GENDER DIFFERENCES REGARDING WORKAHOLISM AND WORK-RELATED VARIABLES

strategies. Women were characterized by a stronger tendency towards self-justification then men. Males, on the other hand, declared stronger emotional resilience than women. These patterns of results are consistent with the results obtained in a previous study regarding gender differences in using self-handicapping strategies (Doliński and Szmajke 1994). The obtained results can be interpreted through the prism of the roles and tasks currently given to women. On the one hand, in light of social norms, a woman should take care of the household and family, while a man is responsible for earning money to support the family (Blair-Loy 2003). On the other hand, participation of women in the workforce is increasing (Peeters, Montgomery, Bakker and Schaufeli 2005), so they may find it difficult to reconcile work and fulfilling the demands of the roles of spouses, mothers or caregivers. Our study show that women may feel more overloaded with work and they have a higher level of emotional factors than men regarding workaholism. However, one may wonder whether women’s workaholism is still underestimated. Women may find it more difficult to admit that they feel an inner compulsion to work due to gender-differen tiated societal norms and expectations (Beiler-May et al. 2017: 109).


INTRODUCTION
The phenomenon of "workaholism" was first defined by Wayne Oates (1971) as an irresistible or uncontrolled need for continuous work. The definition of workaholism has since broadened. It includes work-oriented behavior and cognition (Snir and Harpaz 2004), and even "addiction to preoccupation" (Wojdyło 2003), which, as researchers point out, indicates that addiction may involve various goal-oriented activities, not necessarily related only to one's professional work (Wojdyło 2003). Some scholars treat workaholism as a psychological addiction (Golińska 2008;Killinger 2007;Robinson and Kelley 1998;Wojdyło 2013), similar to physical addictions such as alcoholism. For instance, Robinson (2014: 7) defined this construct as "an obsessive-compulsive disorder that manifests itself through self-imposed demands, an inability to regulate work habits, and an overindulgence in work to the exclusion of most other life activities." The multidimensionality of workaholism is also highlighted, although there is no agreement among researchers as to which aspects should be distinguished. For example, Ng, Sorensen and Feldman (2007) define workaholism to include the coexistence of three factors: affective, cognitive and behavioral. The definition of Spence and Robbins (1992) is also popular, in which they stipulate the existence of three aspects of workaholism (the so-called workaholic triad): work involvement, feeling driven to work, and work enjoyment.
Some researchers point to the positive -from the organization's point of view -effects of workaholism (Killinger 1991;Korn, Pratt and Lambrou 1987;Machlowitz 1980;Sprankle and Ebel 1987). For example, after interviewing a hundred workaholics, Machlowitz (1980) stated that they are productive and feel fulfilled. Other researchers, in turn, present workaholics as unhappy people who perform their duties compulsively but not necessarily effectively, being unable to cooperate and confrontational (Naughton 1987;Oates 1971;Porter 1996). The negative effects of workaholism in one's professional life include frequent absences from work due to stress, anxiety and depression (Worrall, Cooper and Campbell 2000); decrease in work efficiency (Garson 2005;Vodanovich and Piotrowski 2006); fatigue (Rosa 1995), burnout (Barnett, Gareis and Brennan 1999); the inability to delegate tasks (Spence and Robbins 1992); setting unrealistic goals (Porter 2001); and reduced creativity (Fassel 1990). It was found that workaholics at managerial levels tend to constantly control their subordinates (Graves, Ruderman and Ohlott 2006). Furthermore, workaholics with a high level of work compulsion are more likely to engage in inappropriate activities towards colleagues, such as public shaming (Galperin and Burke 2006). Workaholism has consequences for one's functioning in the non-professional area as well. These include cognitive and emotional exhaustion (Taris, Schaufeli and Verhoeven 2005), low levels of life satisfaction (Burke, Burgess and Fallon 2006), and financial issues (Hanson 1985). Workaholism also affects one's partner relationships: as indicated by Reeves (2005), the divorce rate among workaholics is about 40% higher than in the case of non-workaholics. Moreover, workaholics' spouses feel ignored, unloved, and emotionally or physically abandoned (Porter 2001). Harpaz and Snir (2003) showed that the major predictor of total paid work time per week is one's gender. According to researchers, workaholism (diagnosed based on the number of hours worked per week) is more widespread in men than women. Similarly, Burke, Davis and Flett (2008: 31) indicate that workaholics are mostly considered to be men. Doerfler and Kammer (1986) studied the connection between workaholism and gender as well as the accepted gender role (male, female, and androgynous) in doctors, advocates and psychologists. They showed that 23% of the respondents were workaholics, and the main effect of sex and professional group was not statistically significant. It is noteworthy, however, that women workaholics assumed a male or androgynous sex role exclusively (for comparison, 20% of female non-workaholic women assumed the female role). Results of other research are non-conclusive. Research by Spence and Robbins (1992) carried out on a sample of academic staff concerned the workaholic triad (work involvement, feeling driven to work, and work enjoyment). It has shown that women show higher levels of feeling driven to work and work enjoyment than men. Moreover, women felt higher levels of work-related stress and reported more health issues. In turn, differences between sexes were not reflected in work involvement and the behavioral correlates of workaholism: perfectionism and difficulties with delegating tasks. Burke (1999) also studied gender differences in the area of the workaholic triad. The results of his research did not show statistically significant differences between men and women in terms of work involvement, feeling driven to work and work enjoyment. However, there were differences in other work-related variables: women experienced work-related stress more often than men and showed a higher level of perfectionism. These results were partially replicated by Burgess, Burke and Oberklaid (2006) in studies carried out on a trial group of psychologists, where men felt a higher level of work involvement and feeling driven to work compared to women. Women, however -as in Burke's study (1999) showed a higher level of perfectionism and experienced higher levels of work-related stress, as well as more somatic symptoms and difficulties in delegating tasks. The situational and demographic variables were similar in both studies: women were younger, worked less and had lower earnings than men ). More recent studies conducted among academic staff showed significant gender differences in the workaholic triad: women felt a higher level of work involvement and feeling driven to work. However, they did not differ from men in terms of work enjoyment (Retna, Smith and Davies 2017).

