Available online at www.sciencedirect.com

www.elsevier.com/locate/paid

Personality and Individual Differences 45 (2008) 75–81

Self-efficacy, risk taking and performance in rock climbing

David J. Llewellyn a,*, Xavier Sanchez b, Amanda Asghar c, Gareth Jones d

a Department of Public Health and Primary Care, Forvie Site, Robinson Way, University of Cambridge, CB2 2SR, UKb Department of Sport and Exercise Sciences, Parkgate Road, University of Chester, CH1 4BJ, UK

c Faculty of Health, Leeds Metropolitan University, LS1 3HE, UKd Carnegie Sports Injury Clinic, Headingley Campus, LS6 3QS, UK

Received 15 January 2008; received in revised form 4 March 2008; accepted 11 March 2008

Abstract

Self-efficacy may be associated with high risk behaviors in climbers operating at outdoor venues, though little is known about climbersrecruited at indoor venues or less risky forms of rock climbing. Two-hundred and one active rock climbers (163 male) aged 16–62 yearswere recruited at five outdoor and six indoor climbing venues in Britain in a retrospective study. The relationship of self-efficacy to thefrequency and difficulty of high and medium risk rock climbing behaviors was modelled using linear regression. Climbers high in self-efficacy engaged in both high and medium risk forms of rock climbing more frequently (b P 0.18, 95% confidence interval [CI] 0.04–0.32)and at a higher level of difficulty (b P 0.20, 95% CI 0.04–0.36). These associations were attenuated slightly with adjustment for covar-iates, though all remained significant. The same pattern of associations was observed for climbers recruited at indoor and outdoor ven-ues, and for male and female climbers. Rock climbers may therefore participate more frequently, take calculated additional risks andattempt harder climbs when they feel confident in their abilities and are high in self-efficacy. Researchers should not assume psychologicalor behavioral homogeneity within risk taking populations.� 2008 Elsevier Ltd. All rights reserved.

Keywords: Self-efficacy; Social cognitive theory; Risk taking; Risk; Performance; Rock climbing; High risk sport

1. Introduction

The popularity of ‘high risk’ sports such as rock climbinghas increased exponentially in Western societies in recentyears despite the high prevalence of injuries and the poten-tially fatal nature of the activities (Florenthal & Shoham,2001; Jones, Asghar, & Llewellyn, in press; Pain & Pain,2005; Turner, McClure, & Pirozzo, 2004). Researchers havetraditionally focused on the psychological and demographiccharacteristics that discriminate between risk takers andcontrols; however considerable heterogeneity may existwithin risk taking populations (Celentano, Cohn, Davis,& Vlahov, 2002; Fave, Bassi, & Massimini, 2003; Zhao, Sei-bert, & Hills, 2005). Recent findings suggest that self-effi-

0191-8869/$ - see front matter � 2008 Elsevier Ltd. All rights reserved.

doi:10.1016/j.paid.2008.03.001

* Corresponding author. Tel.: +44 (0) 1223 767148; fax: +44 (0) 1223330330.

E-mail address: dl355@medschl.cam.ac.uk (D.J. Llewellyn).

cacy may vary within high risk sports populations(Slanger & Rudestam, 1997), and that active outdoor rockclimbers may take calculated additional risks to challengethemselves when they feel confident in their ability to man-age those risks (Llewellyn & Sanchez, 2008). It is thereforeimportant to examine whether self-efficacy and climbingbehaviors are associated in climbers recruited at indoor ven-ues, and examine the association between self-efficacy andmedium risk forms of rock climbing behavior.

Bandura (1997) hypothesizes that people take risks andchallenge themselves because they believe themselves capa-ble of coping with the situation, and have feelings of self-efficacy. Self-efficacy refers to an individual’s ‘‘belief inone’s capabilities to organize and execute the courses ofaction required to produce given attainments” (Bandura,1997, p. 3). Those with high self-efficacy are more likelyto set themselves challenging goals, expend effort, and per-sist in the face of adversity for longer (Bandura, 1997).

