Grzegorz Pajestka Social Courage at Work Scale – psychometric evaluation of the tool and preliminary evidence of its validity

Artykuł jest dostępny na warunkach międzynarodowej licencji 4.0 (CC BY-NC-ND 4.0).

Courage, present in philosophers’ considerations for millennia, has attracted the interest of psychologists relatively recently. This was mainly due to positive psychology, which creators see it as one of the six virtues determining a “good life”, next to wisdom, humanitarianism, justice, moderation (temperance), and transcendence (Peterson, Seligman, 2004). Courage is defined by them in a somewhat literary way, as emotional strength that, when engaged in will allow them to achieve goals in the face of adversity, both external and internal. These forces include authenticity (understood as honesty and truthfulness), bravery, perseverance, and vitality (understood as enthusiasm or life energy). However, courage defined this way has not gained much recognition among researchers dealing with this construct. The lack of a commonly accepted definition and tools for measuring courage resulted in a lack of empirical research on it (Howard et al., 2017). This research flourished with the appearance of the first questionnaires. One of them was the Woodard Courage Scale (2004), which, as a result of further research and psychometric analysis, was replaced by the Woodard and Pury Courage Scale (Woodard-Pury Courage Scale-23, WPCS-23; Woodard, Pury, 2007). Its authors assumed that courage should be defined as a voluntary act of action undertaken to achieve some important, perhaps moral, goal. This action may or may not be accompanied by the emotion of fear. This definition is consistent with the implicit theory of courage, the content of which was established based on a series of studies (Rate et al., 2007). They show that people consider an act of action to be brave if it is: a) voluntary and intentional, b) undertaken as a result of reflection, c) involving a significant risk for the actor, d) serving a noble (pro-social) goal, e) undertaken (perhaps) despite the fear experienced. However, there are some doubts regarding the last aspect of courage – it seems that from the perspective of an observer, it is not a key aspect of the experience of a person showing courage. Consequently, courage defined in this way could be identified with fearlessness.  

Norton and Weiss (2009) define courage in an even different way – for them, experiencing fear or anxiety is a key aspect of courage, which they understand as persistence or perseverance despite having fear (p. 213). Earlier, a similar definition was proposed by Putman (1997) who called overcoming one’s irrational fears psychological courage. Putman also proposed two other categories of courage: physical – related to the risk of loss of health or life, and moral – referring to the risk taken in defense of some moral good. These three types of courage are mentioned most often in the literature, sometimes with some modifications (Putman, 1997; Lopez, O’Byrne, Peterson, 2003; Pury, Kowalski, Spearman, 2007; Howard et al., 2017). For example, in the typology proposed by Lopez et al. (2003), instead of psychological courage, there is vital courage as a reaction to the disease and side effects of treatment. Some authors also distinguish, apart from moral courage, civil courage, understood as behavior accompanied by anger and indignation, and which aims to enforce social and ethical norms, without taking into account its own social costs (Kastenmüller et al., 2007). The definition of civil courage, in turn, is close to the definition of social courage proposed by Howard et al. (2017), which refers to the implicit theory of courage proposed by Rate et al. (2007). Social courage is therefore an intentional, after-thought, altruistic act that exposes the actor to social costs in the form of loss of respect from others, e.g. co-workers.

These costs may include loss of face (social image) or deterioration of relationships with others. These may be caused by the actor’s behavior consisting in reprimanding co-workers who make mistakes at work or reporting detected irregularities to their superiors for the good of the organization (whistleblowing).  

Courage is described in the literature as a trait (disposition) that promotes courageous behavior (Kastenmüller et al., 2007; Woodard, Pury, 2007). In another approach, it is treated behaviorally – researchers ask about how brave people behave, not what they are like (Rate et al., 2007; Howard et al., 2017). In other words, they are interested in the act of action itself, which results in the actor being recognized by the observer as brave. This division corresponds to the proposal of Pury and Starkey (2010), according to which courage can be treated in terms of accolade (distinction, glory) or a process. Courage as an accolade refers to acts that can be described as spectacular and heroic – so these will be primarily acts of physical courage, such as saving someone’s life. Importantly, these acts must succeed, otherwise, they do not meet the criterion of courage categorized in this way. It does not matter whether the person committing such an act experienced fear or not – what matters is that they took a risk for a noble aim and the action was successful. Demonstrating this type of courage is usually recognized by society, which is reflected in official awards, e.g. medals.  

Courage should be understood differently as a process. An act of courage may be mundane (overcoming one’s fear of, e.g. flying or speaking in public), nor does it have to succeed. What is important is that a person tries to do something, to face some adversity or limitations. The definition of courage proposed by Norton and Weiss (2009) fits well into this processual understanding of courage. Researchers emphasize that courage understood in this way is important in the process of trauma therapy involving exposure to a threatening stimulus. Lack of courage may result in avoidant behavior, as a result of which the person does not undertake therapy or discontinues it. In such a situation, psychological courage would be, apart from anxiety, a factor determining or moderating the basic behavioral tendency of approach/avoidance. Norton and Weiss (2009) conducted a study aimed at verifying this hypothetical relationship. The participants of the study were people with an increased level of fear of spiders, who believed that they should approach the threatening stimulus (tarantula) at a distance they considered safe. This study found that psychological courage, next to anxiety, is a significant predictor of behavioral rapprochement. The results of this study – the first in which courage was tested empirically – supported the concept of courage as perseverance despite fear/anxiety, while proving the criterion validity of the instrument developed by Norton and Weiss – the Courage Measure. This study was the first and for a long time the only one in which courage was tested empirically. Subsequent studies using an analogous procedure, i.e. the Behavioral Approach Test and the Courage Scale, led to similar conclusions: psychological courage is a variable manifested in behavior; as a behavioral tendency, it has a different direction than anxiety, but is not its simple opposite – it makes a unique contribution to the explained variance of behavioral approach (Chockalingam, Norton, 2019) or moderates the effect of negative emotions so that the relationship between anxiety and behavioral approach becomes less negative (Pajestka, 2023). 

