Condom inclusion in cognitive representations of sexual encounters.

Risk or preventive behavior is frequently understood as a result of the conscious and rational analysis of all behavior-related pros and cons (Edwards, 1954; Linville, Fischer, & Fischhoff, 1993). The presumption that individuals weigh all the elements in a risk situation and make inferences and decisions according to fundamental principals of logic and rational choice, however, has been challenged. Systematic biases occur when people make decisions (Kahneman, Slovic, & Tversky, 1982; Slovic, Fischhoff, & Lichtenstein, 1987).

Since risk or preventive behavior does not necessarily involve rational and deliberate analysis, decisions could be based on more automatic or mindlessness processes (Bargh, 1994; Langer, 1978, 1994), possibly explaining the risk that individuals apparently embrace when participating in unsafe sex. Researchers have since begun to search for variables and models of spontaneous processing between social and cognitive factors and HIV preventive behavior (e.g., DiClemente & Peterson, 1994; Miller, Bettencourt, DeBro, & Hoffman, 1993; Norman & Conner, 1996).

The possibility that HIV preventive behavior could be based on less deliberate and more automatic processes instigated our research on the contribution of knowledge structures, namely, the script, to sexual risk behavior.

The script is a representation of the sequence of expected events in a given context and leads, almost automatically, to our understanding of the events and the guidance of our actions (Schank & Abelson, 1977). The more often a particular type of behavior is enacted in an activity, the more likely it is that it will be present in the script for that situation (Langer, 1978; Triandis, 1979; Verplanken, Aarts, & Knippenberg, 1997). A script increases the probability that individuals will rely on it for the accomplishment of the activity, since it reduces the required cognitive effort in its performance (Langer, 1978).

Three theoretical assumptions justify script automatism functioning. The first involves the economy of individuals' cognitive resources. The script reduces the uncertainty of the following step in the situation, thus, freeing cognitive resources for other less common activities requiring reflection and deliberation (Langer, 1978; Schank & Abelson, 1977). The second is based on individuals' unawareness of script use. The frequent and consistent association between a psychological phenomenon (e.g., affective and behavioral reactions) and situational characteristics creates an association, which, in the presence of environmental stimuli, triggers an automatic activation of the associated mental representations (Bargh & Chartrand, 1999). As the user unconsciously activates the scripts in the presence of situational characteristics, this unawareness is a further indication of the script's possible automatism (Schank & Abelson, 1977). Finally, the third assumption is based on the fact that repetitive behavior tends to be controlled by automatic processes (Aarts, Verplanken, & Knippenberg, 1998; Langer, 1978; Ronis, Yates, & Kirscht, 1989) and it is activity practice that underlies the construction of most scripts through schematic organization of their main regularities (Galambos & Rips, 1982; Schank & Abelson, 1977). Confronted with similar structural situations, information from the surrounding environment becomes dispensable, rendering cognitive evaluations and decisions superfluous and the script is instantiated. Therefore, scripts are considered one of the knowledge structures that contributes to automatism in decision making. Even if not automatic, scripts imply less effort in time and attention (Fayol & Monteil, 1988). In empirical terms, with script activation, decision processes are less elaborate and more automatic (Aarts et al., 1998; Langer, Blank, & Chanowitz, 1978; Verplanken et al., 1997). Response spontaneity is a criterion commonly used to determine the presence of a behavior in the script (Edgar & Fitzpatrick, 1993), as is the speed with which a situation is responded to (Verplanken et al., 1997).

Thus, we propose to identify the sexual scripts and evaluate the extent to which condoms are a part of these structures. We believe that the lack of condom use is partly a result of its absence in the sexual scripts, leading us to assume that the absence of protection might be nondeliberate.

Scripts and Sexual Behavior

Script information intervenes not only in social behavior (Langer, 1978; Langer et al., 1978) but also in sexual behavior (Maticka-Tyndale & Herold, 1999), since it is stereotyped sequences of actions associated with specific sexual encounters, some of which have been the object of empirical research (e.g., Rose & Frieze, 1989; Edgar & Fitzpatrick, 1993; Mewhinney, Herold, & Maticka-Tyndale, 1995).

Collectively, these studies show the existence of sexual scripts that influence the actions of the individual and provide a structure for understanding and anticipating the behavior of others. In the domain of sexual protective behavior, condoms have been mentioned at a very advanced stage in the sexual encounter (Edgar & Fitzpatrick, 1993). Indeed, the use of condoms increases with the likelihood of their inclusion in the sexual script, as supported by research (Maticka-Tyndale & Herold, 1999).

Script in Memory Tests

Early schema researchers proposed different cognitive representations for the perception of typical and atypical information, (Bartlett, 1932). A "typical" action is expected information, while "atypical" information is unexpected in the situation's knowledge structure. The internal representation of stimuli consisted of a schema containing typical information that required schema corrections when confronted with atypical information.

The prominent theoretical model, SC + T (schema copy plus tag) makes two predictions of script action memory. Atypical actions are easier to discriminate than typical actions, since the latter are general pointers and atypical actions are tagged and organized by a more salient process. Furthermore, there is no memory discrimination for very typical information because its memory representation, resulting from a pointer that activates the script as a whole (Graesser, Gordon, & Sawyer, 1979; Graesser, Woll, Kowalski, & Smith, 1980), makes it difficult to discriminate between explicitly stated actions and unstated or inferred actions from the generic script. Thus, better memory discrimination is expected for atypical over typical actions in recognition memory.

By better memory we mean better discrimination between the presented/not presented information, regardless of the response's criteria. In a recognition task, an individual may deliberately make excessive mistakes to increase the probability of a hit, while also increasing the probability of mistakes or being more demanding and only recognizing information when certain of its presentation. In other words, the strategies used involve response criteria that affect memory result but have hardly anything to do with it. The signal detection theory allows the separation of careful detection or discriminative memory from the criteria of response shared by individuals (Coombs, Dawes, & Tversky, 1970). Thus, a better memory should mirror the distinction between information presented and correctly identified (hits) and information not presented and incorrectly identified (false alarms).

Given the generic activation of the script when confronted with a typical action, more false alarms are expected for script actions. Great differences between typical and atypical actions are not expected for the hits. There is no linear relation between typicality and hits, since both an increase in hits with typicality is expected, due to these actions being more likely to be copied to memory, as well as a decrease since the probability of their being saliently targeted decreases with the action's typicality (Graesser et al., 1979, 1980).

This study was designed to gain understanding of the status of condom use in sexual scripts. Whenever people are used to certain behavior and have little time to think about their actions, it is believed that they will resort to the use of their scripts for the situation. When behavior is guided by script, condom inclusion in such representation will increase the likelihood of protection; however, if the condom is not part of the script, it will not become mentally accessible and will not be used. Therefore, its presence in the script is essential. Interventions may focus on the optimization of its use whenever the condom is part of a script, but they should also imply greater effort in the development of cultural expectations, namely, through media, school, and family messages.