76 D.J. Llewellyn et al. / Personality and Individual Differences 45 (2008) 75–81

Previous congruous studies suggest that mastery andaccomplishment are important factors influencing themotivation to participate in high risk sports (Ewert, 1994;Slanger & Rudestam, 1997), and enactive mastery experi-ences are an important source of self-efficacy in high risksports (Brody, Hatfield, & Spalding, 1988; Norris & Wein-man, 1996). Similarly, high risk sports participants oftenmention the need to be in control, and risk taking itselfappears to represent a challenge to some experiencedclimbers (Fave et al., 2003; Robinson, 1985). Self-efficacybeliefs are formulated by processing information derivedfrom direct domain-specific engagement (enactive masteryexperiences), and to a lesser degree verbal persuasion,vicarious experiences, and physiological states (Bandura,1997). Hence, sports participants high in self-efficacy maybe less likely to fear failure, and more likely to set them-selves difficult goals and take calculated, as opposed toreckless risks (Kontos, 2004).

Slanger and Rudestam (1997) examined the relation-ships between self-efficacy, sensation seeking and risk tak-ing among male high risk sports participants includingskiers, rock climbers, kayakers and stunt flyers. Partici-pants were categorized into ‘extreme’ and ‘high’ risk takinggroups depending upon their sport-specific behaviors. Forexample, climbers who ascended routes without protectiveropes (soloing) were extreme risk takers, whereas thosewho only climbed using ropes were high risk takers. Self-efficacy was measured using a Physical Self-Efficacy Scale(PSES) developed by them to assess self-efficacy in highrisk sports, and sensation seeking was assessed using theSensation Seeking Scale V (Zuckerman, 1994). Theirresults indicated that extreme risk takers had higher levelsof self-efficacy than high risk takers, though no associationwith sensation seeking was observed. They concluded thathigh risk sports participants with elevated levels of self-effi-cacy were more likely to take greater risks due to greatermastery needs and the ability to manage their emotionalstates (Slanger & Rudestam, 1997). Participants alsonamed confidence specifically as the most important disin-hibiting factor in qualitative items. However, the numberof risk takers in this study was small (N = 40), risk takingwas operationalized in a simplistic way, and the exclusivelymale sample did not allow the study of female high risksports participants.

Recently Llewellyn and Sanchez (2008) investigated risktaking and individual differences in 116 active outdoor rockclimbers (24% women). They found that risk taking in rockclimbing was relatively common as only 22% of climbersavoided risky forms of climbing (i.e. soloing and tradi-tional leading). Self-efficacy was measured using the PSES,and impulsivity and sensation seeking were assessed usingthe Impulsive–Sensation Seeking Scale (Zuckerman,1994). Those high in self-efficacy and males were mostlikely to take risks in climbing. Small associations with sen-sation seeking, impulsivity and age were also observedthough these were not in the hypothesized direction or sig-nificant after adjustment for potential confounders. They

concluded that to challenge themselves, active outdoorrock climbers may take calculated additional risks whenthey feel confident in their ability to manage those risks(Llewellyn & Sanchez, 2008).

Taken together previous findings suggest that self-effi-cacy may be the key to understanding risk taking in highrisk sports. However, it is not known whether these resultsgeneralize to active indoor climbers, or whether self-effi-cacy is also associated with medium risk forms of rockclimbing. The purpose of the present study was thereforeto investigate the relationship between self-efficacy and awide range of rock climbing behaviors in active rock climb-ers recruited at both indoor and outdoor venues.

2. Methods

2.1. Participants

Of the 205 active British rock climbers approached fol-lowing institutional ethics approval, 201 (98%) gave writteninformed consent, and completed a self administered anon-ymous questionnaire. Climbers were recruited from fiveoutdoor climbing venues (N = 100), and six indoor climb-ing facilities (N = 101) in Britain. The 163 male climbers(M = 35.2, SD = 11.8 years of age) and 38 female climbers(M = 35.1, SD = 10.7 years of age) did not differ signifi-cantly in age. There was no significant difference in yearsclimbing experience between men (M = 13.9, SD = 11.8)and women (M = 11.6, SD = 9.2).