DIFFERENCES BETWEEN THE SEXES IN WORKAHOLISM
Other self-reported studies on workaholism did not confirm the existence of gender differences (Aziz, Cunningham 2008;Bakker, Demerouti and Burke 2009). Some scholars tend to conclude that if there is a connection between gender and workaholism, it may be a weak one at best Russo and Waters 2006). On the other hand, Beiler-May et al. (2017) have proven that the differences between men and women in self-reported studies on workaholism may result from the way of responding to test items in questionnaires measuring workaholism. Analyzing the differences in the functioning of test items, these researchers showed that women would less often admit to devoting more time to work than other activities and to working while others are no longer at work. Beiler-May et al. (2017) indicate that workaholism in women is underestimated due to cultural norms. According to traditional expectations regarding the gender roles of men and women, men are to work and provide financial support for the family, while women are to perform most household duties and look after children. These diverse expectations may lead to a more complex experience of workaholism in workaholic women, because they must find a way to reconcile family responsibilities with their internal compulsion to work (Beiler-May et al. 2017). Results showing that workaholism is higher in women than in men were also obtained by Wojdyło (Wojdyło 2005, as cited in: Lewandowska-Walter and Wojdyło 2011; Wojdyło and Lewandowska-Walter 2009) and Behson (2002).
The presented ambiguity of research results may be due to several factors, including selection of the study participants. In most studies, the study groups were homogeneous and included for example people working in managerial positions (Aziz andCunningham 2008, Burke 1999) or academic staff (Retna et al. 2017;Spence and Robbins 1992), or they were confined to highly educated people (Doerfler and Kammer 1986) or even representatives of one profession, such as psychologists . The ambiguity of the results may also reflect cultural differences. After all, the described studies were conducted in different cultures: The United States (Aziz and Cunningham 2008;Behson 2002;Doerfler and Kammer 1986), Canada (Burke 1999 A significant research trend may be the study of gender differences in personality traits that may contribute to the development of workaholism (Retna et al. 2017). Previous research results have shown that certain personality traits may play a role in the development of addictions (Scott, Moore and Miceli 1997;Ng et al. 2007). It seems that such traits as perfectionism, conscientiousness, obsessive-compulsive personality or the drive for achievements may affect the development of workaholism through the predisposition of individuals to addictive behaviors. PERFECTIONISM One concept of perfectionism is Hamachek's (1978) proposal that perfectionism consists of a positive (adaptive, healthy) component and a negative (non-adaptive, neurotic) one. The former is associated with the pursuit of high albeit reasonable standards, the fulfillment of which leads to satisfaction and complacency. The latter includes tendencies towards setting unrealistic standards, and the fear of failure and disappointing others. This strongly and positively correlates with neuroticism and a tendency towards self-handicapping (Szczucka 2010). The maladaptive aspect of perfectionism, unlike in its adaptive aspect, is associated with experiencing higher levels of stress (Dunkley et al. 2003), and on the plane of professional functioning, with higher work overload (Ozbilir, Day and Catano 2015). Perfectionism is mentioned as a trait predisposing for addiction; it is for example a predictor of Internet addiction (Şenormancı et al. 2014). It is also associated with addiction to physical exercise (Lichtenstein et al. 2014). In addition, maladaptive perfectionists more often use alcohol to cope with stress compared to adaptive perfectionists and nonperfectionists (Rice and Van Ars dale 2010). A few studies have so far investigated the link between perfectionism and workaholism (Slaney et al. 2001;Spence and Robbins 1992;Stoeber, Davis and Townley 2013), showing a positive correlation between the two constructs. Szpitalak (2012) recognizes destructive perfectionism as one of the components of workaholism.

