A theory-based approach to understanding sexual behavior at Mardi Gras.Mardi Gras Mardi Gras (mär`dē grä), last day before the fasting season of Lent. It is the French name for Shrove Tuesday. Literally translated, the term means "fat Tuesday" and was so called because it represented the last opportunity for merrymaking and excessive indulgence in food and drink before the solemn season of fasting. is an annual festival in New Orleans that begins 47 days before Easter and ends the Tuesday before Ash Wednesday (Gotham, 2002; Sexton, 1999). Its roots are grounded in French culture, as early explorers brought the tradition of Mardi Gras with them to Louisiana (Gotham; Sexton, 1999). The French originated the now-famous masked balls and parties in the early 1700s; however, by the late 1700s, Spanish governors banned these festivities (Gotham). Mardi Gras celebrations returned to the city by the early 1800s, and in order to preserve the festivities, a secret society of men called the Mystic Krewe of Comus Comus (kō`məs), in late Roman legend, god of mirth and revelry. A follower of Dionysus, he was represented as a drunken youth bearing a torch. In Milton's poetic masque, Comus, he is the mischievous son of Bacchus and Circe. planned the first official Mardi Gras parade in 1857. Today Mardi Gras symbolizes a time to celebrate the rich culture and history of New Orleans. Parades, costumes, and beads permeate the environment from morning to night (Gotham). Also synonymous with Mardi Gras, however, are the wild crowds, the abundance of alcohol consumption, and public sexual expression (Shrum & Kilburn, 1996). Noise levels are high, peaking at 90 decibels. The smell is acrid, with beer and alcoholic beverages being consumed, spilled, and crushed underfoot on busy, colorful streets teeming with people (Shrum & Kilburn). Exposed body parts, public sex acts, and flashed breasts, buttocks, and penises are commonplace in many of the parade areas (Redmon, 2003a). Mardi Gras behaviors 1. The actions or reactions of persons or things in response to external or internal stimuli. 2. The manner in which one behaves. be·hav have been studied most commonly from the
sociological perspective. Researchers have noted that during Mardi Gras,
behaviors are permitted that are normally considered socially
inappropriate (Forsyth, 1992). In fact, some have argued that the
festival actually encourages the inversion of standard norms of public
conduct (Jankowiak & White, 1999). Redmon (2003a) labeled this
behavior "playful deviance." Playful deviance occurs most
frequently when groups of tourists travel to leisure locations and
engage in types of behaviors that they would not normally enact at home.
Most often these places of leisure are located inside "themed
environments" and incorporate a series of normative instructions
for how to perform playful deviance. Goffman (1959) suggested that three
criteria must be fulfilled for playful deviance to occur: (a) the
setting and props for the behaviors must be appropriate; (b) individuals
must choose the setting to enact the deviant behaviors and terminate the
behaviors upon departure from the locale; and (c) the setting must offer
protection from social sanction for the individuals engaging in the
behaviors. ior·al adj.All of these specifications are met at Mardi Gras. The Mardi Gras environment is replete with symbols distinguishing its unique cultural context. Images and themes of eroticism, public nudity, public sex, and alcohol consumption encourage tourists to participate in the festivities as a part of legitimate fun (Redmon, 2003a). Wearing beads and masks transforms the individual into a performer engaging in behavior that would otherwise be considered inappropriate (Jankowiak & White, 1999). Mardi Gras behaviors are short-lived, specific, and localized to the Bourbon Street area (Redmon, 2003a; Shrum & Kilburn, 1996). Participants in playful deviance feel safe engaging in a range of typically taboo behaviors with relative security (Redmon, 2002). The anonymity of the massive, likeminded crowd protects the individual from stigma and judgement. It is well-known that "what one does during Mardi Gras does not count as a mark on one's character" (Forsyth, 1992, p. 395). In fact, crowds at Mardi Gras embrace and reward sexualized behaviors (Redmon, 2003a). Mardi Gras has also been considered a "time out" place (Jankowiak & White, 1999) or, similarly, a "backspace (1) To move the screen cursor one column to the left, deleting the character that was in that position. A backspace to the printer moves the print head one column to the left. (2) To move to the previous block on a magnetic tape." (Redmon, 2003a). A backspace is an area where individuals can transgress norms and participate in playful deviance without fear of reproach. This concept has also been described as "liminality"--a sense of inhabiting a margin or space where social rights and obligations may be temporarily suspended (Apostolopoulos, Sonmez, & Yu, 2002; Ford & Eiser, 1996, Shields, 1990). The more specific term of "situational disinhibition 1. A loss of inhibition, as through the influence of drugs or alcohol. 2. A temporary loss of an inhibition caused by an unrelated stimulus, such as a loud noise. Sociologists have attempted to study what characteristics of individuals lend themselves to immersion in the Mardi Gras festivities and playful deviance. Social control theory has been put forth to explain such behaviors. Social control theorists suggest that lack of connections or ties to the community lead to deviant acts (Redmon, 2002). Low self-control theory has also been used as a framework for understanding deviant behavior. Individuals who participate fully in the Mardi Gras environment might not perceive the negative consequences of their actions; instead, they develop an immediate-gratification orientation (Redmon, 2003b). No support has been found for these theories. Mardi Gras provides an opportunity to escape the rules and boundaries that social and self-control create and maintain in everyday life. Further, some researchers have discounted the importance of background variables, such as personality attributes, in influencing deviance in favor of emphasizing foreground variables, such as characteristics of the immediate environment (Katz, 1988). Public health and sexuality researchers have also acknowledged the important role context can play in shaping sexual behaviors. In particular, the Triandis theory of interpersonal behavior has been used to identify important predictors of sexual behaviors, including characteristics of the situation and personality variables (Maticka-Tyndale, Herold, & Mewhinney, 1998; Matricka-Tyndale et al., 2003). The Triandis theory of interpersonal behavior (TIB) is similar to other cognitive models, such as the theory of reasoned action and the theory of planned behavior (Bamberg Bamberg (bäm`bĕrk), city (1994 pop. 70,770), Bavaria, S Germany, a port on the Regnitz River. It is an industrial and commercial center; its manufactures include textiles, electrical equipment, clothing, machinery, and beer. Bamberg was the capital of a powerful ecclesiastical state from 1007 to 1802. In 1803 it passed to Bavaria. & Schmidt, 2003; Boyd & Wandersman, 1991; Godin et al., 1996). These models emphasize the role intentions play in determining behaviors, and how attitudes and values combine to create intentions. Triandis goes beyond these concepts in his tri-level model by adding a previous experience construct he labels "habit" and the presence of conditions that may facilitate or hinder its performance (Boyd & Wandersman; Godin et al.; Triandis, 1977, 1980, 1994). At the first level, he theorizes about how personal characteristics (such as gender, race, and social class) and prior experiences shape personal attitudes, beliefs, and social norms related to behavior. At the second