The Courage Scale became a popular research tool, but its operational definition of courage raised theoretical doubts. Howard and Alipour (2014) argue that this scale does not measure courage, but a construct they call Persistence Despite Fear (PDF). Assuming that the most theoretically refined and empirically grounded definition of courage is the one proposed by Rate et al. (2007) who created their own tool for measuring social courage at work and in organizations.  

Although courage is likely a multidimensional construct, previous attempts to create instruments that comprehensively address this construct have been unsuccessful because the resulting scales were characterized by an unclear factor structure (Woodard, 2004; Woodard, Pury, 2007). Howard et al. (2017), proposed to adopt a simpler solution to these difficulties and create tools examining individual types of courage separately, as one-dimensional constructs. Such a tool is the Workplace Social Courage Scale they created. The psychometric properties of the WSCS in American studies were satisfactory and in this sense the scale turned out to be a better tool than other popular measures of courage, such as WPCS-31 (Woodard, Pury, 2007) or CM (Norton, Weiss, 2009). These studies also revealed that social courage is a construct different from risk-taking and prosocial motivation, although both of these variables are related to social courage by definition. Moreover, testing the relationship of the WSCS with personality traits (Big Five), basic self-esteem, as well as organizational variables such as citizenship behavior, voice, and pro-social rule-breaking, yielded results supporting the validity of the scale and its usefulness in organizational research (Howard et al., 2017).  

The psychometric properties of the WSCS were also tested in studies conducted in Italy. On their basis, it was determined that the shortened, 8-item version of the scale fits the data better than the original version of the scale consisting of 11 items. Three items that were excluded from the scale are, according to the authors of the Italian adaptation, characteristic of American culture and the local work ethos, but they have no reference in Italian culture. Therefore, WSCS turned out to be partially culture-dependent, which constitutes a certain limitation for researchers planning to use it in international research. Despite this, an attempt was made to adapt this scale to Poland. The main reason is the lack of tools in the Polish language to test courage, apart from the recently adapted Norton and Weiss Courage Scale (Pajestka, 2023). Taking into account the satisfactory psychometric properties of the original version of the scale and the validity of the tool verified in American and Italian studies (for the shortened version), this decision seems justified. It is also important that this scale is based on a coherent concept of courage developed in 2007 by a group of prominent researchers: Christopher R. Rate, Jennifer A. Clarke, Douglas R. Lindsay and Robert J. Sternberg. The lack of such a well-established theoretical background is a weakness of many other scales developed in the English language (Howard et al., 2017).  

To adapt the WSCS scale to Polish, two studies were conducted using data from two research samples. The first study tested the psychometric properties of the scale – its factor structure and internal consistency. Additionally, an analysis of the measurement equivalence of the tool was performed in groups differing in age and gender. The second study verified the convergent and discriminant validity of the tool as well as the predictive capabilities of the construct.

Study 1: adaptation of the scale and analysis of its psychometric properties 

Method 

Respondents 

Sample 1 consisted of 234 respondents aged 20 to 65 years (Mage = 41.75, SDage = 13.21, 53.4% female). All respondents are working people. Detailed characteristics of the sample are presented in the supplementary materials. 

Tools 

Howard et al.’s Social Courage at Work Scale (2017). In the Polish version – the scale consists of 11 items to which respondents answer using a 7-point Likert scale, where 1 = I strongly disagree and 7 = I strongly agree.  

Procedure 

Scale translation procedure. After receiving the consent of the first author of the WSCS scale for its adaptation (MH), it was translated into Polish by two psychologists fluent in English – each of them translated it independently. Then, two researchers (one of them was the author of this article) based on two versions of the translation, established one, which was subjected to a back-translation procedure carried out by a professional English translator. The original and back-translated versions were rated as compatible by a person who spoke American English as a native language. The translated scale was subjected to psychometric verification.  

Data collection procedure. Participants were recruited using the Ariadna Nationwide Research Panel. The socio-demographic profile of people registered in the Panel reflects the profile of Poles using the Internet. Their participation in the study was voluntary, for which they were rewarded with points exchanged for gifts, in accordance with the Panel’s loyalty program. Applying recommendations for data quality assurance in Internet research (Berinsky, Margolis, Sances, 2014), in Sample 1 two attention tests were administered in the form of instructions such as: “If you are reading this, check three.” These commands were placed in the middle and end of the form, between the items of the questionnaires used to validate the WSCS (see Study 2). Sample size and all statistics are reported only for people who passed both attention tests, data from the remaining participants (eight people) have been removed.  

Data analysis procedure. In study 1, to verify the psychometric properties of the Polish version of the WSCS scale (WSCS-PL), internal consistency analysis (Cronbach’s α) and confirmatory factor analysis (CFA) were conducted. Using the sample size calculator for structural equation models (Soper, 2022), it was calculated a priori that the minimum sample size sufficient to verify the model structure, assuming a minimum predicted effect size of .10 and a minimum statistical power of .80 is 200 subjects. Data collected in Study 1 were also used to verify the measurement equivalence of the tool concerning gender and age. For this purpose, multigroup confirmatory factor analysis (MGCFA) was performed. Groups differing in age were distinguished based on the division proposed by other authors (e.g. Ginevra et al., 2020). People aged up to 39 years were included in the group of young adults (n = 109), and people aged 40 to 65 years (n = 125) were included in the group of middle-aged adults.

