The Effects of Stimuli That Vary in Erotic Content on Cognitive Processes.

The present study is a replication and expansion of a previous study by Wright and Adams (1994) to develop a procedure to assess sexual interest that utilizes choice reaction time (CRT) as the dependent variable. The CRT procedure was based largely on a model of sexual arousal proposed by Singer (1984), which suggested the possibility of assessing sexual interest by measuring the orienting response of an individual to preferred-gender stimuli. This methodology is also supported by the work of Kohlers (1976), who reported that viewing time is greater for material that is motivating and interesting.

The CRT procedure in the pilot project by Wright and Adams (1994) recorded reaction time latencies to three types of stimuli: nude men, nude women, and neutral scenes. Eighty participants, 20 from each of four groups--heterosexual men, heterosexual women, gay men, and lesbian women--completed the CRT task. Please refer to the pilot study for the specifics on procedures and methodology. The study yielded the following results: 76 of 80, or 95% of the participants, scored as predicted on the CRT task, indicating they had longer latencies to preferred-gender stimuli as compared to nonpreferred-gender and neutral stimuli. Upon completion of the pilot study, a few variables that were not adequately controlled became of interest, and these variables serve as the basis for this project. Sexual motivation, stimuli that vary in erotic content, and incidental learning are the prime variables of interest in this study.

Sexual interest is assumed to have motivational properties--it leads to sexual arousal, which can affect one's mood state. If this is so then sexual interest, like depression and anxiety, when elicited, should influence one's ability to process information. It has been demonstrated, for example, that both anxiety and depression decrease performance on a cognitive task such as the Stroop Color-Naming Task (Fox, 1993; Segal, 1995). The Stroop Color-Naming Task is a complex task of attention and interference involving compound stimulus (MacLeod, 1991). Consequently, it could be hypothesized that if sexual interest was elicited during a similar discrimination task, similar interference effects should occur. This hypothesis received support in the pilot project by Wright and Adams (1994) which demonstrated that slides of nude adults caused interference and increased reaction time during a choice discrimination task. Furthermore, the interference was only present if the sexual stimuli were congruent with the individual's sexual orientation. The fact that each group demonstrated longer latencies on preferred-gender stimuli indicates that the interference only occurred if sexual interest was present. These results support a relationship between the motivational aspects of sexual interest and cognitive processes. In addition, it is possible that these results provide a basis for measurement of sexual interest and sexual orientation.

The present study was designed to further explore the motivational and interference effects of sexual interest on cognitive functioning. While the original study demonstrated an interference effect, it ignored the motivational aspects of sexual arousal. In other words, stimuli which elicit greater interest should have a larger influence on one's ability to process information as demonstrated by Pallak, Pittman, Heller, and Munson (1975) using the Stroop procedure.

The present study was designed to replicate the original study by Wright and Adams (1994) using more rigorous controls, and to demonstrate that the degree of sexual interest is of importance in this procedure. According to the trichotomy of sexual arousal theory put forth by Singer (1984), if a stimulus or an aspect of the environment initiates the sexual response, then the individual should attempt to maintain contact with that stimulus. And, while attempting to maintain contact with the stimulus, the individual will be distracted and will experience decreased performance on a task requiring attention and concentration. Thus, the first hypothesis of this study is that individuals will have longer contact or reaction time with stimuli which are sexually appealing as originally shown by Wright and Adams (1994). The second hypothesis is that participants will experience greater sexual interest and, thus, have longer latencies to slides that are more provocative (i.e., stimuli of nude vs. clothed individuals). The third hypothesis is that if slides of nude preferred-gender stimuli elicit sexual interest and interfere with cognitive processes by directing attention to the more salient background image of the stimuli that is competing for attention, then memory for these stimuli should be greater during a recognition task (Kintsch, 1974; Simon, 1994). Thus, incidental learning should occur and recognition should be greatest for slides of nude preferred-gender stimuli during an incidental learning task. This study was designed to evaluate these hypotheses.