level, Triandis explains how cognition, affect, social determinants, and personal normative beliefs influence the formation of intentions with regard to a specific behavior (Godin et al.). Finally, at the third level, Triandis proposes that intentions regarding the behavior, prior experience with the behavior (or very similar behaviors), and situational conditions predict whether or not the individual will engage in the behavior (Triandis, 1980). The TIB has been particularly effective at explaining complex behaviors that may be impacted by the social or physical environment. Most relevant for sexual health research are the constructs situated at the second and third levels, for these are perhaps the constructs that lend themselves to change through intervention. Cognitive and affective components work together with social determinants and personal normative beliefs at the second level of Triandis theory to predict intentions to engage in a behavior (Triandis, 1980, 1994). The cognitive and affective components refer to judgments and feelings the individual may have regarding the behavior of interest. Specifically, the cognitive component is evaluative of the probable consequences of engaging in the behavior (e.g., good/bad, positive/negative, responsible/irresponsible). The affective component refers to feelings one might anticipate as a result of engaging in the behavior (e.g., glad/sad, empowered/disempowered, proud/guilty). The social determinants at the second level incorporate role beliefs and subjective social norms. The perceptions regarding what is appropriate for a member in one's position make up role beliefs. For example, the individual evaluates whether engaging in a behavior would be acceptable for someone of his or her gender or at his or her age. Subjective social norms refer to the perceived group norms and expectations of the individuals' peer group. Specifically, one might estimate whether his or her same-gender friends or opposite-gender friends would perceive the behavior to be appropriate. Internalized standards or moral codes make up the final construct shaping intentions at the second level of the Triandis theory, in the construct labeled personal normative beliefs. In deciding whether or not to engage in a behavior, an individual must assess it in terms of his or her own morals and values. Considerable attention has been devoted to the relations between behavioral intentions and subsequent behavior. However, Triandis has specifically considered the roles that prior experience and situational conditions play in guiding behavior (Maticka-Tyndale et al., 1998). Triandis proposed that the influence of prior experience is strongest when the behavior of interest most closely parallels the previous behaviors and when the previous behavior occurred frequently. Triandis also emphasized that the situational conditions be directly relevant to the behavior of interest. While intentions and prior experiences are the primary influences on behavior, the TIB suggests that situational conditions can facilitate or impede behavior. For example, if an individual has previous experience with casual sex and expects he or she will engage in casual sex, but if no willing partners are available, the conditions of the situation preclude casual sexual behavior. Using the TIB, researchers have studied the role the environment can play in influencing sexual behavior in three settings: spring break in the United States, "schoolies week" in Australia, and beachfront holidays in Europe. Maticka-Tyndale and colleagues (1998) used the TIB as the guiding framework for their study on casual sex during spring break. Intentions to engage in casual sex on spring break and participation in spring break experiences differentiated between those who did and did not engage in casual sexual behavior. In a later study focused on predicting condom use with the TIB, Maticka-Tyndale and Herold (1999) found that rate of condom use was associated with all predictor variables in the TIB--intentions, prior condom use behavior, and situational conditions facilitating condom use. Conditions facilitating condom use explained the largest portion of variance in rate of condom use. Situational conditions related to alcohol use explained the most variance in casual sex experience in a similar study of spring break behavior (Apostolopoulos et al., 2002). Situational conditions reflecting condom availability were the strongest predictors of condom use (Apostolopoulos et al.). In qualitative studies about the spring break experience, participants have described risky sexual behaviors as the norm and excused various behaviors as being the result of the vacation experience (Mewhinney, Herold, & Maticka-Tyndale, 1995). Smith and Rosenthal (1997) reported on young people's experience of "schoolies week," a week when new high school graduates typically travel to the Australian Gold Coast for a beachside holiday. Almost half of participants (39%) reported engaging in sexual intercourse, and almost half of those (40%) did so within the first four days. Using the TIB to explain schoolies week sexual behavior at Surfer's Paradise in Australia, Maticka-Tyndale and colleagues (2003) found that the extent of participation in the activities associated with the Surfer's Paradise environ ment was the strongest predictor of engaging in casual sex for males and the second strongest predictor (following intentions) for females. Studies of sexual behavior among young people on holiday in Torbay Torbay, city (1991 pop. 54,430) and district, Devon, SW England. The district comprises the former municipal borough of Torquay and the urban districts of Paignton and Brixham. On Tor Bay is a noted tourist resort area, known as the "English Riviera." William of Orange landed at nearby Brixham in 1688. A cave on Windmill Hill contains animal bones and flint implements from the Paleolithic period., England, found the degree of situational disinhibition to be a significant predictor of sexual contact (Eiser & Ford, 1995; Ford & Eiser, 1996). In fact, 64% of participants who agreed "I am a different person on holiday" and 74% of participants who endorsed "I am not so worried about what other people think of me when I am on holiday" indicated that they had intercourse with two or more people at Torbay (Ford & Eiser). In a qualitative study of women's experiences of sexual intercourse on holiday in predominantly European holiday resorts, women perceived holidays to be a period of time removed from everyday life (Thomas, 2000). The women noted that the alcohol consumption and the perception of personal anonymity facilitated sexual involvement. Previous research indicates that sexual behavior is influenced by characteristics of the setting or context. From a sociological perspective, it seems the Mardi Gras environment facilitates sexual behavior; however, this has not yet been studied from a public health perspective. Therefore, the purpose of our investigation was to study the sexual behavior of young people at Mardi Gras using the theory of interpersonal behavior as a guiding framework. METHOD Questionnaire Development Following Triandis' theory (1980), the measurement tool was developed such that each construct was operationalized with respect to a specific time, context, and behavior of interest. Triandis provided the structure and form for each concept (i.e., personal attitudes are measured using semantic differential scales, social norms are measured using Likert-type scales), but specific items and indicators must be developed based on preliminary elicitation research. Elicitation research for this study was conducted following Triandis' guidelines. Two focus groups were conducted, one with male participants (N = 4) and one with female participants (N = 4), all of whom had attended Mardi Gras within the previous four years. Content analysis of the focus group