7Herbst.indd The traditional roles of women in society were those of wives and mothers, with the childcare and domestic responsibilities ascribed to them (Biernat & Wortman, 1991). It is only in recent history that women became a sizeable segment of the work force and made inroads in previously all-male occupations (Senior, 2003). Statistics South Africa (2004) indicated that 1,7 million women entered the labour market between 1991 and 2001, pointing to a 23,6% increase in women entering the labour market during this period. These statistics confirm that more women pursue careers nowadays and are required to contribute to the income of the household (Acker, 1992; Brannen & Moss, 1991; Shaw & Burns, 1993; Spade, 1994), which results in families operating as dual career couples (Klerman & Leibowitz, 1994; Schwartz, 1989). However, in dual career couples women still perform most of the domestic chores and remain responsible for child care and domestic responsibilities (Douglas & Michaels, 2004; Maconachie, 1990; Ozer, 1995). Being a working mother, does have some implications in terms of overall well-being (Sears & Galambos, 1993; Vosloo, 2000). Major difficulties for career women are the perceived incompatibility of their roles as wife and mother with their roles of being employed (Shipley & Coats, 1992; Vosloo, 2000). Coping with the various roles usually contribute to increased levels of stress and difficulty for working mothers (Chusmir & Durand, 1987; Sears & Galambos, 1993; Vosloo, 2000). On the other hand, it also has some implications for the organisation with regards to its productivity and effectiveness, as most working mothers entering the workplace are still responsible for the more traditional domestic and child care responsibilities (Douglas & Michaels, 2004; Senior, 2003). Nevertheless, Welbourne (1999) indicated that having women in top management teams resulted in higher earnings, greater shareholder wealth and better long-term performance. Ruderman, Ohlott, Panzer and King’s (2002) research also suggested that the multiple roles women play in their personal lives provide practice for multi- skilling, opportunities to interpersonal skills and leadership practice that in the end enhance effectiveness in their working role. Therefore, most organisations nowadays appear to realise the increased importance of utilising, developing and retaining female talent (Senior, 2003). In a South African study, Vosloo (2000) observed that working women do not function in isolation, but in a bigger system that encloses the organisation and their work. But they also function in an individual system that consists of their families and demographic variables (Vosloo, 2000). In terms of the interaction of these two systems and the conflict that may arise from this interaction, it is necessary that mechanisms be found for handling these various roles effectively. Working women’s personality characteristics, sense of coherence and coping skills may be regarded as vital mechanisms that may influence the manner in which they deal with these various roles effectively. Personality One area for distinguishing different behavioural responses is to investigate personality dimensions/traits. Trait theories describe personality as a combination of traits – consistent attributes that characterise what a person is like (stable aspects of personality) (Saville & Holdsworth, 1999a). It is believed that all individuals have essentially the same traits and that they differ only in terms of the extent to which they manifest each trait (Sternberg, 1994). The trait theories provide a rationale for individuals’ different responses to stress in their environments, promote an understanding of sources of what proves stressful for people, and more importantly, assist in identif ying potential strategies for coping more effectively (Saville & Holdsworth, 1999a). The Occupational Personality Questionnaire (OPQ) model of personality (Saville & Holdsworth, 1999a) is based on a definition which postulates that personality deals with individuals’ typical or preferred ways of behaving, thinking and feeling (the conative, affective and cognitive dimensions of personality). It is proposed that personality is concerned with three main domains namely the relating, thinking and feeling domains. These domains are potentially joined by a fourth domain, the dynamism domain, characterised by traits such as vigour, competitiveness and decisiveness (Saville & Holdsworth, 1999a). It is postulated that personality traits, as measured by the OPQ, can contribute to a better understanding of how working adults (in this case mothers) differ in their behavioural responses to their environments (Saville & Holdsworth, 1999a). Personality traits may therefore be variables that influence the degree of stress likely to be experienced by working mothers in a given situation. Sense of coherence Antonovsky (1979) proposed that the human condition is stressful and rejected the assumption that stressors are inherently bad. He developed the salutogenic (origin of health) paradigm and its core construct, namely sense of coherence. The salutogenic LETTIE HERBST Lettie.Herbst@ExecuJet.co.za SANET COETzEE* coetzsc@unisa.ac.za DELéNE VISSER Department of Industrial and Organisational Psychology University of South Africa ABSTRACT The purpose of the study was to examine the relationships between personality dimensions, sense of coherence and coping styles among working mothers. The OPQ, OLQ and COPE questionnaires were administered to 120 married, working mothers. In view of contrasting