METHOD

Subjects

A total of 80 participants completed this study. Twenty subjects comprised each of the following four groups: heterosexual men, heterosexual women, gay men, and lesbian women. Participants were selected from the University of Georgia Psychology Department Research Pool and residents of the Athens-area community. Fewer than necessary gay and lesbian individuals from the research pool volunteered, so an advertisement was placed in a campus-wide newspaper to assist in recruitment. Participants were screened during large group testing sessions at which time they completed a consent form, a short demographic questionnaire, and a modified version of the Kinsey Heterosexual-Homosexual Rating Scale (Kinsey, Pomeroy, & Martin, 1948). Many of the gay and lesbian individuals had engaged in both heterosexual and homosexual behaviors, which made classification difficult using the Kinsey scale. In order to correctly categorize individuals who admitted both heterosexual and homosexual experiences, participants were asked to verbalize that they self-identified as gay or lesbian. If an individual reported being bisexual, they were excluded from participating. Participants were not screened for other sexual experiences, such as upsetting or traumatic sexual experiences.

The participants who met criteria for the study were contacted by telephone at a later date to schedule individual appointments to complete the laboratory portion of the experiment. Participants from the Research Pool received partial course credit. Participants from the community were paid $10.00.

Response Measures

The stimuli were counterbalanced so that one half of each group completed the nude trials first and one half completed the clothed trials first. The incidental learning task was presented after the completion of both sets of experimental trials.

Choice Reaction Time Task

This task consisted of two series of 60 slides: one series contained nude stimuli and one series contained clothed stimuli. There were 20 slides from each of the following three categories in each series: men, women, and neutral slides. The slides used in this study were photographed by the investigator. The slides of nude individuals were made from 8-in. by 10-in. pictures that were obtained from commercial sexually-explicit magazines. The stimuli of clothed individuals were photographed from popular clothing and fashion magazines. The neutral slides consisted of a single solid-colored (blue) background, purposely devoid of content, in order to measure pure reaction time.

Each slide had a single white dot superimposed in 1 of 5 locations, the four corners or the center. At a distance of fifteen feet the white dot appeared 1 1/2 inches in diameter and could easily be located. Each participant was instructed to locate the white dot as quickly as possible and to record his or her selection by pressing a button on a hand-held box that corresponded with the position of the dot on the slide. The placement of the dots was counterbalanced so that a total of twelve slides, four from each of the three categories, had dots in each of the five locations. The slides were arranged in a Kodak XL 100 projector carousel in random order; randomization was determined by a computer-generated table of random numbers. The slides were positioned in every other slot in the carousel to provide a blank space between stimuli presentations. The empty slot served as the intertrial interval.

The hand-held button-box, which the participant used to record his or her dot placement selections, was connected to an IBM micro computer via a serial port interface. The box was constructed from a black plastic video cassette case and contained six momentary buttons. A square figure was scribed in the top portion of the button-box to replicate the shape of the slide. Five of the buttons were located in the square figure in the same positions as the dots on the slides, one in each of the four corners and the center. A sixth button was located in the center of the bottom of the box and was used to advance the slide. A computer program was written to integrate the computer, the slide projector, and the hand-held box. The investigator monitored the experiment from a control room that was separate from the experimental chamber where the subject sat. The experimental trials began when the experimenter started the computer program, which advanced the first slide of the series and started the timer. Upon receiving a selection response from the participant via the hand-held box, the computer stopped the timer, recorded whether or not the selection was correct, and recorded the choice reaction time for the trial. When the subject pressed the sixth button located in the bottom center of the hand-held box, the computer recorded the delay time for that trial and advanced the slide.(1) A 3-second intertrial interval separated each experimental trial, during which the projector lamp illuminated the screen. At the end of the intertrial interval the computer advanced the slide projector for the next trial. This process continued until all 60 slides in that series were concluded. At the end of each series of slides the computer program sorted the responses into the three categories (men, women, and neutral), totaled the correct responses, and calculated the means and standard deviations for the CRT for each category. Then the next series (clothed/nude) of slides was administered and the same sequence repeated. Data from the incorrect responses were excluded from the analysis of response times. Five practice trials consisting of neutral slides preceded the first series of experimental trials only to accustom the participant to the procedure.