Studies

Three studies were conducted to determine whether condom use is included in sexual scripts. These three ways involved the identification of sexual scripts (Bower, Black, & Turner, 1979), the devising of more spontaneous conditions for the investigation of condom inclusion in sexual scripts (Langer, 1978; Verplanken et al., 1997), and the study of the memory consequences of the script--(a)typicality of condom-related actions (Graesser et al., 1979). Condom references were examined through free-generation methods, open-ended narratives ending right before sexual intercourse, and a recognition memory test to clarify the typicality of condom use action in the sexual scripts. We hypothesized that if condoms are not part of the script, they will be less mentioned, due to the use of more spontaneous generation methods, and memory discrimination tests can be used to clarify condom status in this knowledge structure.

Three hundred and sixty undergraduate students (177 men), aged 18 to 30 years (98.9% White) participated from three universities in Lisbon. With faculty permission, students were requested to participate in a psychology research study. Those wishing to participate provided voluntary informed consent.

Study 1

This study's aim was to identify sexual scripts and condom presence. Identification of expected actions and events, as well as the sequence concerning both sexual encounters in stable and casual relationships were investigated.

Also, successful scripts (culminating in sexual intercourse) and unsuccessful scripts (ending without sexual intercourse) were compared. Stereotyped sequences of actions and events organized into scripts were expected for the more frequent types of sexual encounters.

Method

Participants. One hundred and twenty university students (60 women) aged 18-29 (M 21.2 and SD 2.58) participated in this study.

Procedure. We manipulated the type of encounter and the presence or absence of intercourse. The scripts created by students with emphasis on condom presence formed the dependent variable.

Participants were asked to write a list of around 20 typical actions/situations in sequential order (beginning with "they were together") describing what people normally do in one of four cases (stable or casual sexual encounter ending with "and they had intercourse" or "but they ended up not having intercourse").

The study followed a 2 type of relationship (stable versus casual) x2 sexual encounter success (intercourse present versus absent) between-participants design.

Fifteen women and 15 men, per script, described the actions for one of the four scripts, with no time limit.

Results

Content analysis. On a more abstract level we used broad categories inspired by literature on main scenes, plans, and strategies used in sexual encounters (e.g., McCormick, 1987; Miller et al., 1993), such as emotional closeness and social verbal interaction, the latter being defined as "verbal contact, which does not imply obvious sexual content, such as talking, talking about themselves, giving compliments to each other." On a more specific level, each category was subdivided into more specific actions/situations to accommodate the participants' thoughts, including categories such as talking about subjects that are trivial and/or removed from oneself and talking about subjects that are more intimate and/or close to oneself, for the above-mentioned abstract category. An abstract category set was created and the actions classified accordingly by one of the authors. This procedure was validated by having a random set of 25% of these actions assessed by a second judge, a trained psychologist, skilled in content analysis and blind to the aims of the study. Overall agreement was 81%. Mismatches were eventually resolved through discussion.

Actions belonging to the abstract categories were included in the scripts when mentioned by more than 30% of participants, a slightly more demanding criterion than the one habitually used in the literature (Maki, 1990). As the abstract categories did not mirror the level of specificity used by participants in their spontaneous production, however, those selected for each script corresponded to the most-used specific categories (the actions selected were mentioned by 17% to 73% of individuals; see Table 1).

The use of merely quantitative criteria to investigate the temporal dimension of the action sequence might have been misleading, since, for example, the reference to a kiss at the beginning and end of the encounter would have led to its inclusion in the middle of the sequence.

The temporal organization of the sequence of actions used the more frequent categories as markers to sequentially organize the categories with more than 30% of responses. Selected markers corresponded to actions mentioned by 75% or more participants: initial physical proximity behavior, sexual exchange behavior, establishment of mutual intention to progress to sexual intercourse, and the first marker to social verbal interaction only for the casual relationship scripts (sets illustrated in Table 2; see Appendix).

Script assembly. The actions found in the scripts are presented in Table I. Condom reference reached 40% in stable relationships with sexual intercourse (script 1) and 37% in casual relationships with sexual intercourse (script 2), and putting it on, an action referred to by 27% of individuals in both scripts. The condom as an obstacle was cited by 30% and 23% of individuals in stable and casual relationships without sexual intercourse (scripts 3 and 4, respectively). References to condom use or absence oscillated between 23% and 40% in the different scripts. Condom referral in the scripts was not significantly different, [chi square] (3, N = 120) = 2.24, p> .05.

Two main different obstacles were detected in unsuccessful sexual encounters. Accordingly, we distinguished three variants for stable sexual encounter scripts (scripts 1, 1a, lb) and for casual sexual encounter scripts (scripts 2, 2a, 2b). One variant was of successful sexual encounters (scripts 1 and 2), or, rather, scripts with sexual intercourse, and two were of unsuccessful encounters where obstacles to sexual intercourse emerged--one related to the absence of the condom (scripts la and 2a) and the other related to a different obstacle (scripts l b and 2b; see Table 2).

We used an interprototype similarity measure inspired by Tversky (1977) and adapted by Cantor, Mischel, and Schwartz (1982) for script comparison (Table 3). We verified that stable sexual encounter scripts, regardless of intercourse, were similar to each other and differed from casual sexual encounter scripts (except the a variants) and likewise for casual encounter scripts. Therefore, being in a stable or casual relationship contributes to greater differentiation of the sexual scripts than the presence or absence of intercourse.

Given the above pattern of similarities and differences, we considered only two scripts, one for stable relationships and the other for casual relationships, each with three variants: one where there is sexual intercourse, the other where obstacles arise, in which one is based on condom-related obstacles and the other on a different one.

Discussion

The results clearly indicate expectations regarding the sexual behavior involved in both stable and casual relationships.

The scripts are differentiated more by type (stable versus casual relationships) than by the actual occurrence of sexual intercourse. Despite the results obtained for condom use in stable and casual relationships in which there is sexual intercourse, the figures belong to the category of actions with the lowest percentage that met our criteria for script inclusion.

The condom is referred to as the most frequent obstacle in unsuccessful sexual encounters, but it is still a relatively infrequent one.

Condom reference is one of the latest actions of sexual interaction and occurs after the mutual intention to engage in sexual intercourse has been established, when sexual intercourse is imminent, in both types of relationships.

One may conclude that condom use related actions appear in successful sexual encounter scripts and also whenever condom absence is an obstacle in unsuccessful encounters. The relatively small percentage of individuals who refer to the condom in the script, as well as its late reference in the sequence, however, lead us to question its unequivocal presence in the script. Undoubtedly the scripts might dictate its presence toward the end of the sexual interaction. Even though silence regarding the condom up to the point of its use in a continued relationship might be due to a previous agreement in past sexual encounters between the partners, however, it is difficult to explain in casual relationships. The condom seems to appear out of the blue, with no previous referral or agreement regarding its use. Furthermore, we may find literature suggesting that this type of behavior is avoided by both sexes, males fearing a negative impact on the acceptance of sexual intercourse (e.g., Bryan, Aiken, & West, 1999), while females associate it with an inappropriate and negative image of themselves (Hynie & Lydon, 1995). This suspicion is reinforced by the specificities of the script identification task. In this task, participants can generate script actions with plenty of time and cognitive resources to engage in deliberate thought and to accommodate social desirability concerns. In fact, condom use related actions in the assembled scripts, due to strong social norms, have already been identified by Edgar and Fitzpatrick (1993). The critical question from a research perspective lies in whether equivalent levels of condom use referral would occur under conditions promoting less deliberate and monitored responses, in other words, under conditions promoting the use of the script and its automatic functioning.