2.2. Rock climbing behaviors

Detailed information regarding the frequency of partic-ipation in key forms of rock climbing (i.e., soloing, tradi-tionally protected leading, indoor sport leading, outdoorsport leading, and bouldering) in the last 12 months wasobtained. Soloing is where a climber ascends without arope, and is completely unprotected in the event of a fall.Traditionally protected leading is where climbers placepieces of climbing protection as they progress up a rockface outside. In this case the outcome of a fall is determinedby the nature of the climb and the skill of the climber,though may be serious. Sport or bolt-protected leading(both outdoors and indoors) is similar to traditionally pro-tected leading, though expansion bolts are prearranged asprotection, and the consequences of a fall are typically lessserious. Bouldering is where a climber tackles particularlydifficult problems on very short climbing routes close tothe ground. The difficulty (climbing grade) at which theycould consistently perform each type of climbing techniqueover the last 12 months was also recorded, giving domainspecific measures of technical difficulty accomplished.Climbers rated difficulty using the British climbing gradingsystem, which we assigned values ranging from 1 for theeasiest (Moderate) to 16 for the hardest (Extremely Severe9) for all forms of leading and soloing, and 1 for the easiest(4a) to 11 for the hardest (7a) for bouldering (Fyffe & Peter,

D.J. Llewellyn et al. / Personality and Individual Differences 45 (2008) 75–81 77

1997). The frequency and difficulty of soloing and tradi-tional leading have previously been identified as usefuloperational measures of risk taking in rock climbing (Lle-wellyn & Sanchez, 2008). We therefore subdivided theclimbing behaviors into ‘high risk’ (soloing and traditionalleading) and ‘medium risk’ (indoor sport leading, outdoorsport leading and bouldering) categories based upon thepotential for falls to result in serious or fatal injury. Yearsof participation in rock climbing was also incorporated asan overall measure of rock climbing experience.

2.3. The Climbing Self-Efficacy Scale (CSES)

A validated measure of self-efficacy specific to thedomain of rock climbing was not available in the literature,though it is recommended that situationally specific mea-sures are used (Bandura, 1986, 1997, 2001). Thus, a climb-ing specific Climbing Self-Efficacy Scale was developed inthe present study (CSES; Fig. 1). First, the major variablescomprising the perception of efficacy in the domain underconsideration – rock climbing – were identified inductivelyby semi-structured interviews with active recreational rockclimbers (N = 5), climbing instructors (N = 2) and aca-demic specialists in the areas of social cognitive theory(N = 2), risk taking (N = 1), and sport psychology(N = 2). The ten variables identified were congruent withself-efficacy theory (Bandura, 1997). Second, we thenpiloted a draft version of the CSES with 30 active climbers(23 male) recruited from three sites who did not take partin the main study. The internal consistency of the draft

Fig. 1. The Climbing Self-

CSES in the pilot study was high (a = 0.92). Lastly, follow-ing feedback from these climbers, minor alterations weremade to the overall format and the wording of the instruc-tions. In the final version of the CSES climbers were askedto rate how confident they feel about their climbing abili-ties at the moment on a scale ranging from ‘not at all con-fident’ (scored 0%), through ‘moderately confident’ (scored50%), to ‘extremely confident’ (scored 100%). The totalscore (ranging from 0 to 1000) gives a measure of an indi-vidual’s expectations or beliefs about their ability to per-form actions necessary to produce a given effect inclimbing. Item analysis revealed that all items were highlycorrelated with the corrected CSES total (r P 0.49).Exploratory principal factor analysis suggested that oneimportant underlying factor, as identified by scree testand eigenvalues (Factor 1 = 4.5; Factor 2 = 0.7) accountedfor 50.3% of the variance. The internal consistency of theCSES in the main study was high (a = 0.88).