Incidental Learning Task

This part of the experiment consisted of a series of 120 slides, 60 men and 60 women, one half clothed and one half nude. One half of the slides were identical to those seen during the experimental trials minus the white dot, and one half were novel but similar to those previously seen. The subjects were instructed to attempt to recall whether or not the slide had been presented during the experimental trials. The subject was instructed to record his or her selection, previously seen or novel, by pressing the lower left button on the button-box for previously seen stimuli or by pressing the lower right button on the button-box for novel stimuli. There was no intertrial interval during this part of the experiment to separate the slides. Upon receiving a selection response from the subject, the computer advanced the slide and scored the response as correct or incorrect. At the conclusion of the 120 trials, the computer sorted and tallied the correct and incorrect responses.

Procedure

The participant was greeted upon arrival, seated in the experimental chamber, and given instructions for part one of the experiment. The specific instructions that were provided to each subject are printed in Appendix A. The experimenter then left the chamber and closed the door that separated the two rooms. Verbal communication was possible between the experimenter and participant without the use of an intercom. Upon receiving a verbal signal from the participant that she or he was ready to begin, the experimenter started the computer program for the five practice trials. Upon completion of the practice trials, the experimenter entered the experimental chamber to answer any questions. The experimenter then exited the chamber and awaited a verbal signal from the participant that she or he was ready to begin part one of the study; the experimenter then initiated the computer program for the first series of slides.

After completion of part one of the study, the experimenter positioned the second series of slides, provided instructions for part two of the study, and answered any questions. The specific instructions for part two of the study are provided in Appendix A.

Upon completion of part two of the study, the experimenter positioned the memory trial slides on the projector for part three of the study, the incidental learning task. The specific instructions provided to each subject are provided in Appendix A. The experimenter then left the chamber and awaited a verbal response from the participant to begin part three of the study. Upon completion of part three, the subject was debriefed and questions were answered prior to his or her departure.

RESULTS

The data for the CRT task were analyzed using raw data, logarithmic transformation scores, and change scores. Since the results of the transformed scores and the change scores were similar to the raw scores, only the raw data analyses are reported.

Reliability Analyses

In order to determine the reliability of the stimuli, the 20 trials were split into even and odd halves and Pearson correlations coefficients with Spearman-Brown corrections were computed. The reliability coefficient was .903 for nude women, .954 for nude men, .924 for neutral slides shown during the nude stimuli, .941 for clothed women, .889 for clothed men, and .922 for neutral slides shown during the clothed stimuli. The reliability coefficients indicate that the stimuli have high reliability.

Choice Reaction Time Analyses

The choice reaction time (CRT) data for the nude and clothed stimuli were analyzed separately. The CRT data for the nude stimuli were analyzed using ANOVA with two between factors (orientation x gender) and one repeated factor (stimulus type). A main effect for stimulus type was significant, F(2, 152) = 54.02, p [is less than] .001, indicating a faster reaction time to neutral stimuli than nude men stimuli or nude women stimuli. The interactions between gender x stimulus type and orientation x stimulus type were significant; F(2, 152) = 3.32, p [is less than] .05 and F(2, 152) = 5.85, p [is less than] .01, respectively. These results were qualified by a significant triple order interaction (gender x orientation x stimulus type), F(2, 152) = 38.87, p [is less than] .001. An analysis of the simple effects using the Tukey HSD test, shown in Table 1, revealed that on the slides of men, the gay men demonstrated significantly longer reaction times than the lesbian women. On the slides of nude women, the heterosexual men had significantly longer latencies than the heterosexual women and the gay men but did not differ from the lesbian women. There were no other significant differences between groups with neutral stimuli. The within-subject effects were analyzed using multiple t-tests. Results indicate that the heterosexual men took significantly longer on the stimuli of women than on the stimuli of men. The heterosexual women did not differ significantly between stimuli of men and women. The gay men took significantly longer on the slides of men than on the slides of women. The lesbian women took significantly longer on the slides of women than on the slides of men. All four groups demonstrated significantly longer reaction times to slides of men and women than to the neutral slides which indicated both types of stimuli had interference effects.