Data analysis 

Scale structure analysis. To confirm the one-dimensional structure of the WSCS-PL, CFA was performed using the maximum likelihood method. Two models were tested, in model 1 the properties of the full version of the scale were tested, and in model 2 properties of the shortened version. The fit of the model to the data was determined using the following indicators: normalized chi-square (χ2/df), CFI (comparative fit index), SRMR (standardized root-mean-squared residual), and RMSEA (root-mean-square error of approximation). Conventional criteria for assessing model fit were used: χ2/df ≤ 3:1, CFI ≥ .95, SRMR ≤ .08, RMSEA ≤ .08 (Hooper, Coughlan, Mullen, 2008; Hair et al., 2014). Due to the violation of the assumption of multivariate normal distribution of variables, the bootstrap technique was used with 1000 samples drawn with a confidence interval of 95% (Brown, 2015). 

Scale measurement equivalence analysis. Using MGCFA, three different levels of measurement equivalence were tested: configurable (construct), metric, and scalar. Since the measurement equivalence analysis process is comprehensively described in the literature (e.g. Lubiewska, Głogowska, 2018), it will only be briefly presented below. Testing configural equivalence allows us to answer the question of whether we measure the same or a different construct in different groups. Metric equivalence refers to equal factor loadings between groups, and its confirmation means that the measurement unit used (e.g. points on a Likert scale) is equal (interpreted in the same way, identical) in all studied groups. Scalar equivalence refers to the equivalence of regression constants (intercepts) for individual scale items in the compared groups. Establishing scalar equivalence validates testing of differences between means in the study groups; establishing metric equivalence only allows the use of a tool to examine relationships between variables analyzed in different groups for which this level of equivalence has been established. Therefore, this is the minimum level of equivalence that should characterize a tool used in scientific research whose aim is to study the relationships between various variables in different groups. Following the recommendations of Chen (2007), all three levels of equivalence were assessed by the deterioration of the values of the CFI and RMSEA parameters, and the cut-off points were the values recommended for small samples (N ≤ 300; ΔCFI ≥ .005; ΔRMSEA ≥ .010). The Δχ2 value was provided for informational purposes only and its significance was not taken into account when determining subsequent levels of measurement equivalence (Putnick, Bornstein, 2016). Psychometric analyses were performed using the SPSS statistical package (version 28) and SPSS Amos (version 26).

Results 

Psychometric properties of the scale 

The reliability (Cronbach’s α) of both versions of the scale turned out to be satisfactory and amounted to .85 and .88 for the 11- and 8-item versions, respectively. However, these versions differed in the goodness of fit of the models to the empirical data, which is presented in Table 1. The fit of model 1 turned out to be poor. Model 2, by the

adopted evaluation criteria, turned out to be a fairly good fit for the data. The exception was the RMSEA index, which exceeded the limit value adopted in this study. Taking into account the values of modification coefficients, it was decided to modify the model by allowing one pair of errors to covariate (items 2 and 7). This improved model turned out to be a good fit for the data (model 2’).  

Table 2 shows the factor loadings for both versions of the scale. In the full version, items 1, 4 and 10 did not exceed the value recommended as the minimum for factor loadings in CFA (> .50, Hair et al., 2014). In the shortened version of the scale, all factor loadings reached satisfactory values. Taking into account the above results, the measurement equivalence analysis and other psychometric analyses were conducted using only the 8-item version of the scale.  

As shown in the data presented in Table 3, taking into account gender, changes in the RMSEA and CFI values are within the limits adopted in this study as indicators of measurement equivalence at the configurational, metric, and scalar levels. The data presented in Table 4 indicate a similar pattern in the case of two different age groups.  

Table 1. Indicators of fit of tested models to empirical data

Table 2. Items, standardized factor loadings and Cronbach’s α internal consistency coefficients of the full and shortened versions of the WSCS-PL

Table 3. Equivalence of measurement with the short version of WSCS-PL in groups differing by gender

Table 4. Equivalence of measurement with the short version of WSCS-PL in groups differing in age

Discussion 

The study confirmed the one-factor structure of the WSCS-PL scale, but its full version turned out to be a poor fit for the data. The factor loadings of the three items were lower than the cut-off point adopted in this study (.50). Removing these items significantly improved the model performance. In the shortened scale consisting of 8 items, the indicators included in the assessment met the criteria for good model fit adopted in the study. The exception was RMSEA, whose value turned out to be slightly higher than the adopted cut-off point. Taking into account the covariance of one pair of errors improved the fit of the entire model, including the RMSEA index, which had an acceptable value in this situation. In CFA, a model fit improvement procedure based on modification factors can be considered justified if items whose errors are correlated load the same factor, because the content similarity between these statements may be responsible for this correlation (Brown, 2015). Given the similarity between items 2 and 7 (both reprimanding co-workers), the use of this procedure can be considered justified.  

