This study examined the relationships among three components of perceived risk of injury: (a) probability of injury, (b) worry/concern of being injured, and (c) confidence in avoiding injury. Participants were 434 athletes from 3 contact sports (hockey, soccer, and football). Correlations between the components showed a positive relationship between worry/concern and probability of injury, and negative relationships between worry/concern and confidence in avoiding injury as well as probability of injury and confidence in avoiding injury. Results also indicated that those athletes who had been previously injured perceived the highest probability of re-injury, demonstrated the greatest worry/concern of re-injury, and had the least amount of confidence in their ability to avoid re-injury. Additional gender differences and type of sport differences were found for certain subscales. When football was excluded from the analyses, a 2 (sport) by 2 (gender) by 2 (injury status) ANOVA demonstrated sport by gender interactions, a gender by injury interaction, and a sport by gender by injury interaction for certain subscales. Findings are discussed in light of self-efficacy and risk taking theory, and the potential effects of socialization on gender differences in perceived risk of injury.

Introduction

       There is little doubt that injuries have a significant impact on individual and team performance (Doyle, Gleeson, & Rees, 1998; Dunn, 1999; Evans, Hardy, & Fleming, 2000; Kolt, 2000; Synder, 1990; Wischnia, 1980). Research shows that the relationship between injury and performance tends to be more negative than positive; meaning that if an athlete is injured, their performance will suffer. From a physical standpoint, the primary reason for the negative relationship is that athletes tend to not attain their pre-injury status upon return from an injury. However, the decrease in performance associated with injury has been attributed to psychological factors more so than physical factors (Doyle et al., 1998; Dunn, 1999; Evans et al., 2000; Synder, 1990; Wischnia, 1980). Heil (1993) stated that injury exerts a direct impact on an athlete’s psychological well-being, which in turn, directly influences health, performance, and the risk of further physical injury. Even injuries that are small or less severe that allow an athlete to still compete may limit preparation and subsequent performance in competition (Kolt, 2000).

       Although there have been a number of psychological factors associated with injuries in sport, Williams and Andersen (1998) suggested that researchers expand the number of psychological variables currently being studied. One potentially debilitative psychological factor affecting performance concerns an athlete’s fear or perceived risk of injury. Fear has been defined as an “apprehensive feeling towards anything regarded as a source of danger, or towards a person regarded as able to inflict injury or punishment” (Simpson & Weiner, 1989). We have defined the fear of injury in sport as the unpleasant feeling of apprehension or distress caused by the anticipation of physical damage to the body or a part of the body.

       Intuitively, most of us would agree that fear of injury has a negative effect on performance. However, researchers have difficulties documenting this relationship. The primary reason for this has to do with the word “fear” itself. Many negative connotations have been associated with the word “fear.” The names “suck” and “sissy” have been used to refer to people who are afraid or who show anxiety (Kidd, 1990). Snyder (1990) conducted a study where gymnasts were labeled according to their risk or non-risk taking behaviors. Gymnasts who were viewed as risk takers were labeled by fellow gymnasts as “kamikazes, gutsy, and animals” while those who were non-risk takers were labeled “wimps, chickens, or balkers.” Research also shows that athletes may attempt to hide their fear because they feel coaches view fear as a weakness or they are afraid of looking foolish. Many athletes feel that they have to be tough and “macho” in order to participate in sports, and thus self-reports of fear of injury may not be as reflective of actual perceptions as they are of socially accepted (e.g., gender stereotyped behaviors) norms. For these reasons, we advocate using the term perceived risk of injury, as suggested by Kontos (2000).

       Our review of the recent literature on perceived risk of injury shows that this construct has been conceptualized in three ways. First, it has been linked to self-efficacy theory with the emphasis being on one’s confidence in his/her ability to avoid injury. Second, perceived risk of injury has been likened to feelings of competitive anxiety, relating to one’s worry or concern about being injured in sport. Finally, it has been looked at as a perceived probability of being injured. Each of these areas is now briefly reviewed.