Table 1. Mean Choice Reaction Times for Nude Stimuli

                            Stimuli Category
           Male Slides         Female Slides       Neutral Slides
Group   M            (SD)    M            (SD)    M         (SD)

He M    1.65(a, b)   (.29)   2.07(b)      (.59)   1.55(a)   (.22)
He F    1.68(a, b)   (.30)   1.62(a)      (.23)   1.53(a)   (.16)
Ho M    1.85(b)      (.29)   1.65(a)      (.22)   1.51(a)   (.13)
Ho F    1.61(a)      (.20)   1.82(a, b)   (.31)   1.52(a)   (.14)

                   t-test(a)
Group     M1 vs M2    M1 vs M3    M2 vs M3

He M      -4.99(*)     3.36(*)      5.73(*)
He F       2.04        3.92(*)      3.94(*)
Ho M       3.66(*)     6.03(*)      5.15(*)
Ho F      -4.57(*)     3.17(*)      5.66(*)

Note: He M = Heterosexual Male; He F = Heterosexual Female; Ho M = Homosexual Male; Ho F= Homosexual Female. Group means in each stimulus category with different superscripts are significantly different at p < .05, Tukey HSD test.

(a) Differences between slides for each group were evaluated using t-tests.

(*) Differences significant at p < .05.

The CRT data for the clothed stimuli were analyzed using ANOVA with two between factors (orientation x gender) and one repeated factor (stimulus type). A main effect for stimulus type was significant, F(2, 152) = 3.60, p [is less than] .05, indicating a faster reaction time to the neutral stimuli. The interaction between orientation x stimulus type was also significant, F(2, 152) = 3.90, p [is less than] .05. These results were qualified by a significant triple order interaction, F(2,152) = 3.66, p [is less than] .05. An analysis of simple effects using the Tukey HSD test revealed that no significant differences occurred between any of the groups on any of the stimuli, as may be seen in Table 2. The within subject effects were analyzed using multiple t-tests. Results indicate that the heterosexual men took significantly longer on the slides of women than on the slides of men. The heterosexual women took significantly longer on both the slides of men and women than on the neutrals but did not differ significantly between the slides of men and women. The gay men took significantly longer on the slides of men than on the slides of women. They also took longer on the slides of women than on the neutral stimuli. The lesbian women took significantly longer on the slides of women than on the slides of men and neutral stimuli, and took significantly longer on the slides of men than on the neutral slides.

Table 2. Mean Choice Reaction Times for Clothed Stimuli

                   Stimuli Category
        Male Stimuli   Female Stimuli   Neutral Stimuli
Group   M      (SD)    M      (SD)      M      (SD)

He M    1.55   (.26)   1.70   (.39)     1.68   (.75)
He F    1.56   (.14)   1.57   (.15)     1.51   (.15)
Ho M    1.68   (.18)   1.56   (.16)     1.48   (.16)
Ho F    1.61   (.20)   1.71   (.28)     1.51   (.18)

                   t-test(a)
Group    M1 vs M2   M1 vs M3   M2 vs M3

He M     -3.47(*)     -.86          .13
He F     -1.26        3.14(*)      4.50(*)
Ho M      4.30(*)     6.56(*)      6.02(*)
Ho F     -2.78(*)     5.54(*)      5.56(*)

Note: He M = Heterosexual Male; He F = Heterosexual Female; Ho M = Homosexual Male; Ho F = Homosexual Female. Group means in each stimulus category with different superscripts are significantly different at p < .05, Tukey HSD test.

(a) Differences between slides for each group were evaluated using t-tests.

(*) Differences significant at p < .05.