The items that were removed from the WSCS-PL scale are the items corresponding to those removed in the Italian version of the scale (Magnano et al., 2022). Although item 1 (Table 2) describes behavior that meets the definitional criteria of social courage, it may as well be interpreted as informing. A person who behaves in the manner described in item 1 may pursue their own goals rather than those of the organization, be extremely competitive, and strive for success at the expense of their co-workers. Polish study participants, like the Italians surveyed, are unlikely to consider such behavior to be brave. The next two items removed (4 and 10, Table 2), as argued by Magnano et al. (2022) reflect the American work ethos, a particular willingness to sacrifice one’s good for the benefit of the organization, which is not typical of Italian workers. Based on the work of Weber (2011) on the influence of Protestant ethics on the development of capitalism, it can be assumed that a similar interpretation is also accurate in the case of Polish workers. The work ethic in Italy and Poland, as Catholic countries, is probably similar and at the same time different than in the USA, a country where the Protestant work ethic is dominant (Feldmann, 2007; Modrack, 2008). However, another explanation also seems plausible, based on one of the dimensions of culture proposed by Hofstede (2001). This dimension is uncertainty avoidance – both Poland and Italy, unlike the USA, rank high on this dimension (Hofstede Insights, 2021). It turns out that this dimension is related to how people behave when their social image is threatened (loss of face). In countries with high uncertainty avoidance, people are more sensitive to the reactions of others and are more afraid of losing face (Merkin, 2017). In response to a threat to their social image, they may react aggressively (Merkin, 2006). Therefore, this threat seems to be more costly for them than for people from cultures that rank low on the uncertainty avoidance dimension. The behaviors described in both excluded items, which concern the direct risk of losing face (I would look stupid, and others would think less of me), in countries with a high index of uncertainty avoidance do not have to be interpreted as examples of courage, but – for example, as a lack of honor (Szmajke, Bąk, Adamus, 2004). It should be noted, however, that two items (3 and 9, Table 2), which describe the situation of compromising one’s image, were characterized by relatively high factor loadings. This means that not every situation in which an actor faces a threat to his or her image is interpreted similarly. Whether the behavior will be considered a manifestation of courage depends on additional factors, but unfortunately, the results of this study do not allow us to answer the question of which ones. One can only assume that admitting a mistake or being able to ask for help are situations in which the actor’s intentions seem clear – someone wants to correct their mistake or understand the task well so as not to make a mistake. It is more difficult to understand the intentions of a person who voluntarily undertakes a project with a chance of failure or speaks in public, even though for some reason he or she looks “dumb” in this situation. It may also be the case that in our culture, these situations are associated more with shame (which should be avoided) than fear or anxiety, which is a reaction to a challenge, and overcoming which can be a manifestation of courage in itself (Norton, Weiss, 2009). It follows that the type of emotions experienced would be the factor that would differentiate brave behavior from other risk-taking behaviors. Courageous behavior, from the actor’s perspective (though not necessarily from the observer’s perspective, cf. Rate et al., 2007), involves acting despite experiencing fear or anxiety, rather than shame or guilt. I believe this hypothesis is worth verifying in further research. 

The obtained results allow us to conclude about the measurement equivalence of the tool for the groups distinguished in the study, differing in gender and age. The full equivalence of the WSCS-PL scale at the configurable, metric, and scalar levels means that social courage measured by this scale is understood similarly regardless of the gender and age of the study participants. The WSCS-PL also turned out to have satisfactory internal consistency measured by Cronbach’s α coefficient. To sum up, the psychometric properties of the scale in its shortened, 8-item version turned out to be satisfactory. The next step was to test the validity of the scale.

Study 2: criterion, convergent, and discriminant validity of the WSCS-PL scale

Since, as I mentioned earlier, there are no courage questionnaires in Polish, the selection of tools to examine the convergent validity of the WSCS-PL was limited to the recently adapted short version of the Courage Scale (CM) into Polish. CM was also used for this purpose in the original study because, according to the authors, it is a better tool than alternatives available in English. Additionally, the relationship between WSCS-PL and resilience was examined. Resilience, defined as the ability to overcome adversity and achieve success despite exposure to high risk, seems particularly close to the definition of social courage adopted by Howard et al. (2017). Both constructs involve taking risks and overcoming difficulties to achieve some goal. The difference between these constructs comes down to the fact that in the case of social courage, by definition, this goal is external – pro-social, while in the case of resilience, it is rather internal – coping and achieving something for oneself.  

Courage examined by self-report, like other socially desirable traits or behaviors, may be particularly sensitive to the need for social approval of the respondents. A strong relationship between these two variables would mean that the WSCS-PL scale is of little use in predicting courageous behavior at work. As for the usefulness of the scale, according to the authors’ intention, the WSCS should predict variables that are important from an organizational point of view. One such variable is organizational citizenship behaviors (OCBs). OCBs are defined as behaviors undertaken intentionally for the good of the organization that goes beyond the employee’s scope of duties. These are prosocial behaviors that involve some risk. For example, employees who voluntarily perform additional work risk ostracism from coworkers who may perceive such activity as threatening: increasing the employer’s expectations regarding employee effectiveness, breaking group solidarity, and reducing the assessment of one’s value as an employee (Jensen, Patel, Raver, 2014). OCBs therefore involve social risks that may be more easily taken by people who demonstrate social courage. If so, then the hypothesis that social courage is a predictor of citizenship behavior is justified. 