data was conducted to determine key features of the Mardi Gras environment. In addition, questionnaires from Maticka-Tyndale et al. (1998) and Maticka-Tyndale and Herold (1999) were used as models. Where possible, similar item terminology was used in order to facilitate comparisons of findings. Measurement of Constructs Affective and cognitive components. Seven-point semantic differential scales consisting of adjective pairs describing the experience and consequences of having sex at Mardi Gras with a new partner comprised the affective and cognitive attitude components. To ensure that these were measures of attitudes toward sex at Mardi Gras and not of condom use, participants were told to make their evaluation "regardless of whether or not a condom is used ..." The adjective pairs measuring feelings about having sex with someone the participant met at Mardi Gras included sexually liberated/sexually conservative, empowered/disempowered, happy/angry, attractive/unattractive, good about myself/bad about myself, proud/guilty, not lonely/lonely, fun-loving/serious, adventuresome/ordinary, and glad/sad. Respondents' scores for each of the 7-point Likert-type items were summed to create a total scale score (M = 44.92, SD = 9.34). High scores indicated positive feelings regarding having sex with someone the participant met at Mardi Gras. Cronbach's alpha for the affective measure was .74. The adjective pairs measuring cognitive evaluations of having sex with a partner met at Mardi Gras included good/bad, responsible/irresponsible, fun/not fun, positive/negative, and smart/stupid. Respondents' scores for each of the 7-point Likert-type items were summed to create a total scale score (M = 24.34, SD = 5.94). High scores indicated positive evaluations of having sex with someone the participant met at Mardi Gras. Cronbach's alpha for the cognitive measure was .54. Personal normative beliefs. The strength of personal norms supporting casual sex was measured by the mean score of responses to two items, each with a 7-point strongly agree to strongly disagree response option. The participants were asked, "What do you think about having sex with someone you meet at Mardi Gras?" with items including, "you would feel guilty if you had sex with someone you meet there" and "it would be against your values to have sex with someone you meet there." Respondents' scores for each of the items were summed to create a total scale score (M = 7.67, SD = 4.08). High scores indicated normative beliefs unsupportive of engaging in sex with someone the participant met at Mardi Gras. The Cronbach's alpha for this measure was .82. Subjective social norms. The measure of subjective social norms assessed the individual's perception of how much one's peer group approves or disapproves of his or her sexual behavior at Mardi Gras, with a summed score for three items with 7-point extremely likely to extremely unlikely response options. Participants were asked, "If you met someone who is sexually attractive at Mardi Gras, how likely would it be that (a) your close male friends, (b) your close female friends, and (c) the friends you are at Mardi Gras with, will think you should have sex with this person?" The mean score on the measure was 13.85 (SD = 5.50). High scores indicated the participant perceived his or her peer group to be supportive of his or her engaging in sex with someone he or she met at Mardi Gras. The Cronbach's alpha for this measure was .83. Role beliefs. Items measuring the role belief construct focused on the degree to which the participant believed having sex with someone he or she met at Mardi Gras would be appropriate for someone of his or her same social status or position. The summed score for three items with a 7-point strongly agree to strongly disagree response option was used to measure this construct. Respondents were asked, "What do you think about having sex with someone you meet at Mardi Gras?" which was followed by the three items: (a) It's okay for someone young like me to have sex with someone they meet at Mardi Gras; (b) It's okay for someone not in a relationship back home to have sex with someone they've met at Mardi Gras; and (c) It's okay for someone of my gender to have sex with someone they've met at Mardi Gras. The mean score on the measure was 15.54 (SD = 5.48). High scores indicated role beliefs supportive of engaging in sex with someone the participant met at Mardi Gras. The Cronbach's alpha for this measure was .91. Situational conditions. During the elicitation phase, students identified specific situations or experiences that had occurred at Mardi Gras that they thought were conducive to or that impeded engaging in sexual intercourse with a new partner. To measure situational conditions, participants were asked, "Since you have been at Mardi Gras, how often have the following happened?" This question was followed by 17 items describing experiences that were consistent with the conditions most consistently reported by participants in the elicitation phase. Principal components factor analysis with varimax rotation was used to determine underlying constructs. Items of each factor were summed to create a subscale. The analysis revealed three factors that accounted for 62.5% of the variance. The first factor, labeled Sexualized Environment, included 8 items with factor loadings ranging from .576 to .779, and accounted for 27.63% of the variance. The items were (a) You "dirty danced" or grinded (i.e., danced in a sexy way) with someone you met; (b) It seems like everyone is having sex; (c) You tried to pick someone up; (d) Someone tried to pick you up; (e) You have been sexually aroused; (f) You've met someone who wants to have sex with you; (g) You wanted to be alone with someone but there hasn't been any place private [reverse coded]; and (h) You have "fooled around" in a sexual way with someone you met. Participants used a 4-point scale ranging from 1 (never) to 4 (frequently) to rate the frequency with which they expected to be involved with each of the experiences. High scores on this subscale indicated greater participation in a sexualized environment at Mardi Gras. The mean score was 20.72 (SD = 5.51). The Cronbach's alpha for this measure was .74. The second factor, labeled Alcohol Use, included five items with factor loadings ranging from .665 to .898. The factor accounted for 20.06% of the variance. The items were (a) You were in a "let loose," "have fun" mood; (b) You drank alcohol; (c) You got drunk; (d) You drank alcohol in a public place (i.e., the street); and (e) You drank alcohol in the middle of the day. High scores on this subscale indicated greater use of alcohol at Mardi Gras. The mean score was 17.87 (SD = 3.15). The Cronbach's alpha for this measure was .89. The third factor, labeled Mardi Gras Culture, included five items with factor loadings ranging from .562 to .777, and accounted for 14.80% of the variance. The items were (a) Someone has offered you beads if you showed a part of your body; (b) You have offered someone beads if they showed a part of their body; (c) Someone has offered you beads in exchange for a sexual favor; (d) You have offered someone beads in exchange for a sexual favor; and (e) You wore a mask or some other Mardi Gras "costume." The mean score was 11.64 (SD = 4.51). High scores on this measure indicated greater immersion in the Mardi Gras culture. The Cronbach's alpha for this measure was .80. Prior experience. The Triandis theory specifies that measures used to assess prior experiences approximate the behavior of interest as closely as possible. Since participants may not have attended Mardi Gras on a previous occasion, the use of experience specific to the Mardi Gras environment was not considered to be feasible. As an alternative, we used three items measuring prior experience with oral sex, vaginal sex, and anal sex with someone the participant