SELF-HANDICAPPING
Self-handicapping is defined as actions that allow an individual to credibly justify a possible failure, or -where said individual succeeds -to intensify one's triumph (Sierota 2003). Self-handicapping strategies can hinder success, but in the event of failure, they help to protect one's self-esteem (Sierota 2003). Dolinski and Szmajke (1994) distinguish three strategies of self-handicapping: behavioral strategies (engaging in activities that make it difficult to achieve success, e.g. drinking alcohol before an exam), demonstrative strategies (feeling and demonstrating one's own weaknesses, e.g. anxiety) and symbolic strategies (negative perception of the task situation, e.g. perceiving a task as more difficult than it really is) (Sierota, 2003). The results of a number of studies show that men use self-handicapping strategies more often than women, but this only applies to behavioral strategies (Berglas and Jones 1978;Harris and Snyder 1986;Kimble, Kimble and Croy 1998;Rhodewalt and Davison 1986). The results of Sierota's research (2003) show that the more women exhibit characteristics typical of the feminine stereotype (e.g. low need for achievement, low appreciation for values related to competences such as ability, ambition, or recognition of social values), the more inclined they are to use self-handicapping strategies. However, the more men display features incompatible with stereotypes of male behavior (e.g. poor pursuit of success or lack of assertiveness), the stronger this tendency is. The results of previous research on self-handicapping in the work context indicate that the use of self-handicapping strategies in social situations is related to burnout (Salmela-Aro, Tolvanen and Nurmi 2011).

HYPOTHESES
The aim of the study was to analyze gender differences in workaholism, perfectionism and self-handicapping strategies.
1) It was assumed that women would have a significantly higher level of workaholism than men. This hypothesis was based on the results of the latest research, which indicate workaholism in women is higher than in men, and it can also be underestimated in selfreport studies (Beiler-May et al. 2017). Considering the scarcity of research on gender differences in workaholism, the study of differences in individual factors constituting workaholism is exploratory in nature. Therefore, no hypotheses were made regarding gender differences in workaholism in its individual aspects.
2) When it comes to perfectionism, it was assumed that women would have a higher level of maladaptive perfectionism than men. This hypothesis was formulated based on the results of previous research, indicating that women have a higher level of negative perfectionism (which may result in lower levels of satisfaction and well-being) than men (Burke 1999;Burgess et al. 2006).
3) A mediation model was also tested, assuming that the connection between gender and workaholism is mediated through maladaptive perfectionism. It was assumed that like other addictions related to perfectionism, maladaptive perfectionism would be associated with workaholism and be a personality variable mediator between gender and workaholism.
4) Regarding self-handicapping strategies, several hypotheses were put forward. First of all, it was assumed that there would be a positive correlation between self-handicapping and workaholism (4.1). One of the personality determinants of workaholism is the fear of failure (Wojdyło 2004), and using self-handicapping strategies is aimed at protecting one's self-esteem from failure (Sierota 2003), so workaholics are more prone to use self-handicapping strategies than non-workaholics. Secondly, it was assumed that women will use self-handicapping strategies more often than men (4.2). This hypothesis was formulated based on the results of research conducted by Doliński and Szmajke (1994), where women showed a significantly higher tendency towards self-handicapping compared to men. It was also assumed that gender differences would appear in the individual factors of the self-handicapping scale. It was assumed that women would achieve a higher score than men on the Self-justification (4.3) and Discipline/ Mobilization (4.4) scales. This hypothesis was based on research results which found that women were more willing to seek excuses and indicate "internal" and "external" circumstances than men, diminishing the impact of women on the results of activities and declared mobilization because of fear of failure in task-related situations (Freeman 2004, Rhodewalt 1990) It was also assumed that men would have a higher score on the Emotional Resilience scale compared to women (4.5). This hypothesis was formulated based on the results of research conducted by Doliński and Szmajke (1994), where men were more convinced about the independence of the results of their actions from their emotional state than women, and the emotional state became independent from external circumstances.