Error Analyses

A total of 28 errors occurred during the experimental procedure. Twenty-two errors occurred during the nude slides (13, 8, and 1 to the men, women, and neutral stimuli, respectively); but only six errors occurred during the clothed slides (2, 4, and 0 to the men, women, and neutral stimuli, respectively). Chi-square tests indicated significantly more errors to slides of men and women than to the neutral slides, [X.sup.2] (2) = 11.68, p [is less than] .01, and more errors to nude than clothed slides, [X.sup.2] (1) = 9.14, p [is less than] .01.

Recognition Score Analyses

The number of stimuli correctly recognized as present or absent during the incidental learning task were analyzed by using ANOVA with two between factors (gender x orientation) and one within factor--presence, (absent vs. present) and stimulus type, (clothed women, clothed men, nude women, nude men). Results indicate that the main effects for presence and stimulus type were significant; F(1, 76) = 221.49, p [is less than] .001 and F(3, 228) = 4.70, p [is less than] .01, respectively. These results indicate that novel slides were more often correctly detected and slides of nude women were better recognized than the other three types. No other main effects were significant. The double order interactions between gender x stimulus type and orientation x stimulus type were significant; F(3, 228) - 4.84, p [is less than] .01 and F(3, 228) = 9.18, p [is less than] .001, respectively. The triple order interactions between gender x orientation x stimulus type and gender x presence x stimulus type were significant; F(3,228) = 26.46, p [is less than] .001 and F(3,228) = 3.18, p .05, respectively. These interactions were qualified by a significant fourth order interaction between gender x orientation x presence x stimulus type, F(3, 228) = 18.72, p [is less than] .001. In general the analysis of simple effects using the Tukey HSD test and t-tests indicate that when novel stimuli were present, there were no significant effects of orientation, gender, or stimulus type on recognition. However, when stimuli had been previously shown to the subjects, each group showed significantly more recognition for their preferred gender when the stimuli were nude. Some of these differences occurred with clothed stimuli, but not all, which indicates a reduced effect for the clothed stimuli. These effects are graphically illustrated in Figure 1.

[Figure 1 ILLUSTRATION OMITTED]

A Pearson correlation coefficient was computed between the total number of stimuli correctly identified as present in each category during the memory task and the mean choice reaction time for the stimulus category during the experimental trials. The correlation between reaction time and recognition scores was r(78) = .253, p [is less than] .05.

Analyses of Correct Identification of Orientation by Various Indices

In terms of percentage of participants who had longer latencies on his or her preferred sex stimuli using the CRT task with nude stimuli, 100% of the heterosexual men, 50% of the heterosexual women, 100% of the gay men, and 100% of the lesbian women scored as predicted. The overall accuracy of the CRT procedure in differentiating among the four groups using the nude stimuli was 87.5%. These data are shown in Table 3. For the CRT task using clothed stimuli, 70% of the heterosexual men, 55% of the heterosexual women, 85% of the gay men, and 90% of the lesbian women scored as predicted. The overall accuracy of the CRT procedure in differentiating among the four groups using the clothed stimuli was 75%.

Table 3. Percent of Subjects Correctly Identified Across
Tasks

            Choice Reaction Time       Recall

Group        Nude     Clothed     Nude    Clothed

He M         100        70         90       50
He F          50        55         55       25
Ho M         100        85         75       90
Ho F         100        90         85       60
Overall
 accuracy    87.5%     75.0%     76.25%   56.25%

Note: He M = Heterosexual Male; He F = Heterosexual Female; Ho M = Homosexual Male; Ho F = Homosexual Female. Subjects identified by longer latencies or greater recall to preferred sex stimuli.

An analysis was conducted to determine if subjects could be classified with regards to sexual orientation based on their ability to correctly recognize preferred-gender slides during the incidental learning task. A change score was computed for number of preferred-gender stimuli correctly recognized during the incidental task by subtracting the number of nonpreferred-gender stimuli from the number of preferred-gender stimuli. Results indicate that for nude stimuli, 90% of the heterosexual men, 55% of the heterosexual women, 75% of the gay men, and 85% of the lesbian women were correctly classified. For the clothed stimuli, 50% of the heterosexual men, 25% of the heterosexual women, 90% of the gay men, and 60% of the lesbian women were correctly identified. The gay men preformed better on clothed preferred-gender stimuli than on nude preferred-gender stimuli. As demonstrated with other indices, recognition and classification accuracy were generally greater with nude stimuli.