had known for less than one week. For oral sex the items was scored dichotomously (1 = yes, 2 = no). For vaginal and anal sex, the items were open-ended and participants were asked to fill in the number of vaginal and anal sex partners they had previously and for whom they had known for less than one week. Peer sexual activity. Triandis (1980) emphasized that internalized perceptions of the peer group can influence behavior. Therefore, a measure of perceived peer sexual activity was added as a predictor of sexual behavior at Mardi Gras. Specifically, participants were asked, "How many of your same-sex friends have had sex with someone they met at Mardi Gras?" Response choices were: (a) none; (b) a few; (c) about half; (d) more than half; (e) most; and (f) don't know. The mean response to this item was 2.02 (SD = 1.55). Outcome variables. Two criterion variables were used in this study. First, intention was measured by asking participants if they expected or intended to engage in oral sex, vaginal sex, or anal sex with a new partner at Mardi Gras. Secondly, behavior was measured by asking participants if they had engaged in oral sex, vaginal sex, or anal sex with someone that they had met at Mardi Gras. Both items were measured with a dichotomous yes/no response option. Data Collection Five female graduate students with experience and training in public health and research methods and who were in the age range of participants of interest attending Mardi Gras collected data during the 2004 Mardi Gras season. Recruitment procedures were similar to those followed by Maticka-Tyndale et al. (1998). Participants were recruited from public areas surrounding Bourbon Street between the hours of 10a.m. and 5p.m. from Saturday to Wednesday. Data were collected on Wednesday in order to capture participants after Fat Tuesday, given that most Mardi Gras participants arrive the weekend prior to Mardi Gras and leave shortly after Fat Tuesday. The timing for data collection each day ensured that participants were likely to be sober, as the heaviest alcohol consumption was expected to occur in the late afternoon (based on the experiences of those reported during the elicitation phase of this study). Individuals were approached by research assistants in teams of two or three who explained the survey procedures and reviewed the ethical guidelines with potential participants. Those who agreed to participate were given a questionnaire on a clipboard and a pencil. Five questionnaires were distributed at one time. Research assistants encouraged participants to complete the questionnaires in a private area removed from their friends and other Mardi Gras participants. Individuals were given condoms and lubricant packets as an incentive to participate. Data Analyses The TIB places personal characteristics such as gender, social class, and ethnicity external to the model, suggesting their influence on intentions and behavior is mediated through their influence on personal attitudes, normative beliefs, and social norms (Triandis, 1977, 1980). However, sexuality research has demonstrated interactions between gender and attitudes, norms, experiences, and behaviors, suggesting that gender may also influence intentions and behaviors from within the model (Maticka-Tyndale et al., 1998). Consequently, we tested for interactions between gender and all independent variables (IVs) before proceeding to test the full model. Given the limited sample size, the presence of interaction effects were tested for each IV by regressing the appropriate intention or behavior on the centered IV, gender, and the interaction between the centered IV and gender. IVs were centered on their means to minimize the effect of multicollinearity between the IV and the interaction term (Cohen, Cohen, West, & Aiken, 2003). Two significant interactions in each of the models predicted intentions to engage in sexual behavior (oral, vaginal, and anal sex) at Mardi Gras: between-gender and the cognitive component and between-gender and subjective social norms. Additionally, there was a significant interaction between gender and role beliefs in the model predicting intentions to engage in anal sex. The presence of these significant interaction terms in models predicting intentions necessitated separate data analysis for men and women for each of the three sexual behaviors. In the models predicting oral, vaginal, and anal sex behavior at Mardi Gras, only one significant interaction was present, specifically, between gender and perception of same-sex friends engaging in sexual behavior at Mardi Gras in the model predicting oral sex behavior. There were no significant interactions in the models predicting vaginal and anal sex; therefore, in models predicting engaging in sexual behavior at Mardi Gras, men's and women's data were analyzed together. The Triandis model, as presented in Figure 1, was tested using multiple linear regression analyses. First, affective and cognitive beliefs, subjective social norms, personal normative beliefs, and role beliefs were used to predict intentions to engage in oral, vaginal, and anal sex. Secondly, intentions, previous experience, situational conditions, and perceptions of peer sexual behavior at Mardi Gras were used to predict oral, vaginal, and anal sex. Regression assumptions, including the absence of multicollinearity, homogeneity of error variances, normality of residuals, and linearity, were assessed, and no serious violations were found. [FIGURE 1 OMITTED] Participants This article reports results of analyses based a sample of 300 Mardi Gras attendees in New Orleans, Louisiana, in 2004 (180 men, 120 women). The mean age was 24.9 years (SD = 3.81, range 18-35). The majority (90%, n = 267) identified as heterosexual, 7% identified as bisexual (n = 22), 1% identified as gay (n = 4), and 1% identified as uncertain or other or did not respond to the item (n = 7). Just less than half of all participants, 43%, had been to Mardi Gras previously (n = 128). Just over one third were students (n = 108) and the remaining two thirds were working professionals (n = 182). Thirty percent were single and not dating anyone (n = 89); 29% were casually dating one or more persons (n = 77). The remaining sample described themselves as being seriously involved with one person, engaged, or married (41%, n = 134). About one third (30%) of those in relationships indicated they had brought their relationship partner with them to Mardi Gras (n = 70). Overall, participants reported that they came to Mardi Gras with an average of 2.3 (SD = 2.61) same-sex friends and 1.8 (SD = 7.19) opposite-sex friends. Participants had been at Mardi Gras an average of 3.8 days (SD = 2.29) when they completed the questionnaire. Table 1 provides an overview of the characteristics of the participants. RESULTS Prior and Mardi Gras Sexual Experiences There was no gender difference in terms of lifetime vaginal sexual experience, with the majority of men and women (91%, n = 273) reporting that they had engaged in vaginal intercourse prior to coming to Mardi Gras. The men reported significantly more lifetime vaginal sexual partners than did women (16.4 and 8.7, respectively, t = 3.32; p = .003). Almost one third of participants (29%, n = 86) reported that they had engaged in anal intercourse before coming to Mardi Gras. The men also reported a significantly higher number of lifetime anal sex partners (4.3 and 1.6, respectively, t = 3.35; p = .009). In general, more men expected to engage in sexual behavior with someone that they met at Mardi Gras. More men (45%, n = 80) than women (14%, n = 17) reported that they had expected to have oral sex with someone they met at Mardi Gras ([X.sup.2][1,298] = 30.93, p = .0001). More men (42%, n = 74) reported intending to have vaginal sex with someone they met at Mardi Gras than did women (12%; [X.sup.2][1,298] = 30.81, p = .0001). Finally, 