2.4. Statistical analyses

Data analysis was carried out using the STATA SE pro-gram version 9.2. (StataCorp, 2005). Descriptive statistics,Pearson product-moment correlation coefficients and least-squares linear regression models were calculated. Age, sex,type of climbing venue at which they were recruited andyears climbing experience were included as covariates inadjusted regression models. We also conducted regressionanalyses separately by location recruited and sex assensitivity analyses, and tested for potential two-way

Efficacy Scale (CSES).

78 D.J. Llewellyn et al. / Personality and Individual Differences 45 (2008) 75–81

interactions between self-efficacy and sex and locationrecruited. Square root transformations were performedafter preliminary analyses to reduce mild positive skewobserved in five variables (years climbing experience, solo-ing frequency, soloing grade, traditional lead frequency,and outdoor sport lead frequency). Two-tailed tests wereused throughout, and statistical significance was preset atp < 0.05.

3. Results

The characteristics of the climbers included in our anal-yses are described in Table 1. Climbers recruited at indoorand outdoor climbing venues did not differ significantly inage, sex, self-efficacy, climbing experience or participationin high risk forms of climbing. The majority of climbersrecruited at both indoor and outdoor venues were male(just over 80%), with the mean age being in the mid-thirties.Mean self-efficacy levels were comparable between venues,and corresponded to values falling between ‘moderate’ and‘extreme’ confidence. Levels of experience varied widely(ranging from 1 to 48 years), though again were similarbetween venues. The majority of climbers had taken partin high risk forms of climbing during the previous 12months (soloing or traditional leading).

Correlations among the measures are shown in Table 2.Age was strongly associated with experience (r = 0.70,p < 0.001), and weakly associated with self-efficacy(r = 0.21, p < 0.01). The relationship between age and thefrequency of climbing behaviors was complex, with twopositive significant associations observed, two non-signifi-cant associations, and one negative significant association.Age did not appear to be associated with the difficulty atwhich the climbers operated, with the exception of a smallpositive association with bouldering grade (r = 0.29,p < 0.001). Self-efficacy was positively and significantly

Table 1Baseline characteristics of rock climbers recruited at indoor and outdoorclimbing venues

Variable Total sample(N = 201)

Recruitedindoors(N = 101)

Recruitedoutdoors(N = 100)

p*

Mean age (SD) y 35.2 (11.6) 34.7 (11.7) 35.8 (11.5) 0.50Women % 18.9 18.8 19.0 0.97Mean self-

efficacy (SD)719.5 (118.9) 728.1 (121.0) 711.0 (116.6) 0.30

Mean climbingexperience(SD) y

13.5 (11.3) 13.4 (11.1) 13.6 (11.6) 0.83

Engaged in highrisk climbing,a

%

89.1 86.1 92.0 0.18

Note. SD = Standard deviation; y = year.* p-Value for difference between climbers recruited at indoor and outdoorvenues (two-tailed).

a Any degree of participation in soloing or traditional leading during theprevious 12 months.

associated with experience (r = 0.29, p < 0.001), the fre-quency of all climbing behaviors (r P 0.18, p < 0.01), andthe difficulty of all behaviors (r P 0.19, p < 0.05). The fre-quency of all forms of climbing behavior were positivelyand significantly associated with each other (r P 0.17,p < 0.05), with the exception of soloing frequency andindoor lead frequency (r = 0.07, p = 0.33), and traditionallead frequency and bouldering frequency (r = 0.04,p = 0.60). Levels of climbing difficulty were all significantlyand strongly associated (r P 0.62, p < 0.001).

Linear regression analyses were performed to investigatewhether self-efficacy predicted the frequency of participationand level of difficulty in rock climbing behaviors (Table 3). Inthe first unadjusted models self-efficacy was significantlyassociated with all forms of climbing behavior (b P 0.18,p < 0.05). In the second models, age, sex, type of climbingvenue and years climbing experience were adjusted for.While the majority of the associations between self-efficacyand the climbing behaviors were attenuated after adjust-ment, self-efficacy remained a significant predictor of allforms of climbing behavior (b P 0.15, p < 0.05).

