High-risk recreation: the relationship between participant characteristics and degree of involvement.
While originally thought of as activities suited only for explorers or daredevils, high-risk recreational pursuits are now sought out by millions of people throughout North America and the world (Ewert, 1989). Terms frequently used to describe these activities include "rugged recreation," "high adventure programming," "thrill sports," and "outdoor pursuits" (Meier, 1978). Throughout this manuscript, however, the term "risk recreation" will be used to describe these endeavors (Ewert & Hollenhorst, 1989). Examples of activities commonly associated with this term include rock climbing, whitewater kayaking and rafting, skydiving, mountaineering, scuba diving, and paragliding.
Models used to explain and predict risk recreational behavior are an integral part of the study of this phenomenon (Schuett, 1993). Using his Adventure Model of risk recreation, Ewert (1987) outlines various participant characteristics and patterns of use regarding risk recreational activities. A review of this research, along with other studies focusing on risk recreation, however, revealed two voids. First, previous studies (e.g., Allen, 1980; Ewert, 1987; Ewert & Hollenhorst, 1989; Robinson, 1992) were limited in their explanation of risk recreation because of methodological drawbacks such as sample design, lack of theoretical foundation, limited field testing, and inconsistent conclusions concerning results (Schuett, 1993). Second, researchers had not yet examined selected predictors of involvement in risk recreational activities. The purpose of the present study was to fill these two research voids. In seeking to promote an understanding of the complex nature of risk recreation participation, this study examined three sets of variables leading to the prediction of high-risk recreation involvement. These variables were selected based on past research targeting personality/demographic variables and/or high-risk recreation participation. The first set involved mortality-related variables, whereas the second and third set focused on psychological and demographic variables, respectively.
Included in the first set, and believed to have played an important role in predicting degree of involvement, was the participant's death anxiety. Becker (1962) hypothesized that human survival has been facilitated by our intelligence, particularly our ability to think in abstract terms, anticipate future events, imagine what does not yet exist, and then realize such possibilities. With this increase in intelligence, however, comes an increased focus on the self. This self-consciousness enables us to focus on our very existence and the associated existential burdens of being aware of our own death and nonexistence.
Given this, it is believed that one method used to achieve control (or at least the illusion of control) over one's mortality is cheating death. Among the many behaviors that allow an individual to cheat death (e.g., using drugs, unsafe sex, etc.), it seems logical that participation in high-risk recreation may provide one method of cheating death and, therefore, achieve the aforementioned illusion of control over one's mortality.
Sensation-seeking was also thought, and has been shown, to have played a part in tendencies to engage in high-risk recreation (e.g., Ewert, 1989; Levenson, 1990; Robinson, 1985; Schuett, 1993; Straub, 1982; Zuckerman, 1983). Zuckerman (1979) defines sensation-seeking as the need for varied, novel, and complex sensations with a willingness to take physical and psychological risks for the sake of those experiences. Logically, participation in high-risk recreation may provide an individual with the aforementioned sensations and experiences.
A number of psychological variables were also believed to have played a significant role in the decision to participate in risk recreation. The first was the participant's locus of control. Research by Ewert and Hollenhorst (1989) suggests that as the level of engagement increases, locus of control shifts from external sources (i.e., group leader) to those more internal in nature (i.e., the individual). One methodological shortcoming of this research, however, was that locus of control was not assessed by any sound psychometric measure. The present study addressed this shortcoming.
Perceived physical self-efficacy was also thought to have played an important part in predicting degree of involvement. Bandura (1977) defines self-efficacy as the strength of an individual's perceived self-confidence that he or she can successfully complete a task (i.e., social, cognitive, and/or physical) through the expression of ability. Several theoretical reasons suggest that a relationship between perceived physical self-efficacy and high-risk recreation involvement exists. Koocher (1971) reported that the acquisition of a leisure or sport skill leads to a significant increase in perceived physical self-efficacy. In support, Marsh, Richards, and Barnes (1986, 1987) found that the Outward Bound Program was particularly powerful in increasing participants' self-efficacy. Therefore, the use of high-risk recreation may provide an individual with an increase in perceived physical self-efficacy.