Study 2

The aim of this study was to assess, less intrusively, the prevalence of sexually protective behavior in a script-based context. In contrast with the previous study (at this point we possessed the sequence of actions and events expected in sexual encounters), we requested information about condom use near the time of its use, in order to make the response quicker and less reflected (Verplanken et al., 1997). We asked participants to continue a script-based story, in accordance with what was expected to happen in a sexual encounter. Even though the knowledge to be activated was based on information regarding what was expected, the involvement created by the script and the speed of the production became allies of the situation's more available information.

We assumed that in situations of rapid response and greater involvement created by the script, there would be fewer references made to condoms if they had been included in the scripts through a longer-lasting methodology, which may have implied greater reflection.

Method

Participants. One hundred and two university students (51 women) aged 18 to 29 (M 19.58 and SD 2.1) participated in the study.

Materials. We used the scripts identified in Study 1 from which the final actions were removed. In the Intention variant of the stable and casual successful scripts (scripts 1 and 2), the text ended when the sexual partners mutually agreed to engage in sexual intercourse. In the Condom-Related Obstacle (scripts 1a and 2a) and in the Condom-Unrelated Obstacle variants (scripts lb and 2b) of the stable and casual unsuccessful scripts, the text ended after the identification of an initial obstacle to sexual intercourse. The obstacle in the Condom-Related Obstacle variant was condom absence and in the Condom-Unrelated Obstacle variants were feeling uneasy/inhibited or having met the partner too recently (final sentence in bold in Table 2).

Procedure. The independent variable consisted of the different scripts and endings. The dependent variable was the sexual protection exhibited in the continuation of the script's narrative by students.

Participants were asked to complete a script-based narrative just after partners get undressed (intention variants) or after finding an obstacle (condom-related or condom-unrelated obstacle variants) according to the following instructions:

"We are developing research about features of sexual interactions to learn about the diversity of behaviors, exchanges, the act of fantasizing and wishes with which everyone, with varying degrees of imagination, feeds his/her sexual experiences (real or imagined ones). We want you to complete the story according to what is expected to happen in that situation. Feel free to mention any kind of exchange evoked by the situation, letting yourself fantasize and be taken by the situation. ... But complete the story based on what is expected to happen."

The study followed a 2 type of relationship (stable versus casual) x 3 script variant (intention versus condom-related obstacle versus condom-unrelated obstacle) between-participants design, 17 participants per script variant.

Results

The inclusion of condom-related actions. The continuation of the scripts was organized according to 2 sexual intercourse (yes versus no) x 2 condom reference (yes versus no) categories.

In Table 4 each category, namely, narratives where protected sexual intercourse occurred--cell 1; unprotected sexual intercourse occurred--cell 2; sexual intercourse failed to occur due to condom-unrelated obstacles--cell 3; and narratives where sexual intercourse failed to occur due to condom absence--veil 4, was analyzed according to the manipulated scripts.

In the same type of relationship, script completion differed according to the variants, in both stable, [chi square] (6, N = 51) = 25.17, p < .01, and casual relationships, [chi square] (6, N = 51) = 32.91, p < .01. Sexual intercourse was mentioned less when the script ended with an obstacle, and sexual intercourse without condom use was referred to more when there were no obstacles. The same did not occur, however, for equal variants of different relational scripts. When the script ended with "they get completely undressed...," there were no differences in the completion of the scripts, [chi square] (3, N = 34) = 2.18, p>.05. The same occurred when the script ended with "they realize that they do not have a condom ...," the condom-related obstacle variant, [chi square] (3, N = 34) = 1.20, p > .05, or, in the condom-unrelated obstacle variant, where uneasiness was present at the end of the scripts, [chi square] (3, N = 34) = 1.18, p>.05. On this basis, unprotected intercourse occurred more frequently in the variants in which no obstacles were introduced and peaked in situations where the story's characters were unhampered by any explicit obstacle.

Condom use in sexual intercourse and the impossibility of having sexual intercourse because of condom absence, or, rather, preventive behavior, obtained 28.7% of all responses. More than half of the sample (53%) did not mention condom use in the continuation of the script-based narratives. Spontaneous reference to the condom was confined to 15% of individuals, who referred to it naturally during involvement in a sexual relationship (column 1 total). This figure remained the same when we removed the responses to condom-related obstacle variants.

When explicitly confronted with condom absence, 41% of students reported protective behavior. Nevertheless, in these circumstances 41% of participants continued the narrative, making the characters engage in unprotected sexual intercourse (column 2).

In the condom-unrelated obstacle variants, obstacles varied but condom absence, as a hindrance to successful completion of the sexual encounter, was never mentioned. Moreover, 53% and 35% of participants (for stable and casual scripts, respectively) constructed narratives in which unprotected sexual intercourse occurred.

Other frequently mentioned actions. Discrepancy between the frequency of spontaneous condom reference in Study 1 (between 23% and 40%) and in this study (28.7% using the condition referring to condom absence or 15% in the case of not considering the condition) led us to investigate the reference to two frequent behaviors likely to occur close to individuals' involvement in sexual intercourse. We selected kissing and caressing.

Both behaviors in Study 1 revealed frequencies above the 30% criterion. The results obtained in this study showed that the action of kissing was referred to by 30% of participants and the action of caressing by 41%. The scripts that presented the condom as an obstacle included fewer kisses and caresses, followed by the scripts that manifested other obstacles. The scripts in which obstacles were not expressed were those that had the highest reference to kisses and caresses.

All in all, these results tied in with the actions identified in Study 1. The greater reference to intimate caresses over kisses, after the partners undress, was common in the scripts and was also less frequent in the scripts in which sexual intercourse did not occur.

Thus, we concluded that condom use differed according to a slightly different sample and way of collecting information, which did not happen with two other actions identified in the scripts.

Discussion

The reference to condom use obtains similar values, but it is lower than those obtained in the script identification. When the results of condom-related obstacle variants (column 4) are excluded, the spontaneous inclusion of the condom is reduced to 15% of participants.

Unprotected sexual intercourse occurs more frequently in the participants' reports when no obstacles are mentioned at the outset. When the people are alerted to the possibility of hindrances to sexual relations, unprotected sexual intercourse occurs less frequently. These results seem to suggest that it is more natural to skip condom use, since it is referred to in the sequence of a cue that is also reinforced by the fact that 41% of individuals continue to report sexual intercourse even when alerted to condom absence. Thus, condom absence does not appear to be a strong enough obstacle to prevent sexual intercourse from taking place.

The differences found between condom inclusion in the cognitive representations of Studies 1 and 2 can be attributed to the fact that although in both studies we request spontaneous responses from the participants, in Study 1 production takes longer and reflects a series of steps leading up to condom reference (which occurs later). In Study 2 the response appears after the reading of a significant part of the script, in which it is natural for people to go with the flow of the situation when there is greater involvement, created by the story. Hence, in conditions in which response production is less immediate and may be more reflected, condom referral is different. Nevertheless, the fact that other behaviors (such as kisses and caresses) are similar in this study and in Study 1, also leads us to believe that the condom has a different status in the script.