We also conducted a series of sensitivity analyses toexamine the robustness of the regression models (dataavailable upon request). The same pattern of associationswas observed when participants recruited at indoor andoutdoor venues were analyzed separately, and there wasno evidence for a two-way interaction between self-efficacyand location. Similarly the same pattern of associationswas observed when male and female climbers were ana-lyzed separately, with no evidence for a two-way interac-tion between self-efficacy and sex.

4. Discussion

The purpose of the present study was to investigate theassociation between self-efficacy and a wide range of rockclimbing behaviors in active rock climbers recruited at bothindoor and outdoor venues. Our data suggest that climbershigh in self-efficacy engage in both high and medium riskforms of rock climbing more frequently and at a higherlevel of difficulty. These associations were generally attenu-ated by adjustment for age, sex and years climbing experi-ence, though remained significant for all ten associations.The same pattern of associations was observed for thoserecruited at indoor and outdoor venues, and for maleand female climbers.

To our knowledge the present study is the first to exam-ine the association between self-efficacy and rock climbingbehaviors whilst incorporating climbers recruited at indoorclimbing venues. Our study was the first to incorporate ameasure of self-efficacy (the CSES) specific to the domainof interest – rock climbing – rather than a measure ofself-efficacy in high risk sports in general (the PSES). Thismay also be the first study to examine the associationbetween self-efficacy and moderate risk forms of rockclimbing behavior (indoor and outdoor sport leading andbouldering). We incorporated the largest sample of active

Table 2Correlations between age, self-efficacy, experience and the frequency and difficulty of rock climbing behaviors (N = 201)

Age Self-efficacy

Experience Soloingfrequency

Soloingdifficulty

Trad. leadfrequency

Trad. leaddifficulty

Sport leadfrequency

Sport leaddifficulty

Indoor leadfrequency

Indoor leaddifficulty

Boulderingfrequency

Self-efficacy 0.21**

Experience 0.70*** 0.29***

Soloingfrequency

0.02 0.21** 0.22**

Soloingdifficulty

�0.06 0.30*** 0.23** 0.31***

Trad. leadfrequency

0.26*** 0.36*** 0.33*** 0.51*** 0.18*

Trad. leaddifficulty

�0.06 0.39*** 0.25** 0.38*** 0.81*** 0.32***

Sport leadfrequency

0.16* 0.25*** 0.31*** 0.20** 0.30*** 0.38*** 0.47***

Sport leaddifficulty

�0.17 0.37*** 0.18 0.27** 0.77*** 0.22* 0.86*** 0.51***

Indoor leadfrequency

�0.00 0.22** 0.14* 0.07 0.35*** 0.28*** 0.46*** 0.50*** 0.43***

Indoor leaddifficulty

�0.14 0.29*** 0.17* 0.29*** 0.73*** 0.23** 0.82*** 0.48*** 0.92*** 0.54***

Boulderingfrequency

�0.23** 0.18** 0.04 0.17* 0.50*** 0.04 0.56*** 0.36*** 0.53*** 0.44*** 0.50***

Boulderingdifficulty

0.29*** 0.19* 0.08 0.28*** 0.62*** 0.09 0.71*** 0.28*** 0.78*** 0.33*** 0.74*** 0.58***

Note. Trad. = traditional.* p < .05 (two-tailed).** p < .01 (two-tailed).*** p < .001 (two-tailed).

D.J

.L

lewelly

net

al./P

erson

ality

an

dIn

divid

ua

lD

ifferen

ces4

5(

20

08

)7

5–

81

79

Table 3Linear regression models illustrating standardized beta coefficients (95% confidence intervals) for the association between self-efficacy and rock climbingbehaviors (N = 201)

Unadjusted models Adjusted modelsa

b (95% CI) p* b (95% CI) p*

High risk behaviors

Soloing frequency 0.21 (0.07–0.34) 0.003 0.15 (0.00–0.29) 0.044Soloing grade 0.30 (0.13–0.46) <0.001 0.24 (0.05–0.43) 0.015Traditional lead frequency 0.35 (0.22–0.48) <0.001 0.28 (0.14–0.42) <0.001Traditional lead grade 0.38 (0.25–0.52) <0.001 0.38 (0.22–0.54) <0.001