As proposed by Howard et al. (2017), some personality traits in the Big Five model, i.e. extraversion and conscientiousness, should promote courageous behavior, while neuroticism should hinder courage. However, in his research, the relationship between these features and social courage turned out to be relatively small – the highest correlations were for extraversion (r = .35) and openness to experience (r = .32), the lowest (r ≤ .20) for agreeableness, neuroticism and conscientiousness. Based on these data, it can be assumed that also in this study, the relationship between personality traits and social courage will be relatively small. At the same time, it should be noted that formulating hypotheses about the relationship between personality traits and other variables should be preceded by indicating the model based on which these relationships are derived because there are more or less subtle differences between different models, e.g. Goldberg’s lexical model vs. Costa and McCrae’s model (Strus, Cieciuch, Rowiński, 2014). In this study, I base predictions on Goldberg’s model and Strus et al.’s description of the Big Five traits (2014). Taking into account the characteristics of these features, it is difficult to identify those that are clearly associated with courageous behavior. If we assume, according to the definition, that courageous behavior is an intentional, deliberative and prosocial act that involves a certain risk taken by the actor and may be accompanied by a feeling of fear, it seems that it may apply to both extroverts and introverts, people less and less emotionally stable, etc. On this basis, I predict that social courage measured by the WSCS-PL scale is relatively independent of the personality traits described in the Big Five model. Additionally, in this study, I assume, following other authors (Byrne, Peters, Weston, 2016), that when examining convergent and discriminant validity, two variables can be considered related to each other if the absolute value of the correlation coefficient between them exceeds .30. Coefficients equal to or lower than this value, even if statistically significant, will be interpreted as no relationship between the variables. Although this approach may be considered arbitrary, I adopt it here consciously to avoid confirmation bias in testing the above hypotheses about the relationship between social courage and other variables. 

Method 

Respondents 

Sample 2 consisted of 225 respondents aged 20 to 65 years (Mage = 40.97, SDage = 13.16, 53.8% female). All study participants are working people. Their detailed characteristics can be found in the additional materials. Study 2 also used data from the previously described sample 1. 

Tools 

Social courage. The shortened Scale of Social Courage at Work, Polish version (WSCS-PL) was used. The scale was described in Study 1.  

Psychological courage. The Courage Scale (Norton, Weiss, 2009) is used to measure courage understood as perseverance or persistence despite feeling fear. The current study used its shortened version, which is superior to the full version in terms of psychometric properties (Howard, Alipour, 2014). The scale in the Polish language version has good psychometric properties, and its criterion validity has been behaviorally verified (Pajestka, 2023). The tool consists of six items, rated on a 7-point Likert-type scale (1 = strongly disagree to 7 = strongly agree). An example statement is: “I try to face my fears.”  

Personality. Personality traits: extraversion, agreeableness, conscientiousness, emotional stability and intellect were measured using the IPIP-BFM-20 questionnaire (Topolewska et al., 2014). The questionnaire is a shortened version of the IPIP-BFM-50 questionnaire (Goldberg, 1992), measuring the Big Five personality traits from a lexical perspective. IPIP-BFM-20 consists of 20 items (four for each of the measured features), to which the examined person responds using a 5-point scale (1 = describes me completely incorrectly, 5 = describes me completely accurately).  

Psychological resilience. The Psychological Resilience Scale (SPP-25) by Ogińska-Bulik and Juczyński (2008) was used to measure resilience as a personality trait that helps cope with stress. The questionnaire consists of 25 items measuring the general level of psychological resilience, as well as its five factors: “perseverance and determination in action”; “openness to new experiences and sense of humor”; “personal competencies to cope and tolerance of negative emotions”; “tolerance for failure and treating life as a challenge”; “optimistic attitude to life and the ability to mobilize in difficult situations.” The questionnaire used a 5-point Likert scale (from 0 – definitely not, to 4 – definitely yes). In the present study, the overall score on the SPP-25 scale was used as a variable. 

Organizational citizenship behavior. The Organizational Citizenship Behavior Questionnaire by Czarnota-Bojarska (2011) was used. This questionnaire consists of 33 statements describing behavior beneficial to the organization, to which the participant responds using a 6-point Likert scale, where 0 means very unusual behavior and 5 means very typical behavior for the research participant. Examples of statements from this questionnaire include: “On my own initiative, I undertake work that is not formally required of me”, and “For the good of the company, I do not hesitate to oppose the majority.” The questionnaire statements refer to various categories of behavior: organizational loyalty, organizational obedience, perseverance, helping others, initiative and civil virtue (understood as a sense of responsibility for the organization). In the study, following the example of other authors (e.g. Szalkowska, Żemojtel-Piotrowska, Clinton, 2015), the overall score in the questionnaire was used as an indicator of citizenship behavior. 

The need for social approval. The Social Approval Questionnaire (KAS) by Drwal and Wilczyńska (1980) was used. It is a questionnaire containing 29 true/false statements. An example statement is “I don’t always tell the truth.” KAS correlates positively with similar measures of the need for social approval: Edwards’ SDS and Marlowe and Crowne’s SDS (Drwal, 1995).

Procedure 

Data collection procedure. As in the case of sample 1, participants were recruited using the Ariadna Nationwide Research Panel. In study 2, a control of respondents’ attention was used, analogous to study 1 (two attention tests placed in the middle and final part of the form, among the statements of the questionnaires validating the accuracy of the WSCS). The reported number of participants in sample 2 is for those who passed both attention tests. Data collected from participants who failed them (nine people) were treated as low quality and removed from the database.  

Data analysis procedure. To examine the convergent and discriminant validity, a series of Pearson correlation analyses of the WSCS-PL with selected variables were performed. Then, to verify the hypothesis that social courage is a predictor of organizational citizenship behavior, a regression analysis was carried out using the input method, including the need for social approval in the model as a control variable. All assumptions for multivariate regression analysis were met (Williams, Grajales, Kurkiewicz, 2013). 

