Part I
Psychosocial processes underlying doping use
1 Doping in elite sport
Linking behavior, attitudes, and psychological theory
Kate Kirby, Suzanne Guerin, Aidan Moran, and James Matthews
Recent years have witnessed an upsurge of research interest in the psychosocial factors associated with competitive athletesâ propensity to use prohibited performance-enhancing drugs (PEDs; e.g., see Barkoukis et al. 2013; Hauw and Mohamed 2015; Hodge et al. 2013; Morente-SĂĄnchez and Zabala 2013). This practice is commonly known as âdopingâ and typically refers to athletesâ proclivity to use âillegitimate performance enhancement substances and methodsâ (Lazuras et al. 2010: 694). Although the problem of doping in sport may appear to be a relatively new phenomenon, it has a surprisingly long history. For example, prohibited substances such as caffeine and cocaine were used by cyclists in a bid to enhance competitive performance as far back as the 1890s (Hoberman 1998). Unfortunately, studies on doping in elite athletes are afflicted by at least two unresolved issues. First, the links between doping attitudes and doping behavior have not received sufficient research attention to date. Second, the role of psychological theory in elucidating these links has not been addressed adequately. Therefore, the purpose of the present chapter is to address these two issues. We shall proceed as follows. Following an overview of âattitudesâ and approaches to attitude measurement, we shall consider the measurement of attitudes to doping in sport. After that, we shall explain how the Theory of Planned Behavior (TPB; Ajzen 1991) has been applied to studies of doping in sport. Finally, we shall explore some potentially fruitful new directions for research in this field.
The nature of attitudes
In psychology, the term âattitudesâ refers to the preferences and evaluations (or likes and dislikes) that people form in relation to specific objects of their thought (Banaji and Heiphetz 2010). Reflecting this view, Eagly and Chaiken defined an attitude as âa psychological tendency that is expressed by evaluating a particular entity with some degree of favour or disfavourâ (1993: 1). Attitudes can be formed to groups of people (e.g., athletes), phenomena (e.g., social media), situations (e.g., competitions), and even to abstract ideas (e.g., sportspersonship), and social practices (e.g., doping in sport). Regardless of the particular target involved, however, attitudes are multidimensional in structure, with cognitive (i.e., beliefs about the target of the attitude), affective (i.e., positive or negative feelings toward such targets), and behavioral (i.e., whether or not any actions occur as consequences of oneâs evaluations) components. To illustrate how these components may interact, one might believe that athletes who engage in doping are âcheatsâ (cognitive component) who deserve scorn (affective component) and who should be banned from the sport in which they participate (behavioral component). Historically, many theorists have traditionally regarded attitudes as relatively stable and enduring personal dispositions that are stored in memory and can be âpulled outâ and used when required. On the other hand, more recent researchers such as Schwartz (2007) favor the view that attitudes are adaptive reactions to environmental demands â and hence, are temporary, context-specific judgments constructed from currently accessible information. In attempting to resolve this disagreement, Bohner and Dickel proposed that researchers should âtake into account both stable and situationally variable aspects of attitudesâ (2011: 394).
The assumption that attitudes predict behavior accurately is highly questionable. For example, a seminal review by Wicker (1969) of 42 relevant studies found that the correlation between attitudes and behavior ranged between 0.15 and 0.30. He concluded that overall âit is considerably more likely that attitudes will be unrelated or only very slightly related to overt behaviors than that attitudes will be closely related to actionsâ (1969: 64) because âonly rarely can as much as 10% of the variance in overt behavioral measures be accounted for by attitudinal dataâ (1969: 65). After Wickerâs (1969) review, social psychologists identified a host of variables moderating the relationship between attitudes and behavior. This line of research led to the conclusion that peopleâs behavior is influenced by many factors (such as situational constraints and peer pressure) other than attitudes.