DISCUSSION

The major purpose of this study was to determine the influence of slides of nude and clothed stimuli on cognitive processes during a choice reaction time task for heterosexual men, heterosexual women, gay men, and lesbian women. It was hypothesized that stimuli which elicited the greatest degree of sexual interest, such as slides of nude preferredgender individuals, would cause the most interference with the discrimination task and increase latencies to that category of stimuli. It was also hypothesized that slides of clothed individuals should elicit less sexual interest and should, therefore, cause less distraction but should cause more distraction than would be expected from neutral stimuli. The results obtained from this study are consistent with these hypotheses. The data indicate that varying levels of erotic stimuli will interfere with a choice discrimination task since even the clothed stimuli caused longer latencies than the neutral stimuli. However, the results of the present study indicate that the interference on the CRT task was significantly greater when the nude stimuli were used. If it can be assumed that nude stimuli are more sexually interesting than clothed stimuli and that the stimuli of the preferred gender elicit greater sexual interest than non-preferred stimuli, then it can be assumed that the interference is due to sexual interest. Further, these results also indicate that the greater the sexual interest, the greater the interference effects.

The present results are also consistent with literature on the Stroop effect. However, the present results expand our knowledge of the role of motivation on cognitive processes in two ways. First, most of the previous studies examining the effects of emotion on a modified Stroop task have been done with negative emotions, such as anxiety and depression. Sexual interest, the motivational variable used in this study, is a positive emotion. Consequently, the present data indicate that both positive and negative emotions influence cognitive processes. The second major implication of the present data is that sexual interest facilitates as well as interferes with information processing. On the incidental learning task, recognition for preferred-gender stimuli that were present during the experimental trials was significantly greater than for non-preferred-gender stimuli, particularly with the more sexually provocative nude stimuli. These results cannot be totally explained by the longer exposure to these stimuli since the correlation between length of exposure and recognition was small, r = .254. This indicates that the valance of the stimuli plays an important role in the amount of recognition.

A rather puzzling, unexplained finding with the memory data is that subjects had excellent detection of novel stimuli which were not present during the experimental trials. Subjects' accuracy rate during the incidental learning task for novel stimuli ranged from 85% to 93%. This contrasts sharply with the accuracy rate for stimuli previously seen, which ranged from 46% to 80% correct identification. The reason for this finding is not obvious.

This study is a replication and extension of a previous study to determine if sexual orientation can be determined by using CRT tasks. The present data support and replicate the finding of Wright and Adams (1994) that CRTs can be used, particularly with men, to identify sexual orientation. This effect is also present with the DT and TT procedures, but to a lesser degree. The present results also suggest that the greater the sexual interest the more accurate the identification of sexual orientation. Nude stimuli cause greater differentiation between groups than clothed stimuli.

While this procedure is appropriate with men, it is somewhat problematic with women, particularly heterosexual women. Heterosexual women generally had longer latencies to slides of nude men, but this difference was small and resulted in a lowered correct identification of sexual orientation. The results may be due to the fact that women are less sexually aroused by visual stimuli as compared to other stimulus modalities than men (Leitenberg & Henning, 1995). It is possible that an item analysis of the nude women stimuli would significantly enhance our ability to detect sexual orientations. These results are consistent with findings of other research on gender differences to erotic stimuli (Dekker & Everaerd, 1988; Geer, 1996; Geer & McGlone, 1990). These studies, however, examined gender differences to erotic stimuli but ignored the differences between individuals with respect to sexual orientation. It is unclear at this time why heterosexual women responded differently than the other three groups, and no explanation is offered at this time.