PARTICIPANTS
The study participants were 314 people, 222 women and 92 men, aged from 17 to 80 (M = 29.29, SD = 12.02). They were working students, employees of Krakow-based companies of various sizes, and people running their own businesses. The mean seniority of the participants was 15.75 years (SD = 15.51). Participation in the study was voluntary and its participants did not receive any remuneration.

MATERIALS
Workaholism was measured using the WART work addiction questionnaire (Work Addiction Risk Test, Robinson, Phillips 1995) in its Polish adaptation by Wojdyło (2005). This tool measures the symptoms of a workaholic behavior pattern, which include five basic factors: (1) Obsession/Compulsion, (2) Emotional Arousal/Perfectionism, (3) Overdoing, (4) Outcome Orientation, and (5) Self-Worth. The questionnaire contains 25 statements, to which respondents respond on a four-point scale. It has satisfactory psychometric properties, comparable to the original version. The reliability of the Polish version of the scale is 0.87, as estimated using Cronbach's alpha method.
Perfectionism was measured using the KPAD Adaptative and Maladaptive Perfectionism Questionnaire (Szczucka 2010), which includes two subscales: Adaptive Perfectionism and Maladaptive Perfectionism. The tool is used to measure the task-oriented functioning of both types of perfectionists. The questionnaire contains 35 statements, to which respondents respond on a five-point scale. The internal consistency of the tool, estimated using Cronbach's alpha coefficient, is 0.82 for the Adaptive Perfectionism scale and 0.95 for the Maladaptive Perfectionism scale. The scale is accurate and has good discriminatory power (Szczucka 2010).
Self-handicapping strategies were assessed using the Anticipatory Self-esteem Enhancement Strategy Scale (ASO, Jones and Rhodewalt 1982) in its Polish adaptation by Szmajke and Świątnicki (Doliński and Szmajke 1994). The questionnaire includes three subscales: Self-Justification, Discipline/Mobilization, and Emotional Resilience. It consists of 25 statements referring to situations threatening one's self-assessment, to which respondents reply on a six-point scale. The tool has good internal consistency (Cronbach's alpha coefficient of 0.80) and validity.

PROCEDURE
The participants filled out the questionnaires on paper in pencil, individually or in groups. Table 1 shows the correlation values between the measured variables.  There were significant positive correlations between workaholism and perfectionism, both maladaptive (r(312) = .58, p < .001), and adaptive (r(312) = .36, p < .001). Workaholism was also statistically significantly correlated with the tendency towards self-handicapping (r(312) = .51, p < .001).
In order to test the hypothesis that gender affects workaholism, and that this effect is mediated by maladaptive perfectionism, mediation analysis was performed (model 4, Hayes 2018). Gender was a predicator in the model, maladaptive perfectionism was a mediator, and workaholism was the resultant variable.

ADDITIONAL ANALYSES
Additional analysis was carried out to determine whether gender moderates the relationship between maladaptive perfectionism and workaholism, and the relationship between adaptive perfectionism and workaholism. PROCESS Software 3.4 (Hayes 2018) was used to perform the analysis. The results are presented in Table 4. As can be seen, the relationship between maladaptive perfectionism and the Obsession/Compulsion subscale of workaholism was significantly influenced by gender. This interaction is illustrated in Figure 1. The interaction was probed by testing the conditional effects of maladaptive perfectionism for women and men. As shown in Table 4, maladaptive perfectionism was significantly related to the Obsession/Compulsion subscale of workaholism in the group of women (p < .001), but not in the men (p = .053).