Self-presentational processes, a third psychological variable, was also believed to have played an important role. Self-presentation refers to the processes by which people monitor and control how they are perceived by others (Schlenker, 1980). Past research has shown that people hold certain preconceptions about individuals who play or participate in certain recreational activities (Sadalla, Linder, & Jenkins, 1988). Given this, people's choices of physical activities may be affected by their perceptions of the self-presentational implications of participation in various activities (Leary, 1992). So, with today's cultural push (i.e., television commercials and increased high-risk recreation coverage) to participate in various types of risk recreation, it was thought that individuals may participate in these activities because of the aforementioned self-presentational processes.
Social-complexity, the final variable in this set, was also thought to have played a significant role in risk recreation involvement. Based on Linville's theory of self-complexity (1987), Wann and Hamlet (1994) have proposed a theory of social-complexity, or the extent to which individuals attempt to join and maintain memberships in a number of diverse groups. Following their model, it was thought that social-complexity may have been related to degree of involvement in high-risk recreation. Individuals scoring higher in social-complexity, referred to as "joiners," tend to seek opportunities to increase their social-complexity. Individuals who are higher in their degree of involvement, along with the associated increases in social involvement with other highly identified high-risk recreation participants, should show an increase in social-complexity.
A number of demographic variables were also thought to have been significantly related to degree of involvement. First, many risk recreational activities require that participants maintain a certain socioeconomic level in order to purchase, maintain, upgrade, and/or rent needed equipment, as well as travel to and from locations that are conducive to participating in risk recreational activities. It was believed, therefore, that socioeconomic status may have been an important predictor and a variable to be studied in greater depth.
Finally, gender was believed to have played an important role in predicting risk recreation participation. Numerous studies assessing gender effects on degree of involvement suggest that the risk taker is usually relatively young, middle class, and male (Ewert, 1985; Ewert & Hollenhorst, 1989; Iso-Ahola, LaVerda, & Graefe, 1988; Klausner, 1968; McIntyre, 1989; Schuett, 1993). Dunn and Gulbis's (1976) research lends support for these findings; however, their results suggest that males, compared to females, tend to participate more in air activities (e.g., skydiving, hang-gliding, etc.) and self-defense activities (e.g., karate, boxing, judo, etc.), whereas land activities (e.g., rock climbing, spelunking, mountain climbing, etc.), water activities (e.g., kayaking, surfing, scuba diving, etc.), and snow activities (e.g., downhill skiing, hot dog skiing, etc.) are participated in by both sexes equally (Dunn & Gulbis, 1976). When examining gender and involvement in high-risk recreation, past research, although somewhat inconsistent, reveals that males are more likely to participate in high-risk recreation.
Thus, based on the aforementioned literature, the following variables were considered predictors of the participant's involvement in high-risk recreation: death anxiety, level of sensation-seeking, perceived physical self-efficacy, locus of control, self-presentational style, social-complexity, socioeconomic status, and gender. In addition, it was hypothesized that death anxiety and gender, alone or in combination, would be the best predictors of involvement in high-risk recreation. It was also hypothesized that death anxiety, after holding the other predictor variables constant, would account for a significant proportion of variability above and beyond these seven predictor variables.
Of the students (87 men, 51.5%, and 82 women, 48.5%; mean age = 21.8 years) who volunteered as participants for this study, 29.6% reported being high school graduates, 60.9% reported having some college, 8.3% reported being college graduates, and 1.2% reported having postgraduate degrees. Individuals were recruited, using a sign-up sheet posted at a predetermined location, from a pool of social science students receiving extra credit for their participation. No individuals were excluded from participating in the study. The respondents' socioeconomic status (students' and/or parents' annual income) ranged from $10,000 or below to over $50,000 with the majority making over $50,000.
Socioeconomic status and gender. A demographics questionnaire was used to assess socioeconomic status and gender. Items assessed the participant's gender and household income. Students under parental financial security were asked to respond with parental income. In addition to these two variables, the participant's age, the education level of both parents, and the education level of the participant were assessed to gain further demographic information about the sample.
Risk recreation. The seven-item risk recreation questionnaire, based on Ewert's (1987) Adventure Model of risk recreation, was used to assess involvement in risk recreation. The items were in Likert format ranging from one to nine and were designed to elicit responses concerning various indices that help define degree of involvement in risk recreation. It should be noted that Ewert's (1987) risk recreation questionnaire divided the risk recreation involvement continuum into three distinct levels. Those reporting scores from I to 3 were grouped as introductory, 4 to 6 as development, and 7 to 9 as commitment. In this study, however, scores were not grouped into engagement levels but were measured on an involvement continuum.