In short, the results of this study increase the suspicion that condom use is not a part of the sexual script, in neither stable nor casual relationships. Experimental conditions created from more spontaneous and involved responses reduce the reference to condom use compared with lengthier conditions that lead to a more thought-out description, influenced by social norms. We believe that social norms tend to be approving of condom use since there is evidence that people tend to overreport it (Garry, Sharman, Feldman, Marlatt, & Loftus, 2002).

Study 3

Condom omission in the script cannot be considered conclusive in Study 2. Overall, the condom is mentioned by approximately 30% of participants, but lower condom reference in script variants, where it is not explicitly mentioned and where it is not encountered for other actions from the onset, may suggest a different status for condom-related actions that is important to investigate. The applied methodology was based on the script's behavior in memory tests, using a recognition memory test for the effect.

Memory discrimination to condom-related actions allows us to assess the typicality of condom use. People are more likely to instantiate a script when encountering a typical action than when encountering an atypical one, and this instantiation brings to light other typical actions, which may become confused with the observed action. This, however, is not expected with atypical actions. On this basis, we expect discriminative memory for typical actions to be less than for atypical actions (hypothesis 1).

Suspicions created by condom referral at the end of the script without previous agreement between partners in Study 1, and fewer condom referrals when participants have no clue to remind them of the condom in Study 2, suggest that condom use is more of an atypical rather than a typical action (hypothesis 2).

Method

Participants. Ninety-six students participated in this study (47 women), aged 18 to 30 (M 22.68, SD 2.81). Forty-two university students (25 women) participated in the evaluation of the typicality of actions.

Materials. Material for the memory test was constructed for each of the three variants of the stable and casual relationship scripts and required the selection of typical and atypical actions in accordance with the script.

All the actions of each variant of the script constituted typical actions, so there was no need to fabricate them (e.g., they smiled, kissed, and caressed each other). Unlikely, atypical information refers to content that is not expected to be included in the script. Hence, 24 actions per script relationship (12 actions for script 1 and 12 for the two variants a and b of the script and the same for script 2, 2a, and 2b) were constructed by one of the researchers. They were conceived to be relatively discreet without impeding the flow of events, or, rather, capable of showing themselves to be irrelevant for the script (e.g., putting a hand into a pocket, cleaning a pair of glasses).

In order to test the appropriate character of the atypical actions, they were organized in three sets of four actions for each variant 1 and 2 of the scripts. For variants a and b of each script, the same three sets of four actions were used. Each script was always made up of the same typical actions and of different atypical actions placed in the same location of the sequence, in a total of three different sets of actions for each script variant.

The typical and atypical actions of each script variant were rated by a normative rating group of participants that used a 6-point scale (from very typical to very atypical) to evaluate each action. Seven students rated each variant, and each participant rated three different sets, with different atypical actions. Evaluations for the typical actions fluctuated between 3.48 and 5.76, with an average of 4.82 (selected typical actions fluctuated between 3.48 and 5.71, with an average of 4.76). The ratings for the atypical actions obtained values between 1.29 and 5.00, with an average of 2.38 (selected atypical actions fluctuated between 1.29 and 2.86, with an average of 2.02).

The sets of typical and atypical actions for each variant of the script were as equivalent as possible. The actions were selected according to two articulated criteria: (1) typical actions were selected from those that obtained higher values of typicality and atypical actions were chosen from those that obtained lower values, and (2) the actions of one set shared values that were as similar as possible to each other (these values obtained maximum fluctuations of 0.34 for the typical actions and 0.43 for the atypical ones).

Then, two experimental versions for each script variant were constructed (versions I and II). These versions included four typical actions and four atypical actions in version I, absent from version II, and another four typical actions and four atypical actions in version II, absent from version I. These actions served as material for the memory recognition test and the evaluation of condom typicality. In Table 5 we have presented an example of a variant of the script, from the experimental versions and from the typical and atypical actions selected for the version.

Procedure. We manipulated the typical and atypical actions presented in the scripts. The dependent variable referred to the hits, false alarms, and the discriminative memory exhibited for these actions.

The procedure consisted of a memory recognition test, with the purpose of distinguishing the typicality of the script actions, based on its experimental versions.

The individuals were informed that they were going to participate in two different studies, one based on the formation of impressions about characters involved in stories they were going to read (our experimental task, even though this was not the objective) and the other on logical reasoning (consisting of a distracting task, but unknown to participants).

Each person had 50 seconds to read a set of three separate stories (two distracting scripts and an experimental one). The distracting scripts consisted of the restaurant script and the script of visiting a doctor referred to in the literature (Bower et al., 1979; Graesser et al., 1979). The experimental script consisted of a version of one of the six variants being studied (see Table 5).

After reading the stories, participants were presented with one of three crypto-arithmetic problems that constituted the distracting task. After 15-20 minutes for the completion of the exercise, a list of experimental script actions was presented for recognition (actions of versions I and II). The order of the actions was presented in two ways. Individuals evaluated each action according to a 6-point scale: I am certain that the item was not presented (1)/was presented (6). Points 4, 5, and 6 were considered affirmative judgments about the presence of the action in the script and points 1, 2, and 3 negative judgments about this presence.

In the first study, we found three main variants for each type of stable and casual relationship: one where there is intention to have intercourse and two where an obstacle prevents intercourse through condom absence or through other obstacles. Then, we were able to construct two versions for each variant (see Table 2). In this study, for each script a sample of typical and atypical actions was included in each script version. The study went on to follow a 3 script variant (intention versus condom-related obstacle versus condom-unrelated obstacles) x 2 script version (I versus II) x 2 action typicality level (typical versus atypical) mixed design, the last factor being within participants. Each participant was only tested in relation to one variant of the script, with a total of 16 individuals per variant.

Results

We used discriminative memory (d' = [micro]s - [micro]n/[sigma]n) as a measure of sensitivity to the discrimination of information presented and not presented, based on the theory of signal detection (Coombs et al., 1970).

The results obtained (see Table 6) were similar to the usual results for sets of typical and atypical actions for the structure of knowledge under study. The level of hits was not significantly different for typical and atypical actions, mixed two-way ANOVA F(1, 94) = 1.51, p > .05, the level of false alarms was significantly different and occurred more often in actions that were part of the script, mixed two-way ANOVA F(1, 94) = 115.59, p < .01. The same occurred for discriminative memory, which turned out to be significantly better for the atypical than the typical actions, t (95) = -10.97, p < .01 (one tailed).

We tested for order effects, and there was no interaction between the level of hits and order, F(1, 94) = .21, p>.05, nor any order effect in the hits results, F(1, 94) = .61, p>.05. There was also no interaction between the level of false alarms and order, F (1, 94) = .20, p > .05, and no order effect in the false alarm results, F(1, 94) = 1.19, p>.05. We did not test the order for memory discrimination because it had already been tested for hits and false alarms.