This study relied on self-report of sexual orientation for classification of subjects, and it is known that this information is easily distorted. Future research utilizing this procedure could involve using vaginal photoplethysmography or penile plethysmography as a screening tool to verify self-report of sexual orientation. By correlating the results of the choice reaction time task and a physiological means of assessing sexual arousal, such as plethysmography, concurrent validity of the choice reaction time procedure would be established.

While many of the individuals in this study were able to be appropriately identified by sexual orientation using the choice reaction time procedure with clothed stimuli, the effects were not as robust as with the nude stimuli. Further, recognition during the incidental learning task was also greater with nude stimuli, which provided for greater classification accuracy than the clothed stimuli. As for the inclusion of the DT and TT tasks to increase accurate classification, the choice reaction time task has 100% correct identification of 3 of the 4 groups--heterosexual men, gay men, and lesbian women. The only group whose accuracy in identification increased with the inclusion of the DT and TT tasks were the heterosexual women. In addition, the accuracy in correct identification for these groups was greatly diminished with the inclusion of the clothed stimuli. Therefore, it is recommended that only the choice reaction time procedure with nude stimuli be used in future projects.

The present data indicate that the CRT procedure has excellent reliability with the various stimuli. While the split-half reliability of the stimuli was quite high, future research may also want to determine test-retest reliability by assessing the same subjects on two different occasions. Another issue which should be investigated in future research is whether choice reaction times change as a result of habituation to the stimuli. Future projects should include a group of bisexual individuals to determine their response patterns. It would be interesting to determine at what point along the continuum of same-gender sexual arousal that a person begins to demonstrate longer latencies to same-gender stimuli. Another issue is whether an individual can fake the results of this task and what a faked profile looks like. A faking study should include the results of the choice reaction time task as well as the incidental learning task since both tasks have high correct identification accuracy rates. Additionally, it would be interesting to screen participants for erotophobia or for having experienced traumatic sexual encounters (e.g., rape, child molestation, etc.), to determine the impact of these experiences on reaction time.

Since this procedure has been shown to be a reliable means of assessing normal sexual interest, it may also be capable of detecting deviant sexual interest. By utilizing stimuli congruent with a person's deviant sexual interest, such as slides of children with pedophiles, this tool could be a promising procedure for assessing deviant sexual interest. This procedure would also be useful for situations in which an individual is unable to achieve an erection, which would invalidate the results of plethysmography (e.g., erectile failure or individuals experiencing severe anxiety during the testing situation). The CRT task may also be a more appropriate assessment tool for youth sex offenders since it is less intrusive. Using the procedure in conjunction with plethysmography and obtaining consistent results could strengthen one's statement about an individual's sexual interests.

(1) A delay time (DT) was computed by measuring the amount of time a participant chose to view the slide after he or she made his or her dot placement selection for the CRT. A total time (TT) was computed by combining the mean CRT and the mean DT for each subject for each type of stimuli. Results for the DT and TT tasks were very similar to the results obtained using the CRT and, therefore, are not described in detail. These data are available from the corresponding author.

REFERENCES

Dekker, J., & Everaerd, W. (1988). Attentional effects on sexual arousal. Psychophysiology, 25, 45-54.

Fox, E. (1993). Attentional bias in anxiety: Selective or not? Behaviour Research & Therapy, 31, 487-493.

Geer, J. H. (1996). Gender differences in the organization of sexual information. Archives of Sexual Behavior, 25, 91-107.

Geer, I. H., & McGlone, M. S. (1990). Sex differences in memory for erotica. Cognition and Emotion, 4, 71-78.

Kinsey, A. C., Pomeroy, W. B., & Martin, C. E. (1948). Sexual behavior in the human male. Philadelphia: W. B. Saunders.

Kintsch, W. (1974). The representation of memory. Hillsdale, NJ: Erlbaum.

Kohlers, P. A. (1976.). Buswell's discoveries. In R. A. Monty & V. W. Senders (Eds.), Eye movement and psychological processes (pp. 371-395). Hillsdale, NJ: Erlbaum.

Leitenberg, H., & Henning, K. (1995). Sexual fantasy. Psychological Bulletin, 117, 469-496.