19% of men (n = 33) and 5% of women (n = 6) reported expecting to engage in anal sex with someone they met at Mardi Gras ([X.sup.2][1, 295] = 11.92, p = .0001). However, there was no significant gender difference in terms of actual sexual behavior at Mardi Gras in any of the three categories. Almost one fifth of participants (16%, n = 49) reported engaging in oral sex with someone they met at Mardi Gras. About one third (32%, n = 95) reported engaging in vaginal sex with someone they met at Mardi Gras. Six percent (n = 16) reported engaging in anal sex with someone they met at Mardi Gras. About 40% of respondents (n = 109) indicated that "at least a few" of their same-sex friends had sex with someone they met at Mardi Gras. Sixty-eight percent of the men (n = 120) and 63% of the women (n = 75) reported having at least 5 or 6 drinks per sitting at Mardi Gras. Almost one quarter of men (24%, n = 42) reported having at least 16 drinks per sitting; 15% of women (n = 18) reported the same. Over one third (37%, n = 102) indicated they were "always" drinking before "fooling around" in a sexual way with people that they met at Mardi Gras. One quarter (24%, n = 67) reported that they "sometimes" drank before fooling around in a sexual way. Predicting Intentions to Engage in Sexual Behavior at Mardi Gras Pearson product moment correlations between all variables used in the analysis ranged from .10 to .53. Subjective social norms were most highly correlated with intentions to engage in all three sexual behaviors with someone met at Mardi Gras (r = .38 for oral sex, r = .36 for vaginal sex, and r = .22 for anal sex, all significant at the p = .01 level). Cognitive beliefs were also similarly correlated with intentions regarding the three categories of sexual behaviors (r = .34 for oral sex, r = .32 for vaginal sex, and r = .25 for anal sex, all significant at the p = .01 level). Role beliefs were also correlated with intentions to engage in oral sex (r = .29), vaginal sex (r = .29), and anal sex (r = .20). Table 2 provides a summary of the descriptive statistics for the constructs of the TIB. For women, the TIB model did not significantly predict intentions to engage in oral, vaginal, or anal sexual behavior. However, the TIB did explain a significant portion of the variance in intentions for men. The TIB explained 20% of the variance in intentions to engage in oral sex with someone the men met at Mardi Gras (F[5,132] = 6.18, p < .001). Specifically, cognitive beliefs (b = .29, p = .005) and subjective social norms (b = .26, p = .005) predicted intentions to engage in oral sex with someone the male participants met at Mardi Gras. A similar pattern of results was found for intentions to engage in vaginal sex with someone the male participants met at Mardi Gras ([R.sup.2] =. 17, F[5,132] = 5.34, p < .001). Again, cognitive beliefs (b = .28, p = .006) and subjective social norms (b = .19, p = .044) predicted intentions to engage in vaginal sex with someone the male participants met at Mardi Gras. The TIB model accounted for the least amount of variance in intentions to engage in anal sex with someone the male participants met at Mardi Gras ([R.sup.2] = .09, F[5,131] = 2.61, p = .028). Only cognitive beliefs approached significance in terms of predicting intentions to engage in anal sex with someone the male participants met at Mardi Gras (b = .19, p = .078). Predicting Sexual Behavior at Mardi Gras Pearson product moment correlations between all variables used in this analysis ranged from -.003 to .62. The situational conditions related to the sexualized environment at Mardi Gras were most highly correlated with engaging in all three sexual behaviors with someone met at Mardi Gras (r =. 16 for oral sex, r = .31 for vaginal sex, and r = .17 for anal sex, significant at the p = .01 level). The perception that same-sex friends had sex at Mardi Gras was also correlated with engaging in oral (r = .24, p = .01) and vaginal sex (r = .28, p = .01). The TIB explained a small but significant amount of the variance in behaviors for men and women. The TIB explained 11% of the variance in oral sex behavior with someone the participants had met at Mardi Gras (F[6,234] = 4.46, p < .001). The perception that same-sex friends had sex with someone at Mardi Gras was the most significant predictor of engaging in oral sex at Mardi Gras (b =. 18, p = .005). Intentions to engage in oral sex with someone the participant met at Mardi Gras were also a significant predictor (b = .15, p = .031). The TIB explained almost one fourth of the variance in vaginal sex behavior at Mardi Gras ([R.sup.2] = .22, F[6,219] = 9.78, p < .001). The perception that same-sex friends had engaged in vaginal sex at Mardi Gras (b =. 19, p = .003), intentions to engage in vaginal sex at Mardi Gras (b = .17, p = .015), and previous vaginal sex experience (b = .17, p = .011) predicted engaging in vaginal sex with someone the men and women met at Mardi Gras. The TIB model accounted 14% of variance in anal sex behavior with someone the participants met at Mardi Gras ([R.sup.2] =. 14, F[6,221] = 5.95, p < .001). Previous anal sex experience (b = .23, p < .001) and situational characteristics associated with the Mardi Gras culture (b = .23, p =.006) significantly predicted anal sex behavior with someone the participants met at Mardi Gras. Figure 2 provides an overview of the results of the regression analyses. [FIGURE 2 OMITTED] DISCUSSION The purpose of this study was to assess the extent to which the Triandis Model of Interpersonal Behavior explained sexual decision-making at Mardi Gras. Given the extent to which this model has been valuable for explaining sexual behaviors at similar events that present intensive environmental conditions, this model appeared to be one of the most suitable for exploring Mardi Gras. This represents one of the first attempts to understand sexual decision-making at Mardi Gras from an empirical and theoretical perspective. While others have examined the event for its cultural and sociological value (Gotham, 2002; Redmon, 2003a; Shrum & Kilburn, 1996), virtually none have explored it specifically for gaining insight into the nature of the sexual behaviors that occur there. For its contributions to sex research, an understanding of sexual behavior in this context could provide further support for the extent to which environment influences the manner in which individuals construct their sexual lives. The primary outcome variables of interest in this study were participants' expectations related to sexual activity and participants' actual reports of sexual activity with someone they met at Mardi Gras. Given the history and reputation of the event, it was not surprising that the participants in this study, particularly the men, reported fairly high levels of behavioral expectations related to sexual activity with a new partner. Close to 50% of the men expected to participate in oral and vaginal intercourse with a new partner, and almost 20% expected to participate in anal intercourse with a new partner while at Mardi Gras. Although these expectations varied by gender, there were no significant gender differences when actual sexual behaviors at Mardi Gras were assessed. While men appeared to have overrated the likelihood that they would participate in certain sexual activities with a new partner at Mardi Gras, women tended to underrate the likelihood of the same. In this environment, it appears that women who expect to have sex are not challenged to do so, given the extent to which male partners are readily available. However, the disproportionate expectations of sexual activity between genders may create challenges for men in that there are simply fewer female partners with similar levels of sexual expectations. The Triandis Model of Interpersonal Behavior offered limited explanatory power for both behavioral intentions and actual