Death anxiety. Templer's (1970) 15-item Death Anxiety Scale (DAS) was used to measure level of verbalized death anxiety. Templer (1970) states that the DAS internal and retest reliabilities are high. A reliability alpha of .76 demonstrates reasonable internal consistency and a product moment correlation coefficient of .83 demonstrates reasonable 3-week retest reliability.
Sensation-seeking. To assess sensation-seeking, Zuckerman's (1979) Sensation-Seeking Scale (SSS) was used. The most recent form (Form V) of the SSS is made up of four factors each consisting of ten items for a total of 40 items. These factors include:
1. Thrill and Adventure Seeking (TAS) which is defined as the seeking of sensation through risky but exciting sports and other activities such as fast driving.
2. Experience Seeking (ES) which is defined as seeking sensation through the mind and the senses and through a nonconforming lifestyle.
3. Disinhibition (DIS) which is defined as the seeking of sensation through social stimulation and disinhibition through social drinking.
4. Boredom Susceptibility (BS) which is defined as an aversion to monotonous, invariant situations and restlessness when exposed to new situations.
Zuckerman (1984) reports that the SSS internal and retest reliabilities are satisfactory, especially for the total score for which internal reliabilities are .85, and 3-week retest reliabilities are .94. Subscale internal reliabilities range from .60 (for the BS scale) to .80 (for the TAS scale).
Internal vs. external locus of control. To assess locus of control, Rotter's (1966) 29-item Internal-External Locus of Control Scale was used. The items are in a forced-choice format and measure individual differences in generalized expectancy for internal-external control. Rotter (1966) reports an internal consistency of .71, and a 4-week retest reliability of .72. These results indicate a satisfactory reliability for the scale. Rotter (1966) also stated that there is sufficient evidence for both construct and discriminant validity for the instrument.
Perceived physical self-efficacy. To assess perceived physical self-efficacy, Ryckman, Robbins, Thornton, and Cantrell's (1982) 22-item Physical Self-Efficacy Scale was used. The items are in Likert format ranging from I (strongly agree) to 6 (strongly disagree). The Physical Self-Efficacy Scale items require the respondent to report generalized expectancies concerning their perceived competence in performing tasks involving the use of physical skills and their level of confidence in displaying these skills and having them evaluated by others. Ryckman et al. (1982) report an internal consistency of .81, and a retest reliability of .80. These results indicate a satisfactory reliability for the scale. Ryckman et al. also conclude that there is sufficient evidence for both concurrent and predictive validity for the instrument.
Self-presentational processes. To assess self-presentational processes, participants completed an 11-item Self-Presentational Style Inventory developed specifically for this study. The items are in Likert format and range from 1 (strongly disagree) to 9 (strongly agree). The scale targeted such self-presentational topics as physical and social identity, and physical and social image. The overall structure of the Self-Presentational Style Inventory, using a principal components factor analysis (with varimax rotation), revealed two factors labeled Physical/Social Image and Physical/Social Presentation Reward. The seven-item Physical/Social Image Factor had an eigenvalue of 5.82 and accounted for 52.9% of the variance. The four-item Physical/Social Presentation Reward Factor had an eigenvalue of 1.24 and accounted for 11.2% of the variance. The two factors, therefore, combined to account for 64.1% of the total variance. Scale reliability, using Cronbach's index of internal consistency, showed that the Physical/Social Image (alpha = .89) and the Physical/Social Presentation Reward (alpha = .84) subscales were reliable.
Social-complexity. To assess social complexity, Wann and Hamlet's (1994) seven-item Joiners' Scale was used. The items are in Likert format and range from 1 (strongly disagree) to 9 (strongly agree) and target individual differences in social-complexity. Wann and Hamlet (1994) report an internal consistency of .81, and a retest reliability of .85. These results indicate a satisfactory reliability for the scale. Warm and Hamlet also conclude that there is sufficient evidence for both criterion and discriminant validity for the instrument.