Results 

Table 5 presents the descriptive statistics, and Table 6 presents the internal consistency coefficients and Pearson’s r correlation coefficients of all variables analyzed in Study 2. Table 7 contains the results of the multiple regression analysis testing the value of the WSCS-PL as a predictor of citizenship behavior while controlling for the social approval variable. 

Table 5. Descriptive statistics of variables included in the study

Table 6. Pearson’s r correlations and internal consistency coefficients (Cronbach’s α) for the variables included in Study 2 

Table 7. Results of the regression analysis of social courage on organizational citizenship behavior while controlling for the need for social approval 

Discussion 

The conducted correlation analyses support the convergent and discriminant validity of the WSCS-PL scale. The scale correlates positively with another measure of courage - CM-PL, as well as with psychological resilience tested with the SPP-25 scale. The relationship between the WSCS-PL and the CM-PL scale is, as in the American studies, moderate. Although both scales measure courage, they have different operational definitions. WSCS-PL is based on a definition derived from the implicit theory of courage, behaviorally understood as an intentional, voluntary and prosocial act undertaken despite the associated risk (for the actor). CM-PL was created based on clinical psychology, based on the definition of courage as a feature that manifests itself in perseverance in action despite experiencing fear. Therefore, psychological courage examined by CM-PL seems to be more related to psychological resilience than to social courage, which is confirmed by the high correlation coefficient between these two variables. Both variables (psychological courage and psychological resilience) concern overcoming obstacles and coping with difficult situations, which are similar to social courage. At the same time, both of these variables are directed pro-self, which distinguishes them from social courage, which, by definition, is directed pro-social.  

The altruistic aspect of social courage measured by the WSCS-PL scale was reflected in the results of the regression analysis, according to which it is a significant predictor of citizenship behavior in the organization. Based on the results obtained, it can be concluded that people demonstrating social courage are willing to take actions beneficial to the organization, going beyond their basic duties. As I argued earlier, such a prosocial attitude may have negative consequences for the actor from their coworkers who perceive such behavior as threatening. In other words, carrying out additional tasks in the organization may be a manifestation of social courage when they are undertaken in the face of the risk of losing face or good relations with coworkers. Pointing out to coworkers that they make mistakes, reporting irregularities in the production process to their superiors, disregarding occupational health and safety regulations, and discriminating against other coworkers are examples of such behaviors that are costly for the actor. People who undertake them behave bravely because they take risks for the greater good, which may be the health or even the life of their coworkers, or the broadly understood success of the organization in which they work.  

The results obtained in the regression analysis support the criterion validity of the WSCS-PL – its relationship with prosocial behavior is understandable, taking into account the components of the definition of social courage, where the altruistic goal of behavior is equally important as intentional, thoughtful risk-taking. The validity of the WSCS-PL scale is also supported by the fact that these results are consistent with the validation results obtained by Howard et al. (2017). In these studies, social courage tested with the full version of the WSCS was a significant predictor of organizational citizenship behavior. 

Social courage measured by the WSCS-PL scale turned out to be relatively independent of personality traits. The correlation coefficients did not exceed .30, which, by the cut-off point for relationships between two variables adopted in this article, supports the discriminant validity of the tool. Social courage also turned out to be independent of the need for social approval. Although as a measure – as one might assume – of socially desirable behavior that brings recognition, it seems particularly sensitive to such a need, the low correlation with the KAS did not confirm these concerns.

General discussion 

The aim of the research project presented in this article was to adapt the Social Courage at Work Scale by Howard et al. into Polish (2017). Courage is a new variable in psychology, which is much easier to study today than it was two decades ago. In recent years, there has been a theoretical development of this construct and, as a consequence of this development, a flourishing of empirical research on courage. These were possible (and certainly easier) to carry out thanks to the constantly emerging new scales for examining courage and courageous behavior. The scale presented in this work is one of many such tools, most of which were published in English and their creators were mostly American researchers. The consequence of this is difficulties encountered in the adaptation process – it turned out that not all statements have an analogous meaning in Polish and American culture. As a result, the Polish version of the scale, like the Italian version, consists of only 8 items. The shortened version of WSCS-PL turned out to be a reliable and accurate tool. The results of confirmatory factor analysis indicate a one-factor structure of the scale consistent with the original. The results of the multi-group confirmatory analysis prove the measurement equivalence of the tool in groups differing in age and gender. Establishing equivalence at the metric and scalar levels proves the value of WSCS-PL as a tool useful in scientific research, both those whose main goal is to study relationships between variables and those in which it is required to compare mean results in groups differing in gender and/or age.  

The results obtained in study 2 support the validity of the WSCS-PL scale – the scale correlated positively with scales measuring similar constructs – psychological courage and psychological resilience. Its relationship with prosocial behavior in the form of organizational citizenship behavior supports the criterion validity of the tool, and low correlations with personality traits support the discriminant validity. 

The scale presented in this article, despite good psychometric properties and evidence supporting its validity, also has weaknesses. One of them is the structure of items – they have a complex form, which undoubtedly may make their unambiguous interpretation difficult. However, taking into account the definition of courage adopted as the basis for the construction of the scale, such a way of formulating statements was, according to the authors of the WSCS, inevitable. One item had to include information about the social risk taken and its consequences implying the pro-social nature of the behavior (e.g. 8. Even if it could damage our relationship, I would confront a subordinate who is disrupting the group’s work.). A version of the scale containing only a description of the behavior, without an accompanying description of the consequences of that behavior, was tested by Howard et al. (2017). It turned out that this scale had an unclear factor structure and evidence for its criterion validity was weak. The authors stuck to complex sentences as they better reflect the meaning of courageous behavior. 