Perceived Risk of Injury as a Lack of Confidence in Avoiding Injury

       Self-efficacy theory has received considerable attention with respect to its relationship with fear of injury. Magyar and Chase (1996) stated that fear of injury exists when an athlete lacks confidence in the ability to perform successfully in a threatening or taxing situation. Injured athletes have expressed low self-efficacy in their perceived ability to perform skills and movements required by their specific sport without re-injuring themselves (Doyle et al., 1998, Evans et al., 2000). For example, a rugby player who had sustained a shoulder injury had low self-efficacy upon his return to competition to take on tackles with the previously injured shoulder even though he had fully recovered (Evans et al., 2000). This lowered confidence in his ability to adequately use his shoulder caused him to use his other shoulder more often, not only decreasing his performance, but also placing him in vulnerable situations to experience another injury.

       Bandura (1997) hypothesized that an athlete with low self-efficacy would also be fearful of the sport environment and consequently lack confidence in attaining a positive outcome (e.g., avoiding injury). Hence, athletes lacking confidence in avoiding injury would be more likely to be injured than those confident in avoiding injury. Another consequence of having low self-efficacy is that a high level performer may decide to participate in activities at a lower performance level to accommodate the injury (Doyle et al., 1998). The athlete may choose to participate only in low risk sporting events in order to avoid excessive stress on the injured body part. A worse case scenario would be an athlete whose self-efficacy is so low that they have an unwillingness to perform because of an injury or the thought of being injured (Synder, 1990).

       Clearly, the return to play, after suffering an injury, is a major decision that should be made by a team consisting of physicians, coaches, athletic trainers, the athlete, and the sport psychologist. An athlete who is not fully efficacious upon returning to competition carries a greater risk of re-injury. It has been suggested that when athletes are given the green light to return to competition many of them are not psychologically ready to compete (Evans et al., 2000). There is ample evidence that shows unnecessary injuries occur and that recovery time is increased because of the failure to adequately address psychological factors (Cupal, 1998). If confidence in the athlete is not expressed by significant others (i.e., coach, teammates) prior to returning to play, the athlete’s self-confidence may be further threatened leading to an increased fear of failure or re-injury (Heil, 1993).

Perceived Risk of Injury as Worry or Concern

       Perceived risk of injury has also been included as a component of competitive worry. Most recently, Dunn (1999) included a subscale relating to an athlete’s worries about potential injury of physical danger in his competitive worry measure. Similarly, in their measure of anxiety, Endler, Edwards and Vitelli (1991) included a “physical danger A-trait” dimension in their anxiety measure. Physical danger A-trait was defined as a person’s predisposition to experience anxiety in situations in which the potential for physical harm exists. Hackfort (1986) also included “physical harm/injury anxiety” as a dimension of competitive trait anxiety. Research has shown that the perceived risk of injury is a common source of worry and a possible reason for leaving competition among young gymnasts (Duda, 1995; Duda & Gano-Overway, 1986; Klint & Weiss, 1986; Weiss, Weise, & Klint, 1989).

Perceived Risk of Injury as a Probability

       Kontos, Feltz, and Malina (2000) recognized the need for more research on fear of injury in sport. They developed a general scale for the measurement of perception of risk of injury as a probability or likelihood for use with contact team sports. The Risk of Injury in Sports Scale (RISSc) consists of 24 items representing 6 injury subscales (i.e., uncontrollable, controllable, overuse, upper body, surface-related, and re-injury).

       Kontos et al. (2000) examined the relationship between the perceived probability of injury (using the RISSc) and gender. Females scored higher on all subscales except for the uncontrollable factor, where males were higher. The “uncontrollable” dimension represents injuries that are more common and involve factors (e.g., collisions) inherent in the team or contact/collision sports. Although they found gender differences, their interpeation of these findings was cautious, as sport type may have been a confounding factor. Kontos (in press) examined the perceived probability of injury (using the RISSc) and gender in soccer where both males and females participate under similar rules, and again found strong support for gender differences in perceived probability of injury. However, low perceptions of risk were predictive of injury regardless of gender, indicating that the gender differences in perceived risk may not translate into a difference in injury risk.