Medium risk behaviors

Sport lead frequency 0.25 (0.11–0.39) <0.001 0.18 (0.03–0.32) 0.016Sport lead grade 0.36 (0.19–0.53) <0.001 0.33 (0.13–0.53) 0.001Indoor lead frequency 0.22 (0.08–0.36) 0.002 0.17 (0.03–0.31) 0.016Indoor lead grade 0.29 (0.15–0.43) <0.001 0.25 (0.10–0.40) 0.001Bouldering frequency 0.18 (0.04–0.32) 0.010 0.17 (0.02–0.31) 0.026Bouldering grade 0.20 (0.04–0.36) 0.013 0.22 (0.04–0.40) 0.019

a Adjusted for age, sex, type of venue where recruited, and years climbing experience.* p-Value for regression coefficient (two-tailed).

80 D.J. Llewellyn et al. / Personality and Individual Differences 45 (2008) 75–81

rock climbers in a study of this kind to date (N = 201).With a response rate of 98%, the present study also com-pares favorably with that obtained in some previous stud-ies. For example, Logan and colleagues obtained aresponse rate of 51% in a postal survey of climbers, whichthey acknowledge as a limitation (Logan, Makwana,Mason, & Dias, 2004). The response rate that we obtainedis consistent with studies in which climbers were alsodirectly approached (Llewellyn & Sanchez, 2008; Wright,Royle, & Marshall, 2001).

Some methodological limitations should also be bornein mind when assessing these findings. While the presentstudy provides important new information about the asso-ciation between rock climbing behaviors and self-efficacy,the mechanisms underlying these relationships remainunclear (Mann, 2003). Our sample excluded those climberswho are no longer active, which may have biased our studytowards those with higher levels of self-efficacy. Furtherresearch is therefore necessary to examine these associa-tions longitudinally.

Our findings are consistent with previous studies report-ing an association between self-efficacy and risk taking inhigh risk sports (Slanger and Rudestam, 1997), and rockclimbing in particular (Llewellyn and Sanchez, 2008). Adomain specific measure of self-efficacy in rock climbingwas incorporated in the present study (the CSES; Fig. 1),whereas previous studies have incorporated the PSES, amore general measure of self-efficacy in high risk sports(Slanger and Rudestam, 1997). The proportion of femaleclimbers in the present study (19%) is similar to thatobserved by Llewellyn and Sanchez (24%), whereas agreater proportion of climbers were engaged in high riskforms of climbing (89% vs. 78%). The present study there-fore extends our understanding by demonstrating that theassociation between self-efficacy and rock climbing behav-iors is not constrained to high risk behaviors (i.e. soloingand traditional leading), nor to climbers recruited at out-door venues.

It seems that some climbers take additional risks andselect harder climbs to ascend if they believe themselvescapable of managing the risks effectively and meeting thechallenges involved. Although self-efficacy significantlypredicted all forms of rock climbing behavior, most ofthese associations were attenuated after adjustment foradditional variables such as sex and experience. This is con-gruous with self-efficacy theory, which posits that enactivemastery experiences are the primary source of self-efficacy(Bandura, 1997). This is also consistent with previous workstressing the role of mastery and accomplishment as keymotivating factors (Ewert, 1994; Slanger and Rudestam,1997), and the importance of enactive mastery experiencesas a source of self-efficacy in high risk sports (Brody et al.,1988; Norris and Weinman, 1996).

In conclusion, self-efficacy was associated with the fre-quency and difficulty of a wide range of medium and highrisk climbing behaviors in climbers recruited at both indoorand outdoor venues. Climbers may participate more fre-quently, take calculated additional risks, and attemptharder climbs when they feel confident in their abilitiesand are high in self-efficacy. Our findings are importantas they extend our understanding of the behavioral andpsychological heterogeneity in high risk pursuits, and rein-force the importance of self-efficacy when ‘‘pushing the lim-its” and taking calculated risks.