Limitations and directions for further research 

The adaptation studies carried out have some weaknesses. Although the samples used are relatively diverse, they are based entirely on participants recruited online. In subsequent studies, care should be taken to confirm the obtained research results on a group of people tested using the paper-and-pencil method. Another weakness of the conducted research is the relatively small number of variables analyzed in the context of the validity of the WSCS-PL scale. In the study by Howard et al. (2017), it turned out that WSCS is associated with several organizational variables not included in this study, such as employee voice, counterproductive workplace behaviors (CWB), prosocial rulebreaking (PSRB) and performance. Most of these variables were measured by scales that have not yet been adapted into Polish – at least until the research described in this work was conducted. The exception is counterproductive behavior, whose negative relationship with social courage, shown in the original study, should be verified in subsequent studies using the WSCS-PL scale. The weaknesses of the research presented in this work can be easily eliminated in the future, contributing to enriching knowledge about both the accuracy of the WSCS-PL scale and knowledge about courage itself, as a new variable in empirical psychology.  

The theoretical outline presented in this article only slightly reflects the complexity of courage as a psychological construct, which may be considered its weakness. However, it should be read as an introduction to research on courage, not only social courage. This article is also the beginning of work on the author’s concept of courage, taking into account the specificity of Polish culture, if not at the theoretical level (after all, there is a chance that this construct is to some extent universal), then certainly at the operational level, in the form of a new courage questionnaire. The adapted tool is a measure that is intended to allow Polish research to be related to research conducted in other countries (so far in the USA and Italy), and thus help determine what is common to people from different countries in terms of courageous behavior. 

Translated by Katarzyna Jenek

References 

Berinsky, A.J., Margolis, M.F., & Sances, M.W. (2014). Separating the shirkers from the workers? Making sure respondents pay attention on self‐administered surveys. American Journal of Political Science, 58(3), 739–753, doi: 10.1111/ajps.12081 

Brown, T.A. (2015). Confirmatory factor analysis for applied research. New York: Guilford Publications. 

Byrne, Z.S., Peters, J.M., & Weston, J.W. (2016). The struggle with employee engagement: Measures and construct clarification using five samples. Journal of Applied Psychology, 101(9), 1201, doi: 10.1037/0033-2909.105.3.456 

Chen, F.F. (2007). Sensitivity of Goodness of Fit Indexes to Lack of Measurement Invariance. Structural Equation Modeling: A Multidisciplinary Journal, 14(3), 464–504, doi: 10.1080/10705510701301834 

Chockalingam, M., & Norton, P.J. (2019). Facing fear-provoking stimuli: The role of courage and influence of task-importance. The Journal of Positive Psychology, 14(5), 603–613, doi: 10.1080/17439760.2018.1497685 

Czarnota-Bojarska, J. (2011). Zachowania obywatelskie w organizacji – próba konstrukcji narzędzia pomiaru [Citizenhip behaviors in organization – the attempt of construction of measurement tool] (pp. 145–157). In J. Czarnota-Bojarska, & I. Zinserling (Eds.), W kręgu psychologii społecznej. Warszawa: Wydawnictwa Uniwersytetu Warszawskiego. 

Drwal, R. (1995). Adaptation of personality questionnaires. Warszawa: Wydawnictwo Naukowe PWN. 

Drwal, R., & Wilczyńska, J.T. (1980). Opracowanie kwestionariusza aprobaty społecznej [Elaboration of the social desirability questionnaire]. Przegląd Psychologiczny, 23(3), 569–583. 

Feldmann, H. (2007). Protestantism, labor force participation, and employment across countries. American Journal of Economics and Sociology, 66(4), 795–816, doi: 10.1111/j.1536-7150.2007.00540.x 

Goldberg, L.R. (1992). The development of markers for the Big-Five factor structure. Psychological Assessment, 4(1), 26–42, doi: 10.1037/1040-3590.4.1.26 

Ginevra, M.C., Santilli, S., Camussi, E., Magnano, P., Capozza, D., & Nota, L. (2020). The Italian adaptation of courage measure. International Journal for Educational and Vocational Guidance, 20(3), 457–475, doi: 10.1007/s10775-019-09412-4 

Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E., & Tatham, R.L. (2014). Pearson new international edition: Multivariate data analysis. Exploratory Data Analysis in Business and Economics, 23–60. 

Hofstede, G.H. (2001). Culture’s consequences: Comparing values, behaviors, institutions and organizations across nations. Thousand Oaks, CA: Sage Publications. 

Hooper, D., Coughlan, J., & Mullen, M.R. (2008). Structural equation modeling: Guidelines for determining model fit. Electronic Journal of Business Research Methods, 6(1), 53–60, doi: 10.21427/D7CF7R 

Howard, M.C., & Alipour, K.K. (2014). Does the courage measure really measure courage? A theoretical and empirical evaluation. The Journal of Positive Psychology, 9(5), 449–459, doi: 10.1080/17439760.2014.910828 

Howard, M.C., Farr, J.L., Grandey, A.A., & Gutworth, M.B. (2017). The Creation of the Workplace Social Courage Scale (WSCS): An Investigation of Internal Consistency, Psychometric Properties, Validity, and Utility. Journal of Business and Psychology, 32(6), 673–690, doi: 10.1007/s10869-016-9463-8 

Jensen, J.M., Patel, P.C., & Raver, J.L. (2014). Is it better to be average? High and low performance as predictors of employee victimization. Journal of Applied Psychology, 99(2), 296–309, doi: 10.1037/a0034822 

Kastenmüller, A., Greitemeyer, T., Fischer, P., & Frey, D. (2007). Das Münchner Zivilcourage-Instrument (MüZI). Diagnostica, 53(4), 205–217, doi: 10.1026/0012-1924.53.4.205 

Lopez, S.J., O’Byrne, K.K., & Peterson, S. (2003). Profiling courage. In S.J. Lopez, & C.R. Snyder (Eds.), Positive psychological assessment: A handbook of models and measures (pp. 185–197). Washington, DC: American Psychological Association.  