       It was also suggested that that the presence of previous injuries might increase subsequent perceptions of risk of injury (Kontos, in press; Kontos et al., 2000). Results showed that males reported more previous injuries than females and that there were higher RISSc scores among athletes who had been injured in the past year compared to those who had not been injured. The type, location, severity, immediacy, and repetitiveness of an injury could influence perceived risk of injury. For instance, an athlete who has incurred multiple, severe ankle sprains may perceive more risk in playing sports than an athlete who has had only one such injury.

       Our review shows that perceived risk of injury has been conceptualized in three ways: (a) as a lack of self-efficacy, (b) as a worry or concern component of competitive anxiety, and (c) as a probability of being injured. No single study has explored the inter-relationships among these three constructs. More research is needed to further define or pinpoint the specific psychological factors associated with perceived risk of injury so practical applications can be developed. This information can then be used by coaches and sport psychologists to design and implement psychological skill training programs to cater to the specific needs of their athletes (Dunn, 1999).

       The purpose of this study, therefore, was to examine the interrelationships among the three components of perceived risk of injury in male and female collegiate, team contact sport athletes. Additionally, this study was designed to assess the three components of perceived risk of injury in relation to gender, sport type, and previous injuries. Based on previous research the following hypotheses were made:

1. A positive relationship between worry/concern and probability of injury;
2. A negative relationship between confidence in avoiding injury and worry/concern of injury and between confidence in avoiding injury and probability of injury;
3. No differences in worry/concern, probability, and confidence in avoiding injury in sport type (i.e., soccer, hockey and football). All sports are similar being contact in nature;
4. Gender differences on all measures (worry/concern, probability and confidence) where females would perceive more worry/concern, more risk and have less confidence in their abilities to avoid injuries compared to males;
5. Differences in worry/concern, probability, and confidence in avoiding injuries related to previous injury status – i.e., if an athlete has been injured in the past, he/she would perceive more worry/concern, greater probability of injury, and less confidence compared to those athletes who weren’t previously injured. In addition, there would be a positive relationship between the number of times an athlete has been injured and worry/concern, and risk of injury, and a negative relationship between the same variable and confidence in avoiding injury; and
6. A positive relationship between the severity of an athlete’s previous injury to their perceived probability of injury and worry/concern about injury, and a negative relationship between injury severity and confidence in avoiding injury.

Participants

       Participants were 434 male and female athletes from 3 contact sports (ice hockey, soccer and football). For ice hockey, there were 76 females and 86 males. For soccer, there were 32 males and 32 females. All football players were male (n = 208). The mean age of the participants was 20.91 (SD = 3.97). The average length of time participants were involved in their sport ranged from less than one year to 25 years (M = 10.79, SD = 4.80). The majority of the sample (58%) indicated that they were a “starter” (i.e., that they played more than 50% of each game). Finally, most of the sample played at the Division 2 level (48.4%) followed by Division 1 (13.4%), Recreational (10.8%), Division 3 (8.8%), Club (8.3%), NAIA (5.5%), and Junior A (3.9%).

Measures

       Demographics. A demographics questionnaire was used to obtain relevant personal information (e.g., age, sport type, starting status, competitive level, etc.), and previous injury information from each participant. An injury was defined as “any injury you had in your sport in the past 12 months that required treatment beyond icing and taping” (see Kontos et al., 2000). If the participant had been injured, (s)he was asked to indicate how many times (s)he had been injured and to describe their most serious and most recent injury. Both the most serious and most recent injuries were rated as mild (“an injury requiring treatment without interruption of training/practicing or playing in games”), moderate (“an injury more severe than a mild injury, that interferes with training/practicing or playing in games”), or major (“an injury more severe than mild or moderate that requires a long duration of downtime, often with surgery or hospitalization”) (Heil, 1993, p. 114).