During the testing session, participants were seated as a group in a large room and asked by the experimenter to read and sign an informed consent form. Assurances were given that their responses would remain anonymous, and that they could discontinue the study at any time without penalty. After signing the informed consent form, participants were given the packet containing the aforementioned questionnaires. All participants received the same questionnaires; however, high-risk recreation participants (see Table 1 for a list of high-risk activities) were instructed in the directions to answer additional questions on the risk recreation questionnaire (in addition to those answered by high-risk recreation nonparticipants). When completed, participants were instructed to insert ali the questionnaires back in the packet upon completion, and then return the questionnaire packet to the experimenter. The average time to complete the questionnaire packet was 40 minutes. When completed, participants were thanked for their participation, and given their extra credit slips and debriefing statements which explained the nature of the study and listed a phone number and a date when the results would be available.
Predicting Degree of Involvement
Because this was a correlational study, potential confounds were controlled statistically using multiple regression analyses. As noted above, high-risk recreation participants answered additional questions on the risk recreation questionnaire; therefore, two separate sets of regression analyses were conducted. The first set focused on high-risk recreation participants and nonparticipants together, whereas the second set focused on high-risk recreation participants separately. Both sets of analyses focused on the relationship between the predictor variables of death anxiety, level of sensation-seeking, perceived physical self-efficacy, locus of control, social-complexity, self-presentational style, socioeconomic status, and gender and the criterion variable of degree of involvement in high-risk recreation (see Table 2 for means and standard deviations for ali participants, high-risk recreation participants, and high-risk recreation nonparticipants).
Table 1 High-Risk Activities and Number of Participants High-Risk Activities n In-line skating 28 Whitewater rafting 25 Paragliding 1 Alpine skiing 18 Whitewater kayaking 2 Waterskiing 2 Mountaineering 3 Rodeo cow-poking 1 Surfing 7 Cliff-diving 2 Spelunking 14 Seadoo racing 1 Rappelling 4 Parasailing 5 Go-cart racing 1 Ice climbing 1 Bungee jumping 18 Mountain biking 22 SCUBA diving 15 Bicycle motocross (BMX) 1 Bull-riding 1 Windsurfing 2 Hang-gliding 1 Skateboarding 19 Motorcycle racing 11 Indoor rock climbing 1 Outdoor rock climbing 13 Motocross 4 Skydiving 1 B.A.S.E. jumping 2 Rock jumping 1 Snow climbing 1 Note. Various individuals participated in more than one high-risk activity.
The first of these analyses, using multiple regression, examined the weighted linear combination of the eight predictor variables. Together these variables accounted for 39% of the variability in degree of involvement in high-risk recreation in ali individuals (high-risk recreation participants and nonparticipants), F(8, 141) = 11.01, p [less than] .0001 (see Table 3). When [TABULAR DATA FOR TABLE 2 OMITTED] high-risk recreation participants were analyzed separately from nonparticipants, the aforementioned predictor variables accounted for 22% of the variability in degree of involvement in high-risk recreation, F(8,72) = 2.50, p [less than] .05 (see Table 4).
The second analysis, a stepwise multiple regression, was used to determine which combination of the eight predictor variables provided the best prediction of degree of involvement in high-risk recreation. Significance level for entry into the model was set at p [less than] .05. For all individuals, the only predictor variables that accounted for a significant proportion of the variability in degree of involvement was gender, level of sensation-seeking, and social-complexity. When high-risk recreation participants were analyzed separately from nonparticipants, the only predictor variable that accounted for a significant proportion of the variability in degree of involvement was gender (see Table 5).
The third analysis examined the relationship between degree of involvement in high-risk recreation and death anxiety after statistically controlling for the other predictor variables. This analysis determined if the increment in variability accounted for after adding death anxiety to the predictor variables was significant. The results of a hierarchical multiple regression revealed that the addition of death anxiety into the regression equation, after holding the other predictor variables constant, did not result in a significant increment in variability accounted for in degree of involvement in high-risk recreation in ali individuals, F(1,141) = 2.71, p [greater than] .10, and when high-risk recreation participants were analyzed separately from nonparticipants, F(1,72) = .06, p [greater than] .80.