MacLeod, C. M. (1991). Half a century of research on the Stroop Effect: An integrative review. Psychological Bulletin, 109, 163-202.

Pallak, M. S., Pittman, T. S., Heller, J. F., & Munson, P. (1975). The effect of arousal on Stroop color-naming task performance. Bulletin of the Psychonomic Society, 6, 248-250.

Segal, Z. (1995). A priming methodology for studying self-representation in major depressive disorder. Journal of Abnormal Psychology, 104, 205-213.

Simon, H. A. (1994). The bottleneck of attention: Connecting thought with motivation. In W. D. Spaulding (Ed.), Integrative views of motivation, cognition, and emotion. (Vol. 41). Lincoln, NE: University of Nebraska Press.

Singer, B. (1984). Conceptualizing sexual arousal and attraction. The Journal of Sex Research, 20, 230-240.

Wright, L. W., Jr., & Adams, H. E. (I 994). Assessment of sexual preference using a choice reaction time task. Journal of Psychopathology and Behavioral Assessment, 3, 221-231.

Manuscript accepted July 29, 1998

APPENDIX A

1. Specific instructions read to participants for Part 1 of the study:

This study examines the effects of sexually explicit stimuli on information processing and memory; it comprises three parts. For the first part, you will view a series of 60 slides. The slides will be of three types: nude women, nude men, and neutral slides/clothed men, clothed women, and neutral slides (depending on their condition). Each slide will have a white dot located in one of five positions: the upper left-hand corner, the upper right-hand corner, the middle, the lower left-hand corner, or the lower right-hand corner. Your task is to locate the position of the dot on each slide and press the button on the hand-held box that corresponds with the dot placement, (e.g., if the dot appears in the upper left-hand corner of the slide, you will press the button on the upper left-hand corner of the hand-held box). You are to locate the dot and press the button as quickly as you can, but try not to make any mistakes. After depressing the button you may look at the slide for as long as you like. When you are ready to advance the slide, press the red button located at the bottom center of the box. You will then have a 3-second intertrial interval during which time the projector lamp will illuminate the screen. At the conclusion of the intertrial interval, the computer will advance the slide projector and the next slide will appear. You will have five practice trials to prepare you for the task; after the practice trials, I will then answer any questions, and the experimental trials will begin. After Part 1 of the experiment is completed, you will receive instructions for Part 2. Do you have any questions before we begin?

2. Specific instructions read to the participants for Part 2 of the study:

This is Part 2 of the experiment. It is similar to Part 1 except that during this part of the experiment the stimuli are of (nude/clothed, depending upon the condition) individuals. Your task is the same as it was during Part 1 of this experiment. Locate the dot on the slide as quickly as you can and press the white button on the button-box that corresponds with its location. Then, when you want to advance the slide, press the red button located at the bottom center of the button-box and await the next slide. Work as quickly as you can to locate the dot and press the white button but try not to make any mistakes. Upon completion of this part of the experiment, I will provide you with instructions for Part 3 of the experiment. Do you have any questions?

3. Specific instructions provided to each subject for Part 3 of the study.

You will now see 120 slides, 60 of which are new and 60 of which you saw during Part 1 or Part 2 of this experiment. One half of the slides are of clothed individuals and one half are of nude individuals. There are no neutral slides in this part of the experiment and none of the slides will have a white dot. Your task is to try to remember if the slide was used in Part I or Part 2 of the experiment, or if it is novel. If you recognize the slide from Part 1 or Part 2 of the experiment, press the lower-left button. If the slide was not presented in Part 1 or Part 2 of the study, press the lower-right button. There are no blank spaces between the slides, and you do not need to press the red button. As soon as one slide is taken away, the next slide will follow it. Do you have any questions? I am now going to leave the chamber. I will ask you if you are ready to begin and will await your response before starting the experiment.

Lester W. Wright, Jr. and Adams Henry E. University of Georgia-Athens

Address correspondence to Lester W. Wright, Jr., Ph.D., Department of Psychology, Western Michigan University, Kalamazoo, MI 49008; e-mail: [e-mail].