These results corroborated the typicality of the actions that constituted the scripts identified in Study 1. Furthermore, since the set of typical actions was differentiated from the set of atypical actions, we were able to compare memory discrimination for condom-related actions with the set of typical and atypical actions for the variants of the scripts where the condom was presented as a typical action (scripts 1, 2, 1a, and 2a). In this comparison we used confidence intervals relative to false alarms of typical and atypical actions, and when confronted with the condom (an action isolated from the others). We selected false alarms for this comparison because they were the main characteristic in which, theoretically, we expected differences between typical and atypical actions.

The confidence intervals obtained for p < .05 fluctuated between .37 and .59 for the typical actions and between .00 and .14 for the atypical ones. False alarms for the condom-related actions never occurred. The false alarm mean for condom-related actions is, thus, outside the confidence interval of typical actions, but within the interval for atypical actions.

Discussion

We obtain better memory discrimination for atypical actions than for typical ones, thus, corroborating hypothesis 1. As predicted, this difference is mainly due to a higher false alarm rate for typical over atypical actions rather than due to differences in hit rates.

In accordance with our second prediction, the other actions included in the scripts (typical actions) are much less discriminated than condom-related actions. Memory recognition for this supposedly typical action presents more similarities with the memory for atypical actions than with the memory for typical ones.

In this study, memory discrimination for condom-related actions approximates not those of irrelevant but those of inconsistent, more salient, atypical actions. Since the latter were selected from emotionally unremarkable and ordinary actions, condom-related actions may be distinguished by their inconsistent and not merely irrelevant character.

It is equally possible, however, to think of a different status for the action of condom use. Such could result from a very intimate action involving different emotions such as modesty, and which, for this reason, may exhibit a different mnemonic behavior, transforming it into an exceptional typical action. The results that corroborate the frequency of other typical script actions, such as kisses and caresses, through the more spontaneous exploration of Study 2, accentuate the unique nature of condom use. We are not aware of studies, however, in which a different mnemonic importance was obtained for typical actions with a particular affective weight.

General Discussion

There are socially shared expectations regarding the sexual behavior involved in stable and casual relationships. The studies conducted seem to suggest, however, that condom-related actions are not part of the sexual scripts, in either type of relationship. Thus, in situations in which behavior is guided by the script, protective behaviors will not become mentally accessible and will not be used.

In situations of greater involvement, such as those illustrated in Study 2, condom-related actions are referred to with less frequency when no clue about them is experimentally introduced. Mnemonic results for condom-related actions also suggest the atypical nature of condom-related behaviors and, therefore, the low probability of protected sexual intercourse in sexual encounters. Moreover, the total absence of false alarms for condom-related actions leads to the idea that these actions are inconsistent and not merely irrelevant as far as sexual scripts are concerned. This, however, does not exclude the possibility that condom-related actions can constitute exceptional typical actions. Therefore, the typicality of the condom found in Study 1 could be due to its normative character. The condom is integrated in a situation in which more thought can be given to the encounter, creating a more deliberate attitude. This is forgotten, however, when the experimental situation takes place in a more rapid and potentially more automatic manner. Indeed, in deliberate situations, the condom is referred to more frequently, despite the fact that this was verified in the comparison between direct inquiry about the use of sexual protection and identification of protection in sexual scripts (Edgar & Fitzpatrick, 1993). Nevertheless, although people are aware of the need for condom use, they do not activate this knowledge in more spontaneous situations and, as anticipated, its integration in sexual encounters is expected to be difficult.

The limited reference to condom use may result, however, from the overobvious nature of the action in such a way that any type of reference in the descriptions of the encounters becomes unnecessary. Nonetheless, this argument does not corroborate the results encountered in the literature regarding condom use, because we are dealing with a practice performed inconsistently by a majority of university students (Alferes, 1997; Alvarez, 2005). Should we consider condom-related actions to be an integral part of sexual scripts? The answer to this question should have an impact on sexual behavior campaigns, but for now our data are only suggestive of a negative answer. Ideally, we would like our work to serve as an incentive for continued research on behavioral interventions, given the relation between the script and people's high-risk behavior.

There are a number of limitations in these studies that need to be taken into consideration. The studies were based on convenient samples and, although the results are expected to generalize to include people with similar demographic and sociocultural characteristics, there is no indication of generalization for other age groups from different geographical regions or with different sociocultural backgrounds. Artificially created situations are a second limitation because they do not guarantee similarity to the actual experience itself. In spite of the fact that these situations were constructed to be as close as possible to the sequence of expected events in the sexual experiences, we are aware of how difficult it is to create a realistic environment.

The implications of these findings are relevant for sexual education programs, as they alert us to the need for creating opportunities for associations between stable and casual sexual encounters and condom use. An action should be considered normative in order to be included in the script, and this characteristic needs to be developed--condoms should be available to students in their social and cultural environments and media interventions showing condoms as an inherent part of multiple sexual situations (Keller & Brown, 2002), including the publication of advertisements and movies, such as the notorious 007 Licence to Kill, the only one where Her Majesty's agent explicitly shows condom use, may be reinforced as well as program interventions proposing materials, videos, and role-plays integrating condom use in sexual encounters (e.g., Fisher, Fisher, Misovich, Kimble, & Malloy, 1996). Once in the script, effort will be needed to make condoms more appealing in the sexual routine so that they become part of the heat of the moment.

The main purpose of these studies was to investigate the status of the condom in sexual scripts. Most scripts are constructed through schematic organization of the main regularities found in the situations and, as a result of their repetitive character, sexual encounters are situations for which scripts can be developed. Our results suggest that condoms are not part of the scripts for the stable and casual encounters investigated. Therefore, we strongly suspect that protective behavior will not be used in sexual encounters where action is guided by script. As sexual encounters are likely to be influenced by script information, future research is needed to explore ways of integrating condoms in the expected routine for sexual encounters.

Appendix

In order to arrive at our temporal arrangement, we were obliged to view the different actions in each set as a single unit and each action as a fraction of the whole. The temporal arrangement was calculated according to the relative position of the actions which were attributed a cumulative value. Take, for instance, the following example: in a set with 5 actions, each action carries a .2 value; in a set with 7 actions each carries a .14 value. The calculation for each action takes into consideration its relative position, so if the same action appears in third position in the first set and in fourth position in the second set (that is, in the middle of both sets), it will acquire .6 in the first and .57 in the second. This corresponds to an equal level of importance, despite the different number of actions in each set. When a set only had one single action its value was calculated differently. Given that throughout our temporal arrangement the last action in a set was always attributed the total value of the unit, whenever a set with a single action occurred it was given a much lower value (.1) so that its relevance would not be underestimated in the overall temporal arrangement. Finally, with a view to considering action repetition in a set's temporal arrangement, only the first was taken into account when the same action appeared in two consecutive positions. However, whenever the same action was repeated but appeared in a different position, it was dealt with as an isolated action.

References

Aarts, H., Verplanken, B., & Knippenberg, A. (1998). Predicting behavior from actions in the past: Repeated decision making or a matter of habit? Journal of Applied Social Psychology, 28, 1355-1374.

Alferes, V. R. (1997). Encenacoes e comportamentos sexuais: para ulna psicologia social da sexualidade [Scenarios and sexual behaviors: For a social psychology of sexuality]. Porto: Edicoes Afrontamento.