results obtained by other researchers regarding the dimensionality of the COPE, its factorial validity and internal consistency were assessed. Exploratory factor analysis indicated that the COPE measured five factors with high degrees of internal consistency. Several personality dimensions and sense of coherence variables correlated significantly with the dependent variables (coping styles). It appeared that these independent variables predicted substantial percentages of the variability in the coping styles. Key words: Sense of coherence, coping, personality, working mothers PeRSonAliTy, SenSe of CoheRenCe And The CoPing of woRKing moTheRS 57 SA Journal of Industrial Psychology, 2007, 33 (3), 57-67 SA Tydskrif vir Bedryfsielkunde, 2007, 33 (3), 57-67 HERBST, COETzEE, VISSER58 paradigm focuses on individuals’ ability to be healthy and resilient under stressful conditions, as well as on optimal coping in everyday living. The focus point of this paradigm is on positive, optimal conditions of psychological health. The paradigm includes the strength that is exhibited in order to manage the tension and stress in their lives and not to fall ill. This approach views stress as a dynamic, manageable process which generates positive outcomes and effective coping and functioning (Antonovsky, 1979). Antonovsky (1979) defined sense of coherence as a global orientation that expresses the extent to which one has a pervasive, enduring though dynamic feeling of confidence that one’s internal and external environments are predictable and that there is a high probability that things will work out as well as can reasonably be expected. Sense of coherence is an internalised sense of control, which guides individuals’ orientation towards events. It consists of three core, interwoven dimensions, namely comprehensibility, manageabilit y and meaningfulness (Antonovsky, 1987). Comprehensibility is the extent to which individuals perceive stimuli that confront them as making cognitive sense, as information that is ordered, structured and clear, rather than chaotic, disordered, random, accidental and unpredictable (Antonovsky, 1984). Individuals find logic in the sequence of events, a degree of consistency from one experience to another, and belief that unexplainable events do not occur (Sullivan, 1993). Manageability can be defined as the extent to which people perceive that resources are at their disposal and are adequate to meet the demands imposed by stimuli (Antonovsky, 1984). Life does not become unmanageable and overwhelming, but rather endurable (Sullivan, 1993). Meaningfulness refers to the degree of commitment one has to various life domains and the importance of shaping not only one’s destiny, but also one’s daily experiences. High meaningfulness includes the feeling that life makes sense emotionally (Antonovsky, 1984). The concept sense of coherence implies that stressors are seen as positive and meaningful challenges which can be managed in such a way as to result in optimal outcomes. Individuals functioning with this view of life usually have resistance to the effects of stress and are more capable of experiencing stressful situations without feeling the negative effects thereof (Antonovsky, 1993). Strümpfer (1990) indicated that sense of coherence not only refers to how people stay healthy, but it also has an impact on how work is approached and performed. A strong sense of coherence would result in a person: l making cognitive sense of the workplace, perceiving stimuli as clear, ordered and consistent, l perceiving work as consisting of experiences that are bearable, with challenges that can be met by availing oneself with personal resources under the control of legitimate others, and l making emotional and motivational sense of work demands, as welcome challenges, worthy of engaging in and investing energy in. Carrim (2000) in her research on working women in their multiple role environment also concluded that there is a relationship between the strength of working women’s salutogenic construct scores (including sense of coherence) and their ability to cope with multiple roles. Antonovsky (1987) proposed that sense of coherence is not a coping style. Stressors in life are varied and there are many possible coping procedures. Consistently adopting only one pattern of coping is to fail to respond to the nature of the stressor and therefore to decrease the chances of successful coping. A person with a strong sense of coherence will select a coping strategy that appears most appropriate to dealing with a specific stressor. Coping and its antecedents Lazarus and Folkman (1984) defined coping as a constantly changing cognitive and behavioural effort aiming to manage specific internal and/or external demands that are appraised as taxing or exceeding the resources of the person. Coping is a dynamic process that changes over the course of a stressful transaction between people and their environment (Lazarus & Folkman, 1984). Carver, Scheier and Pozo (1992) defined coping as efforts to create conditions that permit individuals to continue moving toward desired goals or efforts to disengage from goals that are no longer seen as attainable. Coping is thus conceptualised as a transactional process or exchange between individuals and their environment with an emphasis on process. Coping styles are coping patterns that appear to be habitual, actions that are characteristic of the individual, with stability and consistency over time and conditions (Lazarus & Folkman, 1984). The transactional perspective of Lazarus (1966) defines two types of coping: problem-focused coping and emotion- focused coping. Problem-focused coping is coping behaviour aimed at solving the problem or doing something to change the source of stress, whereas emotion-focused coping is aimed at reducing or managing the emotional distress that is associated with a situation (Carver, Scheier & Weintraub, 1989). Carver et al. (1989) argued that each broad type of coping (emotional and problem focused coping) may involve several distinct strategies (e.g. denial, seeking social support), each possibly having very different implications for adjustment. The problem- and emotion-focused types of coping of Lazarus and Folkman (1984) were an important distinction, but this model was too simple according to Carver and Scheier (1989). Carver et al. (1989) subsequently distinguished between five types of problem focused coping, five types of emotion-focused coping, and three lesser used coping styles, namely focus on and venting of emotions, behavioural disengagement and mental disengagement. The questionnaire that was developed, based on this model, was used to measure coping in this study. People tend to use both emotion-focused and problem-focused coping in managing stressful events. This approach appears to be functional, for it allows for both the regulation of emotion and the management of the stressor (Lazarus & Folkman, 1984). To ensure personal coping effectiveness, it is necessary to incorporate relevant problem-solving and/or emotion-focused coping skills. Using different strategies in different combinations may be a better way to manage stress than responding reflexively with the same limited response to varying stressors. A number of studies concluded that having a versatile coping profile is associated with good adjustment (Mattlin, Wethington & Kessler, 1990; Pearlin & Schooler, 1978). This is however not true across the board. For instance, being able to call on more coping mechanisms in occupational stress does not minimise the chances that role strain will result in emotional stress (Pearlin & Schooler, 1978). The results of a given coping style are determined by the interaction of personal needs and preferences and the constraints of the current situation. Applying the same coping strategy across all situations is not likely to be adaptive (Collins, Taylor & Skokan, 1990). Specific coping strategies are more or less effective depending on the type of stressor encountered (Pearlin & Schooler, 1978). At the workplace the adoption of an effective coping style is very important, because this will have a definite impact on individual effectiveness, which in turn will have an impact on various organisational factors, such as productivity, organisational effectiveness, job satisfaction and turnover. Although coping can change from moment to moment, people do develop habitual ways of dealing with stress and these habits or coping styles can influence their reactions in new situations (Carver & Scheier, 1994). COPING OF WORKING MOTHERS 59 Individual differences affect coping in two ways: l Stable (dispositional) coping strategies may be used without any reference to stressful demands (Carver et al, 1989). People also use various coping responses (situational) based on the nature of events, the assessment of events, or their own resources (Lazarus & Folkman, 1984). l Personality dimensions/traits may also predict or correlate with certain coping behaviours (Eksi, 2004). Previous research found relatively modest links between coping, dispositions (like sense of coherence) and traditional personality variables. Eysenck (1983) reported a negative association between neuroticism and task-orientated coping efforts (problem-focused coping). McCrae and Costa (1986) indicated that dimensions of personality are associated with certain preferred modes of coping. Their study indicated that neuroticism was associated with the use of hostile reaction, escapist fantasy, self-blame, sedation, withdrawal, wishful thinking, passivity and indecisiveness. They also found that extraversion correlated with rational action, positive thinking, substitution and restraint. Similarly, Vollrath (2001) reported that there is a positive correlation between extraversion and conscientiousness on the one hand, and problem-focused coping, that is active, rational, planned coping strategies, on the other hand. As can be expected, seeking social support has been shown to be predominantly related to extraversion. As also found by McCrae and Costa (1986), neuroticism was linked to coping styles related to avoidance (Vollrath, 2001). More recently, Roesch, Christian and Vaughn (2006) found that neuroticism was also positively related to emotion-focused coping (specifically emotional support). Their study also confirmed that extraversion, conscientiousness and openness were positively related to problem-focused coping (active coping, planning) and emotion-focused coping (specifically positive reframing and acceptance). In addition, agreeableness was positively associated with active coping. In South Africa, Storm and Rothman (2003) reported that emotional stability (low neuroticism), extraversion, openness to experience, and conscientiousness are associated with constructive coping strategies, such as active problem solving, seeking social support (instrumental reasons), positive reinterpreting stressful situations and acceptance of stressors. Agreeableness and conscientiousness were associated with acceptance of stressors and turning to religion. On the other hand, passive coping strategies, such as focus on venting of emotions, denial