Lubiewska, K., & Głogowska, K. (2018). Zastosowanie analizy równoważności pomiarowej w badaniach psychologicznych. Polskie Forum Psychologiczne, 23(2), 330–356, doi: 10.14656/PFP20180207 

Magnano, P., Faraci, P., Santisi, G., Zammitti, A., Zarbo, R., & Howard, M.C. (2022). Psychometric Investigation of the Workplace Social Courage Scale (WSCS): New Evidence for Measurement Invariance and IRT Analysis. Behavioral Sciences, 12(5), 119, doi: 10.3390/bs12050119 

Merkin, R.S. (2006). Uncertainty avoidance and facework: A test of the Hofstede model. International Journal of Intercultural Relations, 30(2), 213–228, doi: 10.1016/j.ijintrel.2005.08.001 

Merkin, R.S. (2017). Saving face in business: Managing cross-cultural interactions. Berlin: Springer. 

Modrack, S. (2008). The Protestant Work Ethic revisited: A promising concept or an outdated idea? Wissenschaftszentrum Berlin für Sozialforschung (WZB). Social Science Research Center Berlin.  

Norton, P.J., & Weiss, B.J. (2009). The role of courage on behavioral approach in a fear-eliciting situation: A proof-of-concept pilot study. Journal of Anxiety Disorders, 23(2), 212–217, doi: 10.1016/j.janxdis.2008.07.002 

Ogińska-Bulik, N., & Juczyński, Z. (2008). Skala pomiaru prężności – SPP-25. Nowiny Psychologiczne, 3, 39–56.  

Pajestka, G. (2023). Facing Fear with Courage: Psychometric and Behavioral Evidence of the Courage Measure (CM) in Poland. Roczniki Psychologiczne, 26(3), doi: 10.18290/rpsych2023.0015 

Peterson, C., & Seligman, M.E. (2004). Character strengths and virtues: A handbook and classification. Vol. 1. New York: Oxford University Press. 

Pury, C.L.S., Kowalski, R.M., & Spearman, J. (2007). Distinctions between general and personal courage. The Journal of Positive Psychology, 2(2), 99–114, doi: 10.1080/17439760701237962 

Pury, C.L.S., & Starkey, C.B. (2010). Is courage an accolade or a process? A fundamental question for courage research. In C.L.S. Pury, & S.J. Lopez (Eds.), The psychology of courage: Modern research on an ancient virtue (pp. 67–87). Washington, DC: American Psychological Association, doi: 10.1037/12168-004 

Putman, D. (1997). Psychological courage. Philosophy, Psychiatry, & Psychology, 4(1), 1–11, doi: 10.1353/ppp.1997.0008 

Putnick, D.L., & Bornstein, M.H. (2016). Measurement invariance conventions and reporting: The state of the art and future directions for psychological research. Developmental Review, 41, 71–90, doi: 10.1016/j.dr.2016.06.004 

Rate, C.R., Clarke, J.A., Lindsay, D.R., & Sternberg, R.J. (2007). Implicit theories of courage. The Journal of Positive Psychology, 2(2), 80–98, doi: 10.1080/17439760701228755 

Soper, D.S. (2022). A-priori Sample Size Calculator for Structural Equation Models [Software]. Available from https://www.danielsoper.com/statcalc 

Strus, W., Cieciuch, J., & Rowiński, T. (2014). Polska adaptacja kwestionariusza IPIP-BFM-50 do pomiaru pięciu cech osobowości w ujęciu leksykalnym. Roczniki Psychologiczne, 17(2), 327–346. 

Szalkowska, A., Żemojtel-Piotrowska, M., & Clinton, A. (2015). Entitlement and organizational behaviors: The moderating role of narcissism. Current Issues in Personality Psychology, 3(4), 230–241, doi: 10.5114/cipp.2015.54437 

Szmajke, A., Bąk, P., & Adamus, E. (2004). Relikty kultury honoru w mentalności Polaków. Czasopismo Psychologiczne, 10, 7–21. 

Topolewska, E., Skimina, E., Strus, W., Cieciuch, J., & Rowiński, T. (2014). The short IPIP-BFM-20 questionnaire for measuring the Big Five. Roczniki Psychologiczne, 17(2), 385–402. 

Weber, M. (2011). Etyka protestancka i duch kapitalizmu. Warszawa: Wydawnictwa Uniwersytetu Warszawskiego. 

Williams, M.N., Grajales, C.A.G., & Kurkiewicz, D. (2013). Assumptions of multiple regression: Correcting two misconceptions. Practical Assessment, Research, and Evaluation, 18(1), 11, doi: 10.7275/55hn-wk47 

Woodard, C.R. (2004). Hardiness and the Concept of Courage. Consulting Psychology Journal: Practice and Research, 56(3), 173–185, doi: 10.1037/1065-9293.56.3.173 

Woodard, C.R., & Pury, C.L.S. (2007). The construct of courage: Categorization and measurement. Consulting Psychology Journal: Practice and Research, 59(2), 135–147, doi: 10.1037/1065-9293.59.2.135 

Additional materials 
Characteristics of the respondents