Alvarez, M. J. (2005). Representacoes cognitivas e comportamentos sexuais de risco." O guiao e as teorias implicitas da personalidade nos comportamentos de proteccao sexual. Cognitive representations and sexual risk behaviors: The script and the implicit theories of personality in sexual protective behaviors]. Published Doctor of Philosophy thesis, University of Lisbon. Lisboa: FCG/FCT.

Bargh, J. (1994). The four horsemen of automaticity: Awareness, intention, efficiency, and control in social cognition. In R. S.

Wyer & T. K. Srull (Eds.), Handbook of social cognition (2nd ed., pp. 1-40). Hillsdale, NJ: Lawrence Erlbaum Associates.

Bargh, J. & Chartrand, T. (1999). The unbearable automaticity of being. American Psychologist, 54, 462-479.

Bartlett, F. C. (1932). Remembering. A study in experimental and social psychology. Cambridge: Cambridge University Press.

Bower, G. H., Black, J. B., & Turner, T. J. (1979). Scripts in memory for text. Cognitive Psychology, 11, 177-220.

Bryan, A., Aiken, L., & West, S. (1999). The impact of males proposing condom use on perceptions of an initial sexual encounter. Personality and Social Psychology Bulletin, 25, 275 286.

Cantor, N., Mischel, W., & Schwartz, J. (1982). A prototype analysis of psychological situations. Cognitive Psychology, 14, 45-77.

Coombs, C., Dawes, R., & Tversky, A. (1970). Mathematical psychology: An elementary introduction. Englewood Cliffs, NJ: Prentice-Hall.

DiClemente, R. J. & Peterson, J. L. (1994). Changing HIV/AIDS risk behaviors: The role of behavioral interventions. In R. DiClemente & J. Peterson (Eds.), Preventing AIDS: Theories and methods of behavioral interventions (pp.1-14). New York: Plenum Press.

Edgar, T. & Fitzpatrick, M. A. (1993). Expectations for sexual interaction: A cognitive test of the sequencing of sexual communication behavior. Health Communication, 5, 239-261.

Edwards, W. (1954). The theory of decision making. Psychological Bulletin, 51, 380-417.

Fayol, M. & Monteil, J.-M. (1988). The notion of script: From general to developmental and social psychology. European Bulletin of Cognitive Psychology, 8, 335-361.

Fisher, J., Fisher, W., Misovich, S., Kimble, D., & Malloy, T. (1996). Changing AIDS risk behavior: Effects of an intervention emphasizing AIDS risk reduction information, motivation, and behavioral skills in a college student population. Health Psychology, 15, 114-123.

Galambos, J. A. & Rips, L. J. (1982). Memory for routines. Journal of Verbal Learning and Verbal Behaviour, 21, 260-281.

Garry, M., Sharman, S., Feldman, J., Marlatt, G., & Loftus, E. (2002). Examining memory for heterosexual college students' sexual experiences using an electronic mail diary. Health Psychology, 21, 629-634.

Graesser, A. C., Gordon, S. E., & Sawyer, J. D. (1979). Recognition memory for typical and atypical actions in scripted activities: Test of a script pointer + tag hypothesis. Journal of Verbal Learning and Verbal Behaviour, 18, 319-332.

Graesser, A. C., Woll, S. B., Kowalski, D. J., & Smith, D. A. (1980). Memory for typical and atypical actions in scripted activities. Journal of Experimental Psychology: Human Learning and Memory, 6, 503-515.

Hynie, M. & Lydon, J. (1995). Women's perceptions of female contraceptive behavior: Experimental evidence of the sexual double standard. Psychology of Women Quarterly, 19, 563-581.

Kahneman, D., Slovic, P., & Tversky, A. (1982). Judgment under uncertainty: Heuristics and biases. Cambridge: Cambridge University Press.

Keller, S. & Brown, J. (2002). Media interventions to promote responsible sexual behavior. Journal of Sex Research, 39, 67-72.

Langer, E. J. (1978). Rethinking the role of thought in social interaction. In J. Harvey, W. Ickes, & R. Kidd (Eds.), New directions in attribution research (Vol. 2, pp. 35-58). Hillsdale, N J: Lawrence Erlbaum Associates.

Langer, E. J. (1994). The illusion of calculated decisions. In R. Schank & E. Langer (Eds.), Beliefs, reasoning and decision making: Psycho-logic in honor of Bob Abelson (pp. 33-53). Hillsdale, NJ: Lawrence Erlbaum Associates.

Langer, E. J., Blank, A., & Chanowitz, B. (1978). The mindlessness of ostensibly thoughtful action: The role of "placebic" information in interpersonal interaction. Journal of Personality and Social Psychology, 36, 635-642.

Linville, P. W., Fischer, G. W., & Fishchoff, B. (1993). AIDS risk perceptions and decision biases. In J. Pryor & G. Reeder (Eds.), The social psychology of HIV infection (pp. 5-38). Hillsdale, NJ: Lawrence Erlbaum Associates.

Maki, R. H. (1990). Memory for scripts actions: Effects of relevance and detail expectancy. Memory & Cognition, 18, 5-14.

Maticka-Tyndale, E. & Herold, E. S. (1999). Condom use on spring-break vacation: The influence of intentions, prior use and context. Journal of Applied Social Psychology, 29, 1010-1027.

McCormick, N. (1987). Sexual scripts: Social and therapeutic implications. Sexual & Marital Therapy, 2, 3-27.

Mewhinney, D., Herold, E., & Maticka-Tyndale, L. (1995). Sexual scripts and risk-taking of Canadian university students on spring break in Daytona beach, Florida. The Canadian Journal of Human Sexuality, 4, 273-288.

Miller, L. C., Bettencourt, B. A., DeBro, S. C., & Hoffman, V. (1993). Negotiating safer sex: Interpersonal dynamics. In J. Pryor & G. Reeder (Eds.), The social psychology of HIV infection (pp. 85123). Hillsdale, NJ: Lawrence Erlbaum Associates.

Norman, P. & Conner, M. (1996). The role of social cognition models in predicting health behaviors: Future directions. In M. Conner & P. Norman (Eds.), Predicting health behaviour (pp. 197-225). Buckingam: Open University Press.

Pryor, J. & Merluzzi, T. (1985). The role of expertise in processing social interaction scripts. Journal of Experimental Social Psychology, 21, 362-379.

Ronis, D., Yates, J., & Kirscht, J. (1989). Attitudes, decisions, and habits as determinants of repeated behavior. In I. A. Pratkanis, S. Breckler, & A. Greenwald (Eds.), Attitude structures and function (pp. 213-239). Hillsdale, NJ: Lawrence Erlbaum Associates.

Rose, S. & Frieze, I. (1989). Young singles scripts for a first date. Gender & Society, 3, 39-48.

Schank, R. & Abelson, R. (1977). Scripts, plans, goals, and understanding. Hillsdale, NJ: Lawrence Erlbaum Associates.

Slovic, P., Fischhoff, B., & Lichtenstein, S. (1987). Behavioral decision theory perspectives on protective behaviour. In N. Weinstein (Ed.), Taking care: Understanding and encouraging self-protective behaviour (pp. 14-41). Cambridge: Cambridge University Press.