Acknowledgements

The first author was supported by The British Academy(grant 39588). The comments of Peter Clarke, Peter Mar-tin, and Gilbert MacKay are gratefully acknowledged.

References

Bandura, A. (1986). Social foundations of thought and action: A social

cognitive theory. Englewood Cliffs, NJ: Prentice-Hall.Bandura, A. (1997). Self-efficacy: The exercise of control. New York:

Freeman.

D.J. Llewellyn et al. / Personality and Individual Differences 45 (2008) 75–81 81

Bandura, A. (2001). Social cognitive theory: An agentic perspective.Annual Review of Psychology, 52, 1–26.

Brody, E. B., Hatfield, B. D., & Spalding, T. W. (1988). Generalization ofself-efficacy to a continuum of stressors upon mastery of a high-risksport skill. Journal of Sport and Exercise Psychology, 10(1), 32–44.

Celentano, D. D., Cohn, S., Davis, R. O., & Vlahov, D. (2002). Self-efficacy estimates for drug use practices predict risk reduction amonginjection drug users. Journal of Urban Health, 79(2), 245–256.

Ewert, A. W. (1994). Playing the edge: Motivation and risk taking in ahigh-altitude wilderness like environment. Environment and Behavior,

26(1), 3–24.Fave, A. D., Bassi, M., & Massimini, F. (2003). Quality of experience and

risk perception in high-altitude rock climbing. Journal of Applied Sport

Psychology, 15(1), 82–98.Florenthal, B., & Shoham, A. (2001). The impact of persuasive informa-

tion on changes in attitude and behavioral intentions toward riskysports for arousal-seeking versus arousal-avoidance individuals. Sport

Marketing Quarterly, 10, 83–95.Fyffe, A., & Peter, I. (1997). The handbook of climbing. London: Pelham

Books.Jones, G., Asghar, A., & Llewellyn, D. J. (in press). The epidemiology of

rock climbing injuries. British Journal of Sports Medicine. doi:10.1136/bjsm.2007.037978.

Kontos, A. P. (2004). Perceived risk, risk taking, estimation of ability andinjury among adolescent sport participants. Journal of Pediatric

Psychology, 29(6), 447–455.Llewellyn, D. J., & Sanchez, X. (2008). Individual differences and risk

taking in rock climbing. Psychology of Sport and Exercise, 9, 413–426.

Logan, A. J., Makwana, N., Mason, G., & Dias, J. (2004). Acute handand wrist injuries in experienced rock climbers. British Journal of

Sports Medicine, 38(5), 545–548.Mann, C. J. (2003). Observational research methods. Research design II:

Cohort, cross sectional, and case-control studies. Emergency Medical

Journal, 20(1), 54–60.Norris, R. M., & Weinman, J. A. (1996). Psychological change following a

long sail training voyage. Personality and Individual Differences, 21(2),189–194.

Pain, M. T., & Pain, M. A. (2005). Risk taking in sport. Lancet,

366(Suppl. 1), S33–S34.Robinson, D. W. (1985). Stress seeking: Selected behavioral characteristics

of elite rock climbers. Journal of Sport Psychology, 7, 400–404.Slanger, E., & Rudestam, K. E. (1997). Motivation and disinhibition in

high risk sports: Sensation seeking and self-efficacy. Journal of

Research in Personality, 31(3), 355–374.StataCorp. (2005). Stata Statistical Software: Release 9.2. College Station:

StataCorp LP.Turner, C., McClure, R., & Pirozzo, S. (2004). Injury and risk-taking

behavior – A systematic review. Accident Analysis and Prevention,

36(1), 93–101.Wright, D. M., Royle, T. J., & Marshall, T. (2001). Indoor rock climbing:

Who gets injured? British Journal of Sports Medicine, 35(3), 181–185.Zhao, H., Seibert, S. E., & Hills, G. E. (2005). The mediating role of self-

efficacy in the development of entrepreneurial intentions. Journal of

Applied Psychology, 90(6), 1265–1272.Zuckerman, M. (1994). Behavioral expressions and biosocial bases of

sensation seeking. Cambridge: Cambridge University Press.