Personality and Hardiness Among Police Students

1. Introduction

In many lines of work, personality and individual characteristics are important determinants that underly the ability to carry out duties, for instance, how personality traits, attitudes, or intellectual ability influence assessments of situations or how tasks or challenges are managed (Furnham, 2008). The police represent such an occupation, where different assignments require different skillsets or competencies tied to individual differences. The scope of the current programme of research was to conduct a pilot study and examine the rudimentary psychometric properties of two combined measures of individual differences, namely personality and hardiness. The study aimed to evaluate the usefulness of these measures for a future large-scale survey that will investigate police studentsʼ personality and hardiness profiles from study admission to completion.

Research has provided us with knowledge of understanding, assessing, and describing various individual phenomena, such as personality: our unique ways of thinking, feeling, and behaving that are relatively stable across time and situations (John et al., 2010). One of the most widely used personality measures is the Big Five taxonomy, which considers personality through five broad factors. Each factor represents a continuum or personality dimension that all people have to a greater or lesser extent. The factors are as follows: Extraversion (e.g., from being reserved to assertive), Agreeableness (e.g., from being unfriendly to kind); Conscientiousness (e.g., from being careless to well organised); Neuroticism (e.g., from being emotionally stable to anxious); and Openness to experience (e.g., from having narrow interests to being original) (Digman, 1990; John & Srivastava, 1999). The well-established Big Five Inventory (BFI) is a 44-item self-report questionnaire developed to capture the prototypicality or core elements of these five factors by respondents indicating how well (or not) they identify themselves with individual personality-related characteristics (John et al., 2008; John & Srivastava, 1999). The current study used a short version of the BFI-44, the BFI-20.

When developing this BFI short-format scale, Engvik and Clausen (2011) selected 20 items from the BFI-44 based on their psychometric properties and an underlying intent to minimalise loss of psychometric quality (through measures of structural validity, maximal representation, maximal stability over time, and maximal criterion validity). The scale measures the following five main factors: Extraversion (active, talkative, dominating, outgoing); Friendliness (warm, non-cynical, caring, friendly); Control (well organised, conscientious, structured, disciplined); Emotional stability (careless, non-sensitive, not nervous, not tense); and Fantasy (original, creative, imaginative, inventive). Even though the BFI-20 contains a spectrum of properties for each of the five factors, it needed to be evaluated by considering what is useful and what is psychometrically acceptable. In this regard, the BFI-20 has shown acceptable values for all factors regarding internal consistency, test-retest reliability, representativeness of the factors as measured by longer measures, and covariations toward factors rated by others (Engvik & Clausen, 2011).

Using 20 items to assess a complex phenomenon like personality is far from ideal. That said, the five factors of the BFI-20 have all shown adequate psychometric properties. The scale measure can be considered useful and convenient in contexts that do not require an optimal personality measurement, such as in large-scale research (which this study is a pilot for). For instance, the BFI-20 has been utilised in large-scale research investigating Big Five personality dimensions among musicians (Vaag et al., 2018), the association between personality traits and consumer behaviour (Ardebili & Rickertsen, 2020), and the relationship between personality and quality of life (Nes et al., 2018).

The second measure employed in the study concerned psychological hardiness (Bartone, 1995; Kobasa, 1979), a personality-related factor believed to function as a buffer between external stressors and physical or psychological effects (Hystad & Johnsen, 2019). The concept can be related to constructs like mental toughness (Ward et al., 2018) or resilience (Janssens et al., 2021). Hardiness is a coping style or a way of operating involving cognitive, emotional, and behavioural processes. Three factors or dimensions constitute the concept. These are (i) Commitment – referring to being confident in oneʼs abilities and engaging in oneʼs life, tasks and relationships, and perceiving these as meaningful; (ii) Control – which concerns the individualʼs belief in their ability to influence outcomes of different situations, and; (iii) Challenge – the extent to which the individual is pro-active and perceives changes and unpredictability as challenges with opportunities to develop from experience (Eid et al., 2008; Eschleman et al., 2010). The three components are interconnected attitudes that influence how individuals experience and perceive potentially stressful events and can assist the person in reducing the negative effects of stressors. Hardiness contributes to stress management through two main aspects: (i) the individualʼs cognitive appraisal of stressors and (ii) their behavioural response. Hardy individuals engage proactively in tasks, appreciate challenges, and have confidence in their ability in ways that lead to effective stress management and personal growth (Hystad & Johnsen, 2019; Maddi, 2002). To assess hardiness, Bartone (1995) established the 30-item Dispositional Resilience Scale, later developed into a 15-item version named The Short-Hardiness Scale or DRS-15-R (Hystad et al., 2010; Johnsen et al., 2004), which we utilised in the current study. Even though police education in Norway does not explicitly target hardiness, students are trained in different ways of coping with the potential stress of working as police officers, for example, with regard to issues such as work overload or physical threats in operational work (Berg et al., 2006). Considering the nature of police work, it is reasonable to assume that both personality and hardiness are influential factors in how police officers carry out their duties. Thus, the scope of the current study was to obtain data on the BFI-20 and DRS-15-R to establish their psychometric properties and evaluate whether they satisfactorily detect individual differences among police students.