DISCUSSION
As predicted, women showed a higher level of workaholism than men. This is an important result in the context of the inconsistency of the current research results on gender differences regarding the discussed characteristics. It is also important against the background of the few studies conducted so far in Poland. Significant differences were revealed in the Emotional Arousal/Perfectionism and Overdoing aspects, and these results are consistent with the results obtained by Wojdyło and Lewandowska-Walter (2010). Work overload may result from women's efforts to meet the challenges of "working two shifts" (Kosakowska and Petrus 2006), that is, being both an employee and a householder. This often results in working at irregular times, such as at weekends or after hours (Behson 2002). At the same time, as Peplińska, Wojdyło, Kosakowska-Berezecka and Połomski (2015) demonstrated, women encounter stereotypes regarding the traditional gender roles, and to be successful in their professional lives they need to prove that they fulfill roles socially assigned to men.
This confirmed the second hypothesis, stating that women exhibit a higher level of maladaptive perfectionism in comparison to men, as well as the third hypothesis, regarding the mediating influence of maladaptive perfectionism on the connection between sex and workaholism. In the context of Internet addiction, Şenormancı et al. (2014) suggest that since the discrepancy between one's actual Self and one's ideal Self is conducive to the development of maladaptive perfectionism, the possibility of creating an ideal avatar on the Internet may lead people to develop Internet addictions. Perhaps a similar mechanism is at work in the case of workaholism. People who have a large discrepancy between their real Self and their ideal Self can perceive work as an opportunity to "demonstrate" and improve their self-esteem, e.g. through long working hours, which makes them workaholics. Perfectionism may explain gender differences in workaholism. Women trying to meet these requirements regarding traditional gender roles and at the same time wanting to develop professionally are setting unrealistic requirements and high standards for themselves, which forces them to work excessively and, in consequence, may lead to work addiction. Additional moderation analysis has shown that maladaptive perfectionism is associated with the Obsessions/Compulsion dimension of workaholism, but such a relationship exists in the group of women but not of men. One possible explanation for this pattern of results may be that the WART subscale of Obsessions/Compulsions most closely reflects the definition of workaholism as harmful coercion into work (Beiler-May et al. 2017). Beyond that, in the research this subscale was found to be a short, valid measure of workaholism (Taris et al. 2005). Another possible explanation may be related to the sample size and variance of the dependent variable. The groups were unequal and in the group of men the variance of results in the Obsession/Compulsions subscale was significantly lower than in the group of women. These factors reduce the test power, meaning it is less likely to find a significant result.
The hypothesis concerning the connection between workaholism and self-handicapping tendency was confirmed: these variables had a strong positive correlation, which is in line with the concept of personality determinants of workaholism according to Wojdyło (2007), where the intensity of work obsession depends on one's fear of failure (the higher the level of fear, the higher the level of workaholism). The results regarding gender differences were in line with the assumptions. Women had a stronger tendency to use self-handicapping strategies compared to men. The resulting pattern was consistent with the results obtained in the studies by Doliński and Szmajke (1994) -women showed a higher tendency to justify themselves and higher levels of discipline and mobilization, whereas men were characterized by significantly higher emotional strength than women.
Thus far, relatively few empirical studies have addressed the issue of gender differences in workaholism and the variables related to it, so it is difficult to draw significant conclusions based on the collected data . This study provides additional knowledge about the differences in workaholism, perfectionism and the tendency to self-handicap in women and men. It covers a sample that is diverse in terms of seniority and occupation, so that its results can be generalized to a wider population than those of some previous studies (e.g. Aziz and Cunningham 2008;Retna et al. 2017), but it has its limitations. First of all, the study did not take into account the adjustment of women and men in terms of demographic variables, such as age, education level, seniority, or having children and their number. Secondly, like all studies based on indirect (self-report) methods, this study might be of limited relevance due to imperfect self-description skills of participants (lack of sufficient self-knowledge), the tendency to choose middle responses, or their need for social approval (Craik 2007;Paulhus and Vazire 2007). In the future, it would also be worthwhile to apply behavioural workaholism indicators, such as overtime, weekend work, or limiting time for other activities outside work (Malinowska, Staszczyk and Tokarz 2015), which can be seen and objectively evaluated by co-workers and family members of the participants. Thirdly, workaholism was diagnosed using the WART questionnaire (Robinson and Phillips 1995;Polish adaptation: Wojdyło 2005), which is based on the concept of workaholism as a disorder involving compulsion to work and work overload. It would be worth investigating gender differences in workaholism using other methods of measuring this characteristic, derived from different concepts of workaholism and gender differences regarding enthusiasm for work (Schaufeli et al. 2002).