Table 3 Summary of Standard Multiple Regression Analysis for Variables Predicting Degree of Involvement in High-Risk Recreation for All Individuals Variable B SE [Beta] TOTSS 0.14 0.04 0.25 TOTPPSE - 0.03 0.03 - 0.06 TOTINEXT 0.04 0.07 0.05 TOTJOIN 0.12 0.05 0.18 TOTSPS 0.02 0.02 0.06 INCOME2 0.27 0.15 0.12 GENDER - 3.05 0.59 - 0.38 TOTDEATH - 0.15 0.09 - 0.12 Note. [R.sup.2] = .39; TOTSS = level of sensation-seeking; TOTPPSE = perceived physical self-efficacy; TOTINEXT = locus of control; TOTJOIN = social-complexity; TOTSPS = self-presentational style; INCOME2 = income; GENDER = gender; TOTDEATH = death anxiety.
The simple correlations between selected variables were examined using the Pearson product moment correlation coefficient (except for gender where a point biserial correlation was utilized). For all individuals, the only predictor variables significantly correlated with degree of involvement in high-risk recreation were level of sensation-seeking, r(168) = .39, p [less than] .0001, social complexity, r(169) = .30, p [less than] .0001, socioeconomic status, r(153) = .23, p [less than] .005, and death anxiety, r(169) = -.22, p [less than] .005. When high-risk recreation participants were analyzed separately from nonparticipants, the only predictor variable significantly correlated with degree of involvement in high-risk recreation was sensation-seeking, r(87) = .24, p [less than] .05. Using a point biserial correlation, gender was also found to be significantly correlated with degree of involvement in high-risk recreation in ali individuals, r(169) = -.39, p [less than] .0001, as well as when high-risk recreation participants were analyzed separately from nonparticipants, r(87) = -.29, p [less than] .01. It should also be noted that different N's were due to missing values which were not included in the relevant analyses.
Table 4 Summary of Standard Multiple Regression Analysis for Variables Predicting Degree of Involvement for High-Risk Recreation Participants Variable B SE [Beta] TOTSS 0.26 0.17 0.17 TOTPPSE - 0.24 0.14 - 0.19 TOTINEXT 0.28 0.27 0.11 TOTJOIN 0.13 0.20 0.08 TOTSPS 0.01 0.08 0.01 INCOME2 0.95 0.62 0.16 GENDER - 7.27 2.49 - 0.34 TOTDEATH - 0.10 0.38 - 0.03 Note. [R.sup.2] = .22; TOTSS = level of sensation-seeking; TOTPPSE = perceived physical self-efficacy; TOTINEXT = locus of control; TOTJOIN = social-complexity; TOTSPS = self-presentational style; INCOME2 = income; GENDER = gender; TOTDEATH = death anxiety.
The first hypothesis, that death anxiety and gender would be the best single predictors of involvement in high-risk recreation, was not fully supported. It was found that level of sensation-seeking and gender were the two variables that accounted for the greatest proportion of variability in high-risk recreation involvement for all individuals, and when high-risk participants were analyzed separately from nonparticipants.
The second hypothesis, that the best prediction of involvement in high-risk recreation would be provided by a combination of death anxiety and gender, was not fully supported. It was found that for ali individuals, the best combination was gender, level of sensation-seeking, and social-complexity. When high-risk recreation participants were analyzed separately, gender, in combination with no other variables, was shown to significantly predict the degree of involvement.
Table 5 Summary of Stepwise Regression Analyses for Variables Predicting Degree of Involvement in High-Risk Recreation High-Risk Participants and Nonparticipants Variable B SE [Beta] [R.sup.2] Step 1 GENDER - 3.60 0.58 - 0.21 Step 2 TOTSS 0.18 0.04 - GENDER - 3.08 0.56 - 0.31 Step 3 TOTSS 0.15 0.04 0.27 TOTJOIN 0.14 0.04 0.22 GENDER - 3.17 0.54 - 0.40 0.35 High-Risk Participants Step 1 GENDER - 7.06 2.31 - 0.33 0.11 Note. TOTSS = level of sensation-seeking; TOTJOIN = social-complexity; GENDER = gender.
The third hypothesis, that death anxiety, after holding the other predictor variables constant, would account for a significant proportion of variability above and beyond the other seven predictor variables, was not supported. It was found that for all individuals, and when high-risk recreation participants were analyzed separately, the addition of death anxiety into the regression equation did not significantly contribute to the prediction of involvement in high-risk recreation.