behaviors for the men who participated in this study. The model explained significant amounts of variance in men's intentions to participate in oral, vaginal, and anal intercourse with a new partner. Specifically, cognitive beliefs and subjective social norms supporting sexual behavior at Mardi Gras were significant predictors of intentions to engage in these behaviors. However, models predicting intentions for women did not reach significance. Given recent research documenting the complex nature of women's sexuality and the many factors that influence arousal and behavior (Graham, Sanders, Milhausen, & McBride, 2004), the model simply did not account for the variables that influenced women's a priori expectations with regard to sexual behavior at Mardi Gras. Additionally, despite the liberal sexual environment that typifies Mardi Gras, perhaps women were concerned about issues related to reputation, consistent with the sexual double standard (Milhausen & Herold, 1999). For men and women, significant amounts of variance in participation in these sexual behaviors at Mardi Gras were explained predominantly by intentions to engage in these behaviors, engaging in similar behaviors previously, and perceptions of peer sexual behavior at Mardi Gras. In previous research, the extent to which a participant is exposed to the environmental factors of an event, measured by the situational conditions variable, has been consistently predictive of behaviors during an event. However, this was not the case for the participants at Mardi Gras, with the exception of anal intercourse reported by male participants. It may be that the intense situational conditions at Mardi Gras are experienced by a larger proportion of those in attendance at the event than would be expected of those attending other events, such as spring break. It may be the case that, given a limited amount of variance in the situational conditions at Mardi Gras (since they are so pervasive), it becomes increasingly difficult to detect statistical differences in the behaviors of interest. In previous research using the TIB in other venues, there was a greater deal of variance in the reported levels of exposure to the measured situational conditions. As a result, researchers may be more likely to detect a stronger level of association between one's score on a measure of situational conditions and one's reports of participation in a certain behavior. An additional reason for the limited explanatory value of the situational conditions variable may be related to the structural nature of Mardi Gras and our resulting study design. The majority of events and revelry occur within a confined geographic area. We only collected data from these central locations of Mardi Gras, given the likelihood that most attendees would pass through them on several occasions during the event. As a result, it may be the case that the vast majority of individuals in the central locations of Mardi Gras, and therefore those who came into contact with our research team, were all exposed to the situational conditions at similar levels. There may be individuals who attend Mardi Gras who purposefully avoid the most intense areas, thereby perhaps having less exposure to the situational conditions that we measured. These individuals would not have crossed paths with the research assistants, and their experiences may not be reflected in this study. It should also be noted that the elicitation phase of our research intentionally sought to gather information from individuals who had been actively involved in the more intense environments of Mardi Gras. This resulted in our ability to develop measures for the theoretical constructs that demonstrated relatively high levels of internal consistency in the actual study sample (with the exception of the cognitive beliefs construct, which had a reliability coefficient of .54). However, one of the challenges of using this model to study behaviors in unique situations such as Mardi Gras is that the researcher is dependent upon data (for the elicitation phase) that may be influenced by situations that elicitation phase participants find themselves in after the event. It could be that the cognitive evaluations of behavior as expressed by individuals in the focus groups were simply different than how those behaviors would be evaluated on the ground during Mardi Gras, contributing to the low reliability of this measure. This might be particularly true for women who are engaged in a focus group with other women. This low reliability on this particular construct may have contributed to the lack of consistency with which it was predictive of behavioral intentions for both men and women. Future research on sexual behaviors at similar events may benefit by sampling across a wider range of venues. However, for this study, the purpose was to understand the nature of behaviors that occurred within the context of the highly sexualized and alcohol-intensive areas of Mardi Gras. This may have contributed to the diminished explanatory power of the model that was used. Future studies could compare samples of individuals from other places outside the center of Mardi Gras with those more directly immersed in the Mardi Gras environment to determine if the latter participants were more likely to engage in sexual behaviors. The absolute revelry of the event is itself worthy of comment from a research perspective. Conducting sexuality-related research that required the collection of data from individuals in an environment of this nature was complex. The research team had underestimated the sheer intensity of the event and the extent to which following research protocols would be challenging. The research team found itself constantly needing to make subjective assessments as to the sobriety of individuals and their ability to provide informed consent. Given the crowds at Mardi Gras and the limited space, it was also challenging to help participants find enough room to complete the study instrument and retain a reasonable amount of confidentiality. Since events of this nature only occur once per year, it is difficult to conduct a true pilot study, particularly for the logistical information that such a study can help one to obtain. However, just as we used an elicitation phase for the purpose of measure development, researchers planning to conduct work in similar areas may want to elicit information from other researchers who have attended these events, either for personal or professional purposes, in order to gain insights into the challenges that one may face when conducting research using methods not typically employed in venues like Mardi Gras. For research conducted at similar events, researchers might also want to consider using more community-based participatory approaches, such as those suggested by Reece and Dodge (2004), which seek to involve community members actively as research partners. Additionally, future researchers interested in Mardi Gras might consider collecting participant contact information at the venue and then mailing surveys or e-mailing electronic links to web-based surveys to facilitate data collection outside the Mardi Gras environment. Another limitation of this study is that the sample contained almost exclusively individuals who described their sexual orientation as heterosexual. Future research at Mardi Gras should attempt to gather samples of both heterosexuals and homosexuals, particularly samples of gay men. There is a distinct area of activity at Mardi Gras that is popular mostly among gay men, and there may be important differences in the nature of behaviors that occur between the two environments. Despite these limitations, the TIB provided a useful framework for understanding men's intentions to engage in sexual activity at Mardi Gras, as well as their actual behavior. It appears that cognitive beliefs and subjective social