Triandis, H. C. (1979). Values, attitudes, and interpersonal behaviour. In H. Howe & M. Page (Eds.), Nebraska symposium on motivation, (Vol. 27, pp. 195-259). Lincoln, NE: University of Nebraska Press.

Tversky, A. (1977). Features of similarity. Psychological Review, 84, 327-352.

Verplanken, B., Aarts, H., & Knippenberg, A. (1997). Habit, information acquisition, and the process of making travel mode choices. European Journal of Social Psychology, 27, 539-560.

Maria Joao Alvarez

Educational Psychology, University of Lisbon

Leonel Garcia-Marques

Social Cognition, University of Lisbon

Correspondence concerning this article should be addressed to Maria Joao Alvarez, University of Lisbon, Educational Psychology, Alameda da Universidade, Lisbon 1649-013, Portugal. E-mail: [e-mail]

Table 1. Actions and Sequence of the Stable and Casual Encounter,
with and without Sexual Intercourse

Script 1--Stable Successful Sexual Encounter

TALK ABOUT SUBJECTS THAT ARE TRIVIAL
  AND/OR REMOVED FROM ONESELF 60%
Go for a walk 20%
PERCEIVE/COMMUNICATE WITH A LOOK OF DESIRING
  SEX AND/OR EMOTIONAL INVOLVEMENT 53%
Laugh/smile 17%
Get closer 17%
Be in the car 17%
Be hand in hand 27%
Caresses 47%#
KISS 63%
EXCHANGE CARESSES (BSE) 57%
PERCEIVE/COMMUNICATE WITH A LOOK OF DESIRING
  SEX AND/OR EMOTIONAL INVOLVEMENT 53%
KISS (BSE) 63%
Be at home 23%
Lie down 17%
Exchange caresses (SSI) 43%#
Remove several pieces of clothing 30%#
Recognize/Feel sexual desire 37%#
KISS (BSE) 63%
Exchange caresses (SSI) 43%#
Remove several items of clothing 30%#
UNDRESS 67%
Exchange caresses 33%#
Put on the condom 27%
Exchange caresses 33%#

Script 2--Casual Successful Sexual Encounter

Look 23%
PERCEIVE/COMMUNICATE WITH A LOOK OF DESIRING
  SEX AND/OR EMOTIONAL INVOLVEMENT 53%
TALK ABOUT SUBJECTS THAT ARE TRIVIAL AND/OR
  REMOVED FROM ONESELF 53%
TALK ABOUT SUBJECTS THAT ARE MORE INTIMATE
  AND/OR CLOSE TO ONESELF 67%
Public gathering place 23%
Drink 33%#
TALK ABOUT SUBJECTS THAT ARE MORE INTIMATE
  AND/OR CLOSE TO ONESELF 67%
Recognize/Feel sexual desire 37%#
Dance 33%#
Caresses 47%#
KISS 73%
KISS (BSE) 73%
Leave the place (not specifying how) 33%#
Exchange caresses (BSE) 40%#
Go towards home 47%#
Make the situation comfortable and/or "romantic" 43%#
KISS (BSE) 73%
Exchange caresses (SSI) 33%#
Kiss (SSI) 17%
UNDRESS 57%
Put on the condom 27%

Script 3--Stable Unsuccessful Sexual Encounter

Talk about subjects that are trivial and/or removed from oneself 40%#
Perceive/communicate with a look of desiring sex and/or emotional
  involvement 37%#
Talk about subjects that are more intimate and/or close to oneself 37%#
Laugh/smile 23%
KISS 53%
KISS (BSE) 60%
Go for a walk 17%
EXCHANGE CARESSES (BSE) 57%
KISS (BSE) 60%
Exchange caresses (BSE) 33%#
Recognize/Feel sexual desire 43%#
Remove several items of clothing 23%
KISS (BSE) 60%
Undress 27%
Not have a condom 27%                or  Turn away /feel afraid 20%
Refuse to have sexual relations 20%      Talk about resistances to 17%
Show understanding 23%                   having sexual relations
                                         Turn away/feel afraid 20%
                                         Show understanding 23%

Script 4--Casual Unsuccessful Sexual Encounter

Be in the disco 17%
PERCEIVE/COMMUNICATE WITH A LOOK OF DESIRING
  SEX AND/OR EMOTIONAL INVOLVEMENT 50%
Introduce oneself 43%#
Talk about subjects that are trivial and/or removed from oneself 43%#
TALK ABOUT SUBJECTS THAT ARE MORE INTIMATE
  AND/OR CLOSE TO ONESELF 60%
Recognize/Feel sexual desire 37%#
Laugh/smile 30%#
Accept 33%#
Drink 37%#
Dance 30%#
TALK ABOUT SUBJECTS THAT ARE MORE INTIMATE
  AND/OR CLOSE TO ONESELF 60%
Caresses 37%#
Kiss 33%#
Caresses 37%#
KISS (BSE) 63%
Exchange caresses (BSE) 40%#
KISS (BSE) 63%
Invite to one's house 40%#
Undress 20%
Not have a condom 23%          or   know each other for a short
                                    time 17%
Refuse sexual intercourse 17%       Talk about resistances to sexual
                                    17% intercourse
                                    A new obstacle arises 17%

In capitals--actions referred to by 50% or more of the participants;
In italics--actions between 30 and 49%; remaining actions between 17
and 30% exclusively.

Note. BSE = behavior of sexual exchange: all behavior related to
physical intimacy that emerges after initial behavior of physical
proximity, maximum of two consecutive actions, as long as clothing
around the genital area has not been removed; SSI = sequence of
sexual involvement: three or more references to behavior related
to physical intimacy, as long as clothing around the genital area
has not been removed.

Note: Actions between 30 and 49% indicated with #.

Table 2. Script Variants for Stable and Casual Sexual
Encounters

Stable Successful Sexual Encounter (Variant S1)

Carlos and Ana are a couple and decide to meet up on this particular
  day. They talk about various subjects, trivial things, during a walk.
  At a given moment they look each other in the eyes. They smile and
  get closer to each other. Inside the car/he takes her hand, strokes
  her hair and they exchange caresses/. They give each other some
  kisses and caress each other. They look at each other and kiss again.
  Once at home they lie down and fondle each other. They remove the
  upper part of the other's clothing. Their hearts start to beat
  faster. They continue to kiss and caress each other. They remove
  some more articles of clothing/. They get completely undressed as
  they continue to exchange caresses. He puts on the condom and,
  continuing to caress each other they have sexual intercourse.

Stable Unsuccessful Sexual Encounter Condom-related Obstacle
  (Variant S1a)

Carlos and Ana are a couple and decide to meet up on this particular
  day. They talk about trivial things and, at a given moment, look into
  each other's eyes. The conversation moves on to more personal
  subjects and they laugh/. They kiss each other/and they carry on
  kissing and walking through the park. Later on, they caress each
  other, they exchange kisses and continue to caress each other. They
  are attracted to each other and begin to remove some items of
  clothing. They continue to take off more clothes and kiss each other.
  They get completely undressed and/when they are about to have
  sexual intercourse they realize they do not have a condom. She does
  not accept having sexual intercourse and he understands. The
  situation ends without sexual intercourse.