       Risk of Injury in Sport Scale (RISSc; Kontos et al., 2000). The original RISSc contained 24-items designed to measure athletes’ perceptions regarding the probability of being injured in team, contact sports. The RISSc consists of 6 subscales dealing with uncontrollable injuries, controllable injuries, overuse injuries, upper-body injuries, surface-related injuries and re-injuries. Previous research has shown adequate validity and reliability for the RISSc (Kontos, in press; Kontos et al., 2000). For the present study, the original RISSc was extended to include two additional hypothesized components of perceived risk of injury, namely, fear/worry and confidence in avoiding injury. The 24 items of the original RISSc remained the same (to maintain the integrity of the original 6 factors), however, the stems were changed to reflect the two new components (see descriptions below).

       RISSc-Probability of Injury (RISSc-P). The RISSc-P consisted of the original 24 items of the RISSc. Each item uses the stem, “What do you think are the chances that you will…” Sample items from each of the factors are as follows: from the uncontrollable subscale: “injure yourself in a collision with an opponent;” from the controllable subscale: “be injured by not paying attention to what you are doing;” from the overuse subscale: “be injured from practicing too hard;” from the upper body subscale: “injure your arm or wrist;” from the surface-related subscale: “fall down and injure yourself;” and from the re-injury subscale: “re-injure an area you have recently injured.” Participants responded using a Likert-type scale ranging from 1 (very unlikely) to 6 (very likely). Reliabilities for the subscales of the RISSc-P in this study were strong, ranging from .77 to .89.

       RISSc-Worry/Concern about Injury (RISSc-W). The RISSc-W was used to assess how worried or concerned an athlete is about being injured in sport. The format was identical to the RISSc-P, except the stem was, “How worried or concerned are you that you will…” Respondents were instructed to respond using a Likert-type scale ranging from 1 (not worried or concerned at all) to 10 (very worried or concerned). The reliabilities for each of the 6 subscales ranged between .82 and .94.

       RISSc-Confidence in Avoiding Injury (RISSc-C). The RISSc-C was used to assess how confident an athlete is in avoiding injury. The RISSc-C used the stem, “How confident are you in your ability to avoid…” To be more similar to other efficacy measures (cf. Moritz, Feltz, Fahrbach, & Mack, 2000), respondents were instructed to choose their answers using a scale that ranged from 0 (not confident at all) to 10 (very confident). Reliabilities for the subscales of the RISSc-C ranged between .84 and .96.

Procedure

       Approval to conduct the study was granted by the Institutional Review Board. Data collection occurred during regular team meetings or practices. Participants took approximately 15-30 minutes to complete the measures.

Results

       The results of this study are presented in 4 sections. The first section contains the descriptive statistics. The second section considers the relationship between the 3 primary dependent variables (probability, worry/concern, and confidence in avoiding injury). The third section includes analyses relating each outcome variable to athletes’ previous injury status. Finally, analyses for gender by sport and injury status are included in the fourth section. All statistical analyses were performed using SPSS (Version 10.0.05).

Descriptive Statistics

       Descriptive statistics for the three components of the RISSc according to sport and gender can be found in Table 1. For all subscales of the RISSc-P as well as the total RISSc-P scores, mean values ranged between 2.12 and 2.83. These values indicate that participants perceived low probabilities of being injured in sport. An Analysis of Variance (ANOVA) showed that there were significant gender differences for the perceived probability of re-injury (F (1, 432) = 7.56, p < .05), uncontrollable injuries (F (1, 432) = 27.54, p < .01), and total RISSc-P scores (F (1, 432) = 5.15, p < .05). In all cases, values were higher for females compared to males. With respect to type of sport, an ANOVA showed that there were significant differences between sport type on the probability of re-injury (F (2, 431) = 3.34, p < .05), surface injuries (F (2, 431) = 4.41, p < .05), overuse injuries (F (2, 431) = 4.48, p < .05), uncontrollable injuries (F (2, 431) = 11.20, p < .01), and on the total RISSc-P scores (F (2, 431) = 3.62, p < .01). Follow-up post hocs using Tukey’s test showed that the soccer players perceived a higher probability of re-injury, surface and overuse injuries compared to football and hockey players. With respect to the uncontrollable injuries, football players perceived the lowest probability compared to the hockey players and soccer players. Finally, for the total perception of probability of injury score, soccer players had the highest means and they differed significantly from the football players.