is therefore necessary to know clectrodermal RPs before any other experimental manipulation. In this regard, the selection of experimental samples according to their right or left RP could help to a better knowledge of the relation between electrodermal and hemispheric asymmetry when specific stimu!ation is employed to attempt the activation of a single hcmispherc; (b) LC obtained from a stimulation-free period appear as an useful index to predict further electrodermal orientation or RP; and (cl RP remains constant for each subject and is independent of the tasks when hemisphere specific task or hemisphere input-techniques were not employed, In relation to the constancy of RPs, these experiments showed that the RP remains constant in each subject throughout the experimental session of 22 or 25 min. At the moment, we do not know if this constancy is maintained over 25 min. The theoretical implications of these results are not clear. In our opinion, when electrodermal activity is employed in the study of individual
79
differences RPs appear as an important subject variable which might reflect differences in peripheral mechanisms of electrodermal system. In this regard, the results obtained in both unilateral and bilateral EDA could be contaminated by different proportions of right-hand or left-hand respo ers in any specific experimental sample.
ACKNOWLEDGEMENTS We wish to thank Pilar Martinez-Pelegrin her review of the English translation.
for
REFEYIENCES American Psychoiogicai Association f 1981) Ethica! principles of psjichoiogists. Am. Fsychologist. 36: 633-638. Boyd. G.M. and Maitzman. 1. (1983) Bilateral asymmetry of skin conductance responses during auditory and visual tasks. Psychophysiology, 20: 196-203. Boyd, G.M. and Maitzman, 1. (1984) Effects of cigarette smoking on bilateral skin conductance. Psychophysiology, 21: 334-341.
Curtis, S. (198.5) The development of human cerebral iateraiization. In Benson, F. and Zaidei, E. (Eds.), The &al brain. Hemispheric spe&lization in humans. New York: The Guiiford Press, pp. 97-i 16. De Bonis. M. and Freixa i Baqui, E. (1980) Stress, verbal cognitive activity and bilateral eiectrodermai responses. Neuropsychobiology, 6: 249-259. Edeiberg, R. (1967) Eiectrtcai propertlrs of skin. In C.C. Brown (Ed.). Methods in Psychophysiology. Baltimore: Willians and Wilkins. pp. l-35. Erwin, R.J., McCianahan, B.A., and Kieinman, K.M. (1980) Effects of level of arousal and type of tasks on bilateral skin conductance asymmetry and conjugate lateral eye movements. Par. J. of Biol. Sci., 15: 292-298. Fedora, 0. and Schopflocher. D. (1984) Bilateral eiectrodermai activity during differential cognitive hemispheric activation. Psychophysiology. 21: 307-311. Freixa i Baque, E., Catteau, MC., Miossec, Y. and Roy, J.C. fi984aJ Asymmetry of eiectrodermai activity: A review. Biol. Psychol., 18: 219-239. Freixa i Baque, E. and De Bonis, M. (1983) Eiectrodermai asymmetry during human sleep. Biol. Psycho!, 17: 14.5-151. Freixa i Baque, E.. Somogyi. M. and De Bonis, M. (1984b3) Etude de i’asimetrie des responses iiectroderrnaies d’orientation et de defense. fsychologie MGdicale, 16: 4Y5487.
Gross, J.S. and Stern, J.A. (1980) An investigation of biiaterai
asymmetries in eiectrodermai activity. Par. 1. of Biol.sci., 15: 74-81. Gruzeiier, J., Eves. F. and Contohy, J. (1981) Reciprocal hemispheric influences on response habituation in the eiectrodermai system. Physiol. Psychol., 9: 313-317. Harris, A.J. (1974) Harris Tests of Lateral Dominance. New York: The Psychological Corporation. Spanish adaptation by Ticnicos Especiaiistas Asociados, S.A. (TEA), 1978. Hugdahi, K. (1984) Hemispheric asymmetry and bilateral electrodermal recording: A review of the evidence. prychophysiology, 2 I : 37 l-393. Ketterer, M.W. and Smith, B.D. (1977) Bilateral eiectrodcrmai activity, iateraiized cerebral processing and sex. &chophysioiogy , 14: 5 13-5 16. Kirk, R.E. (1982) Experimental design: Procedures for the Beharioral SLiences. Belmont: Wadsworth, Inc. Lacroix, J.M. and Comper, P. (1979) Lateraiization in the eiectrodermai system as a function of cognitive/ hemispheric manipulation. Psychophysiology, 16: 116-129. Martinez-Seiva. J.M.. Roman, F., Garcia-Sanchez. F.A. and Gomez-Amor, J. (1987) Sex differences and the asymmetry of specific and nonspecific eiectrodermai responses. ht. J. of Psychophysiol.. 5: 155-160.
O’Brien, R.G. and Kaiser, M.K. (1985) MANOVA method for analyzing repeated measures designs: An extensive primer. Psychological Bull., 97: 316-333. Obrist. P.A. (1963) Skin resistance levels and galvanic skin response: Unilateral differences. Science, 3: 252-258. Patterson, J. and Venabies, P.H. (1978) Bilateral skin conductance and skin potential in schizophrenic and normal subjects: The identification of the fast habituator group of schizophrenics. Psychophysiology. 15: 556-560. Robinson, T.N. and Zahn, T.P. (1981) Bilateral EDA, hand preference and hand position (Abstract). Psychophysiology, 18: 137-138.
Romin. F., Martinez-Seiva, J.M., Garcia-Sanchez, F.A. and Gomez-Amor, J. (1987) Sex differences, activation level, and bilateral eiectrodermai activity. Par. J. of Biof. Sci., 22: 113-l 17. Roman, F.. Garcia-Sanchez, F.A., Martinez-Seiva, J.M., Gomez-Amor. J. and Carriiio. E. (1989) Sex differences and bilateral eiectrodermai activity: A replication. Par-. J. of Biol. Sci., 24: 150-155. Smith, B.D. and Ketterer. M.W. (1982) Lateraiized cortical/ cognitive processing and biiarerai clectrodermai activity: Effects of sensory mode and sex. Biol. Psychol., 15: 191201. Smith, L.D., Ketterer, M.W. and Concannon, M. (1981) Biiaterai eiectrodermai activity as a function of hemisphere specific stimulation, hand preference, sex and familial handedness. Biol. Psychol., 12: i-11. Wilkinson. L. (1988) Systat: The system for statistics. Evanston, IL: Systat. Wyatt, R. and Turskey, B. (1969) Skin potential levels in right and left-handed males. Psychophysiology, 6: 133-137.