norms are particularly predictive of intentions to engage in sexual behavior at Mardi Gras for men. Similarly, the perception of same-sex friends having sex with someone they met at Mardi Gras was the greatest predictor of engaging in oral and vaginal sex among both men and women. Interventions to decrease sexual risk-taking at Mardi Gras and other similar events might target perceptions of peer sexual behaviors. At both spring break and Mardi Gras, most men expected to engage in sexual behavior, but far fewer actually did (Maticka-Tyndale et al., 1998). It may be that men assume their male peers are engaging in more sexual activity than they actually are, creating an exaggerated perceived norm for sexual behavior. This study makes a significant contribution to our understanding of the sexual activity of young people at Mardi Gras, yet it also creates many questions. Foremost, although the TIB incorporated cognitive and environmental factors and took into account individuals' previous sexual experiences, it offered little explanatory value in terms of predicting intentions and actual behaviors at Mardi Gras. What is missing from the model that would help explain, to a greater degree, men's and women's experiences at Mardi Gras? Also, what type of sexual health programs would be most effective at decreasing levels of sexual risk-taking, particularly those that occur following high rates of alcohol use, at Mardi Gras by both genders? The TIB and other models that assess environmental influences on sexual behavior afford researchers a more complete understanding of the complexity of sexual decision-making in these unique environments, and additional research will help us understand these behaviors better and identify those intervention mechanisms that are appropriate and effective. Note. We wish to acknowledge the natural disaster that impacted the city of New Orleans in late summer 2005 and extend our sympathies for the losses that the city has incurred. We hope the rich cultural traditions of the city, among them Mardi Gras, will soon return with vibrancy and pride. We would also like to acknowledge the members of the Sexual Health Research Working Group at Indiana University for their assistance with this project, particularly Forest King, Erin Hoschouer, Erin Brown, and Jacynda Lacy for their assistance with data collection. Manuscript accepted August 2, 2005 REFERENCES Apostolopoulos, Y., Sonmez, S., & Yu, C. H. (2002). HIV-risk behaviours of American spring break vacationers: A case of situational disinhibition? International Journal of STD & AIDS, 13, 733-743. Bamberg, S., & Schmidt, P. (2003). Incentives, morality, or habit? Predicting students' car use for university routes with models of Ajzen, Schwartz, and Triandis. Environment and Behavior, 35, 264-285. Boyd, B., & Wandersman, A. (1991). Predicting undergraduate condom use with Fishbein and Ajzen and the Triandis attitude-behavior models: Implications for public health interventions. Journal of Applied Social Psychology, 21, 1,810-1,830. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd Edition). Mahwah, NJ: Lawrence Erlbaum Associates. Crowe, L. C., & George, W. H. (1989). Alcohol and human sexuality: Review and integration. Psychological Bulletin, 105(3), 374-286. Eiser, J. R., & Ford, N. J. (1995). Sexual relationships on holiday: A case of situational disinhibition. Journal of Social and Personal Relationships, 12, 323-339. Ford, N. J., & Eiser, J. R. (1996). Risk and liminality: The HIV-related socio-sexual interaction of young tourists visiting Torbay. In S. Clift and S. J. Page (Eds.), Health and the international tourist (pp. 152-178). London: Routledge. Forsyth, C. J. (1992). Parade strippers: A note on being naked in public. Deviant Behavior, 13, 391-403. Godin, G., Adrien, A., Willms, D., Maticka-Tyndale, E., Manson-Singer, S., & Cappon, P. (1996). Cross-cultural testing of three social cognitive theories: An application to condom use. Journal of Applied Social Psychology, 26, 1,556-1,586. Goffman, E. (1959). The presentation of self in everyday life. New York: Doubleday. Gotham, K. F. (2002). Marketing Mardi Gras: Commodification, spectacle and the political economy of tourism in New Orleans. Urban Studies, 39 (10), 1,735-1,756. Graham, C. A., Sanders, S. A., Milhausen, R. R., & McBride, K. R. (2004). Turning on and turning off: A focus group study of the factors that affect women's sexual arousal. Archives of Sexual Behavior, 33, 6, 527-538. Jankowiak, W., & White, C. T. (1999). Carnival on the clipboard: An ethnological study of New Orleans Mardi Gras. Ethnology ethnology /eth·nol·o·gy/ (eth-nol´ah-je) the science dealing with the major cultural groups of humans, their descent, relationship, etc., 38, 335-349. Katz, J. (1988). Seductions of crime. New York: Basic Books. 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(5), 1,010-1,027. Maticka-Tyndale, E., Herold, E. S., & Mewhinney, D. (1998). Casual sex on spring break: Intentions and behaviors of Canadian students. The Journal of Sex Research, 35(3), 254-264. Maticka-Tyndale, E., Herold, E. S., & Oppermann, M. (2003). Casual sex among Australian schoolies. The Journal of Sex Research, 40(2), 158-169. Mewhinney, D. M., Herold, E. S., & Maticka-Tyndale, E. (1995). Sexual scripts and risk-taking of Canadian students on spring break in Daytona Beach, Florida. Canadian Journal of Human Sexuality, 4(4), 273-288. Milhausen, R. R., & Herold, E. S. (1999). Does the sexual double standard still exist? Perceptions of university women. The Journal of Sex Research, 36, 361-368. Redmon, D. (2002). Testing informal social control theory: Examining lewd behavior at Mardi Gras. Deviant Behavior, 23, 363-384. Redmon, D. (2003a). Playful deviance as an urban leisure activity: Secret selves, self-validation, and entertaining performances. Deviant Behavior, 24, 27-51. Redmon, D. (2003b). Examining low self-control theory at Mardi Gras: Critiquing the general theory of crime within the framework of normative deviance. Deviant Behavior, 24, 373-392. Reece, M., & Dodge, B. (2004). A study in sexual health using the principles of community-based participatory research. Archives of Sexual Behavior, 33(3), 235-247. Sexton, R. L. (1999). Cajun Mardi Gras: Cultural objectification and symbolic appropriation in a French Tradition. Ethnology, 38(4), 297-313. Sexton, R. L. (2001). Ritualized inebriation inebriation /in·e·bri·a·tion/ (in-e?bre-a´shun) drunkenness; intoxication with, or as if with, alcohol. in·e·bri·a·tion (  n-, violence, and social
control in Cajun Mardi Gras. Anthropological Quarterly, 74(1), 28-38.Shields, R. (1990). The system of pleasure: Liminality and the carnivalesque at Brighton. Theory, Culture, and Society, 7, 39-72. Shrum, W., & Kilburn, J. (1996). Ritual disrobement at Mardi Gras: Ceremonial exchange and moral order. Social Forces, 75, 423-458. Smith, M. A., & Rosenthal, D. (1997). Sex, alcohol and drugs? Young people's experience of Schoolies Week. Australian and New Zealand Journal of Public Health, 21(2), 175-180. Thomas, M. (2000). Exploring the contexts and meanings of women's experiences of sexual intercourse on holiday. In S. Clift and S. Carter (Eds.), Tourism and sex: Culture, coercion, and commerce. New York: Pinton. Triandis, H. C. (1977). Interpersonal behavior. Monterey, CA: Brooks/Cole. Triandis, H. C. (1980). Values, attitudes, and interpersonal behavior. In H. Howe & M. Page (Eds.), Nebraska symposium on motivation 1979 (pp. 195-295). Lincoln, NE: University of Nebraska Press. Triandis, H. C. (1994). Culture and social behavior. New York: McGraw Hill Inc. Robin R. Milhausen University of Windsor, Canada Michael Reece Indiana University Bilesha Perera University of Ruhuna, Sri Lanka Address correspondence to Michael Reece, Ph.D., MPH, Department of Applied Health Science, Indiana University, Bloomington, IN, 47405; e-mail: mireece@indiana.edu.