Stable Unsuccessful Sexual Encounter Other Obstacle (Variant S1b)
(the same as variant S1a until they get completely undressed and/) she
  withdraws a little as she is not quite at ease. He tries to find out
  what is wrong and they talk about it for a while. When they get close
  to having sexual intercourse she withdraws again and he finally
  accepts her choice. The situation ends without sexual intercourse.
  Casual Successful Sexual Encounter (Variant S2)

Carlos and Ana do not know each other and, on this particular day,
  they meet in a disco. They notice each other and look each other in
  the eye/. He starts to chat to her and they talk for a while about
  trivial things and flatter each other before approaching more
  personal subjects. In the disco they have a few drinks, they chat and
  try to get to know each other better and they are physically
  attracted. They dance and/start to touch each other/. They kiss and
  continue to kiss and then they both decide to leave the place. Their
  caresses become more intimate and they decide to go to one of their
  houses. They listen to music, dim the lights and begin to kiss and
  exchange caresses again/. They get completely undressed, he puts on
  the condom and they have sexual intercourse.

Casual Unsuccessful Sexual Encounter Condom-Related Obstacle
  (Variant S2a)

Carlos and Ana do not know each and, on this particular day, they
  meet in a disco. They are in the disco when they stare at each
  other/. He introduces himself and she does the same and they chat
  about fairly general subjects until the conversation begins to get
  more personal. They feel attracted to each other and exchange
  smiles. She accepts a drink and they go for a dance. The
  conversation becomes more intimate and/they touch each other. They
  begin to kiss/and caress each other and kiss again. They caress
  each other and he invites her back to his house. They get completely
  undressed, they realize they do not have a condom and she refuses to
  have sexual intercourse. They end up not having sexual intercourse.

Casual Unsuccessful Sexual Encounter Other Obstacle (Variant S2b)
(the same as variant S2a until they get completely undressed) but she
 doesn't think they have known each other for long enough. They discuss
the resistance she is experiencing but further obstacles keep emerging
and they end up not having sexual intercourse.

Table 3. Similarly between the Scripts

Comparison     Type of Relationship      Sexual Encounter Success

Similarity *    S1/S1a,b   S2/S2a,b    S1/S2   S1a/S2a,b   S1b/S2a,b
                0.81 (a)   1.00 (a)    0.50    0.80 (a)    0.50 (a)
                0.76 (b)   0.93 (b)     --     0.48 (b)    0.58 (b)

Comparison            Without Relation Between Scripts

Similarity *              S1/S2a,b       S2/S1a,b
                          0.40 (a)       0.52 (a)
                          0.38 (b)       0.50 (b)

* a measure of inter-prototype similarity (this measure corresponds
to the division of the number of actions common to two scripts by
the sum of the number of single actions in each script).

(a)--comparison with variant a; (b)--comparison with the variant b.

S1--stable successful sexual encounter; S1a--stable unsuccessful
sexual encounter, condom-related obstacle: S1b--stable unsuccessful
sexual encounter, other obstacle.

S2--casual successful sexual encounter; S2a--casual unsuccessful
sexual encounter, condom-related obstacle: S2b--stable unsuccessful
sexual encounter, other obstacle.

Table 4. Inclusion of Condom-Related Actions in Open-ended Narratives
that Stopped Right Before Sexual Intercourse

                                               1               2

                                            Sexual         2 Sexual
                                          Intercourse     Intercourse
                                            With a         Without a
N = 102                                   Condom (%)      Condom (%)

Script 1 (Stable Intention)                   18              70
Script 2 (Casual Intention)                   24              76
Script 1a
  (Stable Condom Related Obstacle)            18              41
Script 2a
  (Casual Condom Related Obstacles)           12              41
Script 1b
  (Stable Condom Unrelated Obstacle)           6              53
Script 2b
  (Casual Condom Unrelated Obstacles)         12              35
Total                                         15              52.6

                                               3               4

                                           No Sexual
                                          Intercourse      No Sexual
                                         Due to Condom    Intercourse
                                           Unrelated     Due to Condom
                                           Obstacle         Related
N = 102                                   Reasons (%)    Obstacle (%)

Script 1 (Stable Intention)                   12               0
Script 2 (Casual Intention)                    0               0
Script 1a
  (Stable Condom Related Obstacle)             0              41
Script 2a
  (Casual Condom Related Obstacles)            6              41
Script 1b
  (Stable Condom Unrelated Obstacle)          41               0
Script 2b
  (Casual Condom Unrelated Obstacles)         53               0
Total                                         18.7            13.7

Table 5. Experimental Versions of the Stable Successful
Sexual Encounter(S1) and Test Actions

S1--Stable Successful Sexual Encounter

Version I
  Carlos and Ana are a couple and on this day decide to meet up.
  They talk about various subjects, trivial things, during a walk. At a
    given time they look each other in the eyes. They smile and she
    chews gum. Inside the car/he takes her hand and they exchange
    caresses. A dog passes nearby/. They give each other some kisses
    and caress each other. They look at each other and kiss again.
    Once at home he takes his wallet out of his pocket, they lie down
    and fondle each other. They each remove the upper part of the
    other's clothing. They continue to kiss and caress each other. They
    remove some more articles of clothing/as they continue to touch.
    He puts on the condom, asks if she likes the sheets and, continuing
    to caress each other, they have sexual intercourse.

Version II
  Carlos and Ana are a couple and on this particular day decide to
    meet up.
  They talk about various subjects, trivial things, during a walk. At a
    given time they look each other in the eyes, get closer to each
    other and he adjusts his watch. In the car/he strokes her hair,
    scratches an elbow and they caress each other/. They give each
    other kisses and caress. He puts his hands on the wheel, they look
    at each other and kiss again. Once at home they caress each other.
    They remove the upper parts of each other's clothing. Their hearts
    start to beat faster. They continue to kiss and caress each other.
    They remove more items of clothing/. They finish by completely
    undressing each other as they continue to exchange caresses.
    She smells an aromatic candle and continuing to caress each other
    they have sexual intercourse.

Test actions for the stable successful sexual encounter:
They smile (Typical I).
He takes her hand (Typical I).
They lie down (Typical I).
He puts on the condom (Typical I).
She chews gum (Atypical I).
A dog passes nearby (Atypical I).
He takes his wallet out of his pocket (Atypical I).
Asks if she likes the sheets (Atypical I).
Get closer to each other (Typical II).
He strokes her hair (Typical II).
Their hearts start to beat harder (Typical II).
They finish by completely undressing each other (Typical II).
He adjusts his watch (Atypical II).
Scratches an elbow (Atypical II).
He puts his hands on the wheel (Atypical II).
She smells an aromatic candle (Atypical II).

Table 6. Average Proportion of Kits and False Alarms and
Discriminative Memory for Typical and Atypical Actions in
the Totality of the Scripts

                  Typical Actions      Atypical Actions

Hits                   .86                   .83
False alarms           .42 *                 .09 *
d'                     .15 *                 .26 *

d' index of discriminative memory.

Mixed two-way Anova tests were computed for hits and false alarms.

* p < .01 and a t test for d', p < .01 .