One possible explanation for the above findings regarding death anxiety may be due to the relatively young age of the sample as a whole (21.8 years). The mean level of death anxiety was low and did not show a large degree of variability (M = 6.86, SD = 3.24, on a 15 point scale). Perhaps young people, as a whole, exhibit a significant level of denial in regards to their own death, regardless of what degree they participate in risk recreation (Westman, Canter, & Boitos, 1984). Future research should attempt to investigate this question with a more representative sample in terms of age.
Another possible explanation for the findings regarding death anxiety involves the religious beliefs of the sample. As a whole, especially in this region of the country (i.e., Kentucky), individuals tend to exhibit higher levels of religiosity (United States Bureau of the Census, 1996). This, in turn, may provide increased hope of immortality and, hence, decrease anxiety about death (Solomon, Greenberg, & Pyszczynski, 1991). In the future, the religiosity of the sample should be controlled for so the effects of death anxiety can be examined more closely in terms of participation in high-risk recreation.
Gender, on the other hand, was found to be a significant predictor of involvement, alone and in combination with other predictor variables. Consistent with past research (Ewert, 1985; Ewert & Hollenhorst, 1989; Iso-Ahola et al., 1988; Klausner, 1968; McIntyre, 1989; Schuett, 1993), the findings of the present study found that a greater proportion of males (62.4%), compared to females (37.4%), participate in high-risk recreation. Gender, in and of itself, however, may not be the best way to approach the underlying question. Perhaps a better method would be to examine the individual's level of masculinity and femininity. Rather than just being "male" or "female," perhaps measuring the respondent's level of masculine traits and feminine traits would provide a different, and more informative perspective in regards to participation in high-risk recreation. Research examining masculinity vs. femininity was stimulated by the development of the Bem Sex Role Inventory (Bem, 1974). This particular scale measures the degree to which an individual exhibits feminine, masculine, or neutral personality characteristics. based on their responses to these items, a classification of masculine, feminine, androgynous, or undifferentiated is assigned. By examining gender in this manner, perhaps we can gain a clearer understanding of the aforementioned relationship.
Regarding locus of control, earlier research conducted by Ewert and Hollenhorst (1989) was found to contain various methodological shortcomings. More recent studies that were conducted to overcome these shortcomings found that locus of control did not significantly contribute to the prediction of involvement in high-risk recreation (Schuett, 1993). The present study was also an attempt to overcome the above methodological shortcomings. The findings, however, were no different in that locus of control did not significantly contribute to the prediction of involvement in high-risk recreation.
Another area of concern is that many individuals participate in multiple high-risk recreational activities (see Table 1). Although the present study did not examine this aspect of risk recreation participation, future research should attempt to investigate the impact of multi-sport involvement as it relates to various predictors. In addition, it should be noted that a number of high-risk activities vary in their degree of risk. For example, in-line skating can be considered less dangerous than scuba diving. Future research should also attempt to address this as it may have a significant impact on the prediction of involvement in high-risk recreation.
In conclusion, one's gender and level of sensation-seeking, rather than gender and death anxiety, appear to be the best single predictors of degree of involvement in high-risk recreation. It also appears that the best prediction of involvement in high-risk recreation is provided by a combination of gender, level of sensation-seeking, and social-complexity, rather than gender and death anxiety, at least when high-risk recreation participants and nonparticipants are analyzed together. When high-risk recreation participants are analyzed separately from nonparticipants, however, gender, in combination with no other variables, is shown to significantly predict the degree of involvement. In addition, it appears that death anxiety, after holding the other predictor variables constant, does not account for a significant proportion of variability above and beyond the other seven predictor variables. It should also be noted that the above explanations are based on post hoc analyses, and that future research in these areas should maintain an empirical nature.
Allen, S. D. (1980). Risk recreation: Some psychological bases for attraction (Doctoral dissertation, University of Montana, 1980). Dissertation Abstracts International, 41, 1766A.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-225.
Becker, E. (1962). The birth and death of meaning. New York: Free Press.
Bem, S. L. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology, 42, 155-162.
Dunn, D. R., & Gulbis, J. M. (1976). The Risk Revolution. Parks and Recreation, 11 (8), 12-16.
Ewert, A. W. (1985). Why people climb: The relationship of participant motives and experience level to mountaineering. Journal of Leisure Research, 17(3), 241-250.