Approaches to attitude measurement: from explicit to implicit
Scientific approaches to attitude measurement began in the 1920s (e.g., Thurstone 1928) and 1930s (e.g., Likert 1932) with the use of verbal self-report scales designed to assess peopleâs beliefs, opinions, and values. The rationale underlying this self-report approach is that people are both willing and able to accurately introspect on, and subsequently report, the contents of their own thoughts. Until the late 1970s, the use of such explicit self-report scales was the most popular approach in this field. However, around that time, two problems occurred which prompted researchers to shift from explicit to implicit (i.e., not accessible to conscious awareness) attitude measurement (Bohner and Dickel 2011; Banaji and Heiphetz 2010). First, response biases such as âsocial desirabilityâ emerged with the discovery that people tend to hide their true attitudes in an effort to present themselves in a positive light (e.g., by suppressing negative attitudes to groups such as immigrants). Second, the assumption that attitudes are open to introspective access was challenged by evidence (see review by Nisbett and Wilson 1977) that people often do not know â and hence cannot report reliably on â the reasons for their own behavior. In an effort to address these problems, certain implicit strategies for measuring attitudes (Fazio and Olson 2003) have been developed (see Chapters 3 and 7). According to Bohner and Dickel (2011), the objectives of these strategies are not only to counteract peopleâs response biases but also to explore those aspects of peopleâs attitudes that are not open to conscious introspection and control. Briefly, both of these approaches assume that evaluative associations in a respondentâs mind should produce different response times to categorical stimuli that represent the attitude object in question. These differences in response time may be used to infer implicit attitudes on the part of the respondent.
The measurement of attitudes to doping in sport â from athletes to entourage
Numerous studies have measured explicit attitudes to doping among a variety of athletic populations. Among the topics investigated in this largely atheoretical research literature are attitudes to doping in different sports (Alaranta et al. 2006), attitudes to doping and supplementation (Bloodworth et al. 2012), projected motives for doping in sport (Laure and Reinsberger 1995; Ăzdemir et al. 2005; Scarpino et al. 1990), and attitudes to drug testing and education (Striegel et al. 2002; Waddington et al. 2005). In such studies, researchers typically developed custom-built, doping-related questionnaires. Unfortunately, given the lack of psychometric detail available in this research, definitive conclusions regarding athletesâ attitudes to doping are not possible (Backhouse et al. 2007).
Fortunately, descriptive, âone-shotâ studies of attitudes of doping have been increasingly supplanted by theoretically driven research. For example, Sas-Nowosielski and Swiatkowska (2008) found that athletes who were high in ego orientation and low in task orientation displayed significantly more positive (permissive) attitudes to doping than those who displayed a low ego, high task orientation. Similarly, Kirby (2011) reported that certain subscales significantly predicted more lenient attitudes to doping, namely: coach criticism, unequal recognition, and ego orientation. These studies also determined that males tended to have significantly more favorable attitudes toward doping than females â a finding that has been replicated in other studies on doping in sport (Alaranta et al. 2006; Lucidi et al. 2008). The cornerstone of such research is the proposition that anti-doping education and deterrence methods cannot be fully effective unless athletesâ attitudes toward doping and their motives for banned substance use are more clearly understood (Lucidi et al. 2008).
In addition to the preceding research, some studies have explored the knowledge and attitudes of coaching staff in relation to doping in sport (Fung and Yuan 2006; Laure et al. 2001; Shirazi and Tricker 2005). Whereas Shirazi and Tricker (2005) set out to explore athletic directorsâ policies on drug testing, Laure et al. (2001) investigated the frequency with which coaches are asked for information about doping by their athletes. Unfortunately, little effort has been made to determine the factors that shape the coachesâ attitudes to doping, or to examine the factors that predicted coachesâ intentions to promote anti-doping among their athletes (Backhouse and McKenna 2013). Interestingly, according to Waddington and Smith (2009), âif we wish to understand the use of drugs in elite sport then it is crucial that we understand the centrality of the relationship between elite level athletes and practitioners of sports medicineâ (2009: 82). However, surprisingly few studies have attempted to understand the doping-related knowledge and attitudes of sports physicians. Indeed, as Backhouse et al. concluded, âthe literature in relation to doping and the medical population is weakâ (2007: 74). The earliest studies that examined attitudes to doping using physicians as participants focused on estimations of the prevalence of doping (Scarpino et al. 1990). This study by Scarpino et al. was significant because it was the first of its kind in Europe to assess the knowledge, attitudes, and perceived prevalence of doping among people involved in sport. However, very little information is provided about the characteristics and specific responses of the 102 physicians surveyed, except that 21 percent of them (compared to 30 percent of coaches, managers, and athletes) indicated that athletic performance can be enhanced by doping. Laure and colleagues examined attitudes to, and knowledge of, doping in sport among 202 French physicians (Laure et al. 2003) and retail pharmacists (Laure and Kriebitzsch-Lejeune 2000).