2. Method

3. Results

4. Discussion

This pilot study aimed to evaluate the suitability of the BFI-20 and DRS-15-R for upcoming police student research. Being short inventories, the usability of researching with these measures has apparent advantages. However, for adequate measurement of personality and hardiness, these instruments should be sensitive enough for detecting individual differences by showing variability in data. The findings from the present study demonstrated firstly that most respondents generally reported high scores, thus resulting in somewhat negatively-skewed distributions. A visual inspection and a skewness value close to zero revealed that scores on Fantasy were quite normally distributed. Our expectations of skewed scores on hardiness and Emotional stability were therefore met. However, the results did not demonstrate the expected pattern of more salient variability on the Extraversion, Friendliness, and Control traits. A possible explanation for this is that ranges restrictions are most likely due to characteristics of our sample, where all participants had been through a selection process emphasising certain similar characteristics (Skoglund, 2018).

The Cronbachʼs alpha values were suboptimal, except for the Extraversion scale (see Table 3). Therefore, caution is warranted in interpreting scores from the other scales. Lower than optimal alpha values are not necessarily indicative of suboptimal measures per se, as the magnitude of the coefficient alpha is a characteristic of the obtained scores of the measure (Streiner, 2003). The suboptimal internal consistency in the scores obtained can perhaps be explained by our sampleʼs marginal number of participants. However, there are of course challenges related to the content domain in such short measures as the ones used in the present study. Nevertheless, we should have larger samples for investigating the internal reliability of scores from police students.

The rudimentary item analyses generally supported that the measures could identify individual differences among police students, as ranges in responses are detectable. However, there was a challenge with the BFI-20 item “Am considerate towards other people”, as no one had marked a score below five on the 7-point scale. It is, of course, possible that this is a valid response for our sample. Another explanation, however, is that such an item is especially sensitive to a socially desirable response.

The predominantly modest correlations of hardiness and personality traits align with earlier findings (Kardum et al., 2012). Hardiness and Emotional stability had a large covariation, which was expected and in line with other research (Eschleman et al., 2010; Oshio et al., 2018). With the correlation of r = .51, approximately 25% of the variance is shared between the two variables. After controlling for the effects of age and gender, the personality traits added explanatory value for the total hardiness score. Among the personality traits, Emotional stability was a strong isolated predictor of the total hardiness score, further supporting the relatedness of these constructs. Our expectations of a higher explanatory power of the personality traits over demograhics were therefore met. While being related to the personality traits, especially Emotional stability, the main result of modest covariations supports the notion that Hardiness seems to be a separate construct relative to personality (Hystad et al., 2015; Kardum et al., 2012). And, indeed, since about half of the variance between Emotional stability and hardiness was not shared, these constructs also can be seen as different constructs, albeit being related.

When considering the studyʼs limitations, first and foremost, sampling bias is an unknown factor as the response rate was quite low. The present studyʼs marginal response rate of 23.35% was probably not due to the measures but rather to the timing of the data collection, which collided with several academic examinations at the end of the semester. With a low response rate, problems related to self-selection bias might come into play. Students with certain personality traits might be more inclined to participate in a study, especially at a time of high academic workload. The distribution of personality traits we found might, therefore, not fully represent the whole police student population. Further data collection is needed to verify or adjust the patterns of findings obtained in this pilot study. Secondly, as the measures used in the study are short and based on self-report, care in score interpretation is warranted. Self-report on personality measures may contribute to socially-desirable responses, thus decreasing the reliability and validity of scores. This could, however, be more actualised in selection settings (Morgeson et al., 2007a, 2007b). Considering the limited length of the measures, this is especially relevant for the BFI-20. Compared to more comprehensive Big Five tests, such as the NEO-PI-R/3, with its 240 items resulting in facet scores in addition to factor scores, it becomes clear that nuances in personality factors are lost with short-form measures. Usability benefits are an obvious advantage of short measures, but it is also of interest if short-form inventories can provide reasonable answers to personality research questions. For the DRS-15-R, the particularly low Cronbachʼs alpha found for the challenge dimension (.46) could limit the credibility of this specific dimension. However, it is not uncommon to find that the challenge scale has a notably lower reliability estimate than the other two dimensions (e.g., Hystad et al., 2010; Sandvik et al., 2020).

To our knowledge, this is the first study to combine the BFI-20 and DRS-15-R. Overall, we consider the measures feasible and suitable for the mentioned upcoming research among police students. These measures are not, however, optimal for nuanced personality or resilience descriptions at the individual level. Still, we see them as useful for quantitative research purposes in which time and recruitment of participants can be a challenge. In our marginal sample, the measures demonstrated somewhat promising psychometric properties, and our findings also replicated some basic findings from the Big Five and hardiness literature. Thus, we conclude that useful knowledge of individual differences can be obtained with these instruments, justifying continued research using the measures on a police student population.