Ewert, A. W. (1987). Risk recreation poses new management problems. Park Science, 8(1),7-8.
Ewert, A. W. (1989). Outdoor adventure pursuits: Foundations, models, and theories. Scottsdale, AZ: Publishing Horizons, Inc.
Ewert, A. W., & Hollenhorst, S. (1989). Testing the adventure model: Empirical support for a model of risk recreation participation. Journal of Leisure Research, 21 (2), 124-139.
Iso-Ahola, S.E., LaVerda, D., & Graefe, A.R. (1988). Perceived competence as a mediator of the relationship between high-risk sports participation and self-esteem. Journal of Leisure Research, 21, 32-39
Klausner, S. (Ed.). (1968). Why man takes chances. New York: Anchor Books.
Koocher, G. P. (1971). Swimming, competence, and personality change. Journal of Personality and Social Psychology, 18, 275-278.
Leary, M. R. (1992). Self-presentational processes in exercise and sport. Journal of Sport and Exercise Psychology, 14, 339-351.
Levenson, M. R. (1990). Risk taking and personality. Journal of Personality and Social Psychology, 58 (6), 1073-1080.
Linville, P. W. (1987). Self-complexity as a cognitive buffer against stress related illness and depression. Journal of Personality and Social Psychology, 52, 663-676.
Marsh, H. W., Richards, G. E., & Barnes, J. (1986). Multidimensional self-concepts: The effect of participation in an Outward Bound Program. Journal of Personality and Social Psychology, 50, 195-204.
Marsh, H. W., Richards, G. E., & Barnes, J. (1987). Multidimensional self-concepts: A long-term follow-up of the effect of participation in an Outward Bound Program. Personality and Social Psychology Bulletin, 12, 475-492.
McIntyre, N. (1989). The personal meaning of participation: Enduring involvement. Journal of Leisure Research, 21 (2), 167-179.
Meier, J. F. (1978). Is the risk worth taking? Journal of Physical Education and Recreation, 49 (4), 31-33.
Robinson, D. W. (1985). Stress seeking: Selected behavioral characteristics of elite rock climbers. Journal of Sport Psychology, 7, 400-404.
Robinson, D. W. (1992). A descriptive model of enduring risk recreation involvement. Journal of Leisure Research, 24 (1), 52-63.
Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs: General and Applied, 80 (1, Whole No. 609), 1-28.
Ryckman, R. M., Robbins, M. A., Thornton, B., & Cantrell, P. (1982). Development and validation of a physical self-efficacy scale. Journal of Personality and Social Psychology, 42 (5), 891-900.
Sadalla, E. K., Linder, D. E., & Jenkins, B. A. (1988). Sport preference: A self-presentational analysis. Journal of Sport and Exercise Psychology, 10, 214-222.
Schlenker, B. R. (1980). Impression management: The self-concept, social identity, and interpersonal relations. Monterey, CA: Brooks/Cole.
Schuett, M. A. (1993). Refining measures of adventure recreation involvement. Leisure Sciences, 15, 205-216.
Solomon, S., Greenberg, J., & Pyszczynski, T. (1991). Terror management theory of self-esteem. In C. R. Snyder & D. R. Forsyth (Eds.), Handbook of social and clinical psychology (pp. 21-40). New York: Pergamon.
Straub, W. F. (1982). Sensation seeking among high and low-risk male athletes. Journal of Sport Psychology, 4, 246-253.
Templer, D.I. (1970). The construction and validation of a death anxiety scale. The Journal of General Psychology, 82, 165-177.
United States Bureau of the Census. (1996). Statistical Abstract of the United States (116th ed.). Washington, DC: The National Data Bank.
Wann, D. L., & Hamlet, M. A. (1994). The Joiners' Scale: Validation of a measure of social-complexity. Psychological Reports, 74, 1027-1034.
Westman, A. S., Canter, F. M., & Boitos, T. M. (1984). Denial of fear of dying or of death in young and elderly populations. Psychological Reports, 55, 413-414.
Zuckerman, M. (1979). Sensation seeking: Beyond the optimal level of arousal. Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers.
Zuckerman, M. (1983). Sensation seeking and sports. Personality and Individual Differences, 4, 285-293.
Zuckerman, M. (1984). Sensation seeking: A comparative approach to a human trait. The Behavioral and Brain Sciences, 7, 413-471.