Unfortunately, as mentioned previously, researchers investigating the attitudes to doping of athletes, coaches and support staff have typically used custom-built measures rather than systematically validated ones. Indeed, many of the preceding studies did not report the rationale underlying their scale development process or provide adequate information on the psychometric adequacy of the scales used (Backhouse et al. 2007) â problems which hamper accurate interpretation of the results of the studies. Accordingly, Backhouse et al. (2007) concluded that direct comparisons across such studies are difficult to make. Furthermore, as the psychometric properties of these scales are rarely reported, doubts exist about the validity and generalizability of findings in this domain. Addressing this issue, PetrĂłczi and Aidman (2009) argued that when test scores are interpreted as attitudes and inferences are made about athletic populations based on these scores, adequate reliability and validity are essential requirements. Unfortunately, many of the studies that report athletesâ, coachesâ, and physiciansâ attitudes to doping fail to adequately define the term âdopingâ itself â or to give sufficient detail on the development or content of the questionnaires employed. In fact, Backhouse et al. (2007) suggested that the term âdoping attitudeâ had been so poorly defined that frequently the doping knowledge of athletes rather than their explicit attitudes has been surveyed.
An interesting addition to the research literature on doping has been the âPerformance Enhancement Attitude Scaleâ (PEAS; PetrĂłczi and Aidman 2009), which is a self-report measure examining generalized doping attitude (defined as âan individualâs predisposition toward the use of banned performance enhancing substances and methodsâ PetrĂłczi 2007: 7) (see also Moran et al. 2008). A positive association with elevated PEAS score and self-admitted doping has been reported (PetrĂłczi and Aidman 2008). Although this finding does not demonstrate conclusively that permissive attitudes to doping predict doping behavior, it does provide preliminary evidence that the two concepts are related.
Nevertheless, self-report scales (such as the PEAS) are marred by problems of response bias â or the tendency of participants to answer questions that reflect what they think the investigator wants to hear rather than as an index of their true beliefs. Recently, Morente-SĂĄnchez and Zabala noted that âthere could be a significant difference between what athletes say and what they really thinkâ (2013: 410). Interestingly, this issue of the validity of athletesâ self-reported attitudes to doping was examined by Gucciardi et al. (2010), who showed that, as expected, favorable attitudes to doping were associated with greater susceptibility to doping. However, the strength of this relationship between attitudes to doping and doping susceptibility was moderated by social desirability â a finding that highlights the importance of controlling for this latter variable when conducting self-report studies of doping in athletes.
To conclude this section, a promising new direction in doping measurement reflects the move toward the use of implicit rather than explicit assessment techniques (see also Chapters 3 and 7 of this volume). This move has been prompted by a desire to circumvent the biases afflicting explicit attitude assessment (for a review see Fazio and Olson 2003). Reflecting this new approach, Brand et al. (2014) argued that indirect attitude measures could prove valuable in studies of doping in sport. However, implicit measures are not unanimously favored. For example, PetrĂłczi (2013) provided a critical consideration of the use of response times in the indirect measurement of doping attitudes.
As psychological research in the field of doping has evolved, researchers such as Lucidi et al. (2008) and Barkoukis et al. (2013) have attempted to explain theoretically, rather than simply describe, key issues. More generally, recent studies have moved away from descriptive accounts and turned increasingly to social psychological theories (e.g., self-determination theory, Deci and Ryan 2000; achievement goal theory, Nicholls 1984) in their quest to understand the relationship between athletesâ attitudes to, and engagement in, doping behavior. Of the various psychological theories available to doping researchers, perhaps the most influential (e.g., see Lucidi et al. 2008; Goulet et al. 2010) is the Theory of Planned Behavior (TPB; Ajzen 1991). This theory, which was developed in order to explain behaviors which are not fully under volitional control, has been used extensively to predict health risk behaviors such as self-harm, driving behavior and substance abuse (see Ajzen 2014; Armitage and Conner 2001). Given this latter application, the TPB offers researchers a useful theoretical perspective from which to investigate doping behavior in sport.
Applying the Theory of Planned Behavior to doping in sport
The TPB has exerted a seminal influence on recent studies of doping in sport (e.g., Barkoukis et al. 2013; Chan et al. 2015). Although space limitations preclude a detailed explanation of the TPB, one of its central tenets is the...