Table 1. Participant Characteristics
Men Women
(N = 180) (N = 120)
Age (a) 25.67 (SD = 3.96) 23.80 (SD = 3.27)
Sexual Orientation (%) (a)
Heterosexual 96.1 81.2
Homosexual 1.7 16.2
Bisexual 1.1 1.7
Uncertain/Did not respond 1.1 .9
Previous Mardi Gras
Experience? (%) 44.4 40.8
Occupation (%) (a)
Student 29.5 48.7
Professional 70.5 51.3
Relationship Status (%) (a)
Single, not dating 33.3 25.2
Casually dating 33.3 23.6
Seriously involved,
engaged, or married 33.4 51.2
Relationship Partner at Mardi
Gras? (%) (a) 24.8 36.9
Number of Same-Sex Friends at
Mardi Gras (mean) (a) 2.62 (SD = 2.69) 1.85 (SD = 2.39)
Number of Opposite-Sex
Friends at Mardi Gras
(mean) (a) .99 (SD = 1.83) 3.04 (SD = 11.42)
Lifetime Vaginal Sex
Experience? (%) 93.8 90.8
Number of Lifetime Vaginal
Sex Partners (mean) (a) 16.81 (SD = 25.33) 9.23 (SD = 13.77)
Lifetime Anal Sex Experience?
(%) 30.5 26.9
Number of Lifetime Anal Sex
Partners (mean) (a) 4.33 (SD = 5.64) 1.59 (SD = 1.34)
Overall Sample
(N = 300)
Age (a) 24.92 (SD = 3.81)
Sexual Orientation (%) (a)
Heterosexual 90.2
Homosexual 7.4
Bisexual 1.4
Uncertain/Did not respond 1.0
Previous Mardi Gras
Experience? (%) 43.0
Occupation (%) (a)
Student 37.2
Professional 62.8
Relationship Status (%) (a)
Single, not dating 30.1
Casually dating 29.4
Seriously involved,
engaged, or married 40.5
Relationship Partner at Mardi
Gras? (%) (a) 29.9
Number of Same-Sex Friends at
Mardi Gras (mean) (a) 2.33 (SD = 2.61)
Number of Opposite-Sex
Friends at Mardi Gras
(mean) (a) 1.76 (SD = 7.19)
Lifetime Vaginal Sex
Experience? (%) 92.5
Number of Lifetime Vaginal
Sex Partners (mean) (a) 13.87 (SD = 21.87)
Lifetime Anal Sex Experience?
(%) 29.1
Number of Lifetime Anal Sex
Partners (mean) (a) 3.29 (SD = 4.69)
(a) Significant gender differences at the p < .05 level.
Table 2. Descriptive Statistics for Triandis Model Constructs
Mean (SD)
Men Women
Affective Component 45.09 (8.94) 44.66 (9.96)
Cognitive Component (a) 25.88 (5.27) 22.16 (6.18)
Personal Normative Beliefs (a) 8.17 (4.18) 6.96 (3.84)
Subjective Social Norms (a) 15.83 (4.67) 11.02 (5.36)
Role Beliefs (a) 16.79 (4.85) 13.69 (5.85
Situational Conditions (Alcohol) (a) 18.26 (2.90) 17.29 (3.43)
Situational Conditions (Sex) 20.98 (5.44) 20.33 (5.61)
Situational Conditions 11.60 (4.51) 11.70 (4.53)
(Mardi Gras Environment)
Range
Total Total
Affective Component 44.92 (9.34) 10-70
Cognitive Component (a) 24.34 (5.94) 5-35
Personal Normative Beliefs (a) 7.67 (4.08) 2-14
Subjective Social Norms (a) 13.85 (5.50) 3-21
Role Beliefs (a) 15.54 (5.48) 3-21
Situational Conditions (Alcohol) (a) 17.87 (3.15) 5-20
Situational Conditions (Sex) 20.72 (5.51) 9-30
Situational Conditions 11.64 (4.51) 5-20
(Mardi Gras Environment)
N
Men Women Total
Affective Component 147 92 239
Cognitive Component (a) 147 103 250
Personal Normative Beliefs (a) 163 112 275
Subjective Social Norms (a) 162 113 275
Role Beliefs (a) 163 110 273
Situational Conditions (Alcohol) (a) 170 113 283
Situational Conditions (Sex) 154 103 257
Situational Conditions 172 113 285
(Mardi Gras Environment)
(a) Significant gender differences at the p < .05 level.
|
|
||||||||||||||||||
Printer friendly
Cite/link
Email
Feedback
Reader Opinion