and seeking social support (emotional reasons) were associated with neuroticism, low agreeableness and low conscientiousness. There is not a body of evidence that relates personality, as measured by the OPQ, to coping. However, in a study by Visser and Du Toit (2004) an exploratory factor analysis of the OPQ yielded a six-factor structure that included five factors corresponding to the Big Five model of personality and a sixth factor which they labelled Interpersonal Relationship Harmony. They found that the sub-scales of Emotional control, Modest, Outgoing and Persuasive of the OPQ loaded on extraversion. The sub-scales Relaxed, Worrying, Tough minded, Socially confident, Social desirability, Optimistic and Outgoing loaded on emotional stability (low neuroticism). Decisive, Democratic, Controlling and Independent loaded on agreeableness. The factor openness to experience included the sub-scales of Conceptual, Data rational, Artistic, Innovative and Critical, whereas conscientiousness included the OPQ sub-scales of Conscientious, Detail conscious, Forward planning, Social desirability, Achieving and Controlling. It can therefore be expected that personality, as measured by the OPQ, will also be associated with specific coping strategies measured by the COPE. Sense of coherence is not a coping style per se, but a person with a strong sense of coherence will be able to select an appropriate coping strategy for a specific stressor (Antonovsky, 1987). In a qualitative study, using a hermeneutic approach, Strang and Strang (2001) found that the manageability component of sense of coherence, for people diagnosed with a life-threatening disease, was achieved by information seeking strategies, social support and positive reinterpretation of the situation. Meaningfulness was created by close relations and faith. These results suggest, firstly, that a relationship between manageability and seeking information, social support and positive reinterpretation may be expected, and secondly that there is a relationship between meaningfulness and turning to religion. Some empirical evidence for the relationship was found by Rothmann and Van Rensburg (2002). They found a practically significant negative relationship of medium effect between sense of coherence on the one hand and behavioural disengagement and alcohol- drug disengagement on the other hand. Rothmann (2004) found practically significant positive relationships (of medium effect) between sense of coherence on the one hand and problem-focused coping and positive reinterpretation on the other hand. Furthermore, Redelinghuys and Rothmann (2005) found a practically significant (large effect) positive correlation between sense of coherence and active or approach coping, and a practically significant (medium effect) negative correlation between sense of coherence and venting of emotions. It is evident that personality and sense of coherence can act as determinants of the manner in which individuals experience and cope with stress. It is hypothesised that these variables will determine the style and manner in which working mothers cope with life’s demands and stresses. Working mothers require skill to be successful spouses, mothers and also productive workers, considering the fact that they are confronted with different ways of being and must decide which way of being-in-the-world they will respond to and to what intensity (Richardson, 1991; Vosloo, 2000). This is especially important for modern organisations that are constantly under pressure to make the best use of their human capital (Senior, 2003). The general objective of this research was therefore to examine the relationships between personality dimensions (as measured by the OPQ), sense of coherence (as measured by the OLQ) and coping styles (as measured by the COPE - questionnaire) in a sample of working mothers and to determine whether personality dimensions and sense of coherence can predict the various coping styles. A secondary objective of the research was to investigate the psychometric properties of a coping questionnaire (that is increasingly used in South Africa to measure coping) within the context of this study. ReSeARCh deSign Research approach A cross-sectional survey design that employed quantitative methodology was used for the present study. The empirical study was descriptive in nature, because relationships between personality dimensions (independent variable), sense of coherence (independent variable) and coping styles (dependent variable) were investigated. Participants The sample was obtained by contacting relevant managers in identified companies within the medical, aviation, motor, food, educational, engineering, mining and financial industries. The convenience sample consisted of 120 working mothers, representing different ethnic groups and occupying various occupational levels in several occupational fields, organisations and industries. A prerequisite was that they should be married, while also pursuing their careers and being responsible for fulfilling the traditional family responsibilities of mother and wife. The sample consisted mainly of Afrikaans-speaking (75,49%) and English-speaking (20,59%) participants, who were between the ages of 31 and 40 (72,55%), with 45,10% having GR12 and 25,49% having a degree. The participants were mainly white (85,29%), while other participants were African (5,88%), HERBST, COETzEE, VISSER60 Indian (4,90%) and Coloured (3,92%). The participants have between 16 and 20 years (31,37%) and 11 and 15 years (29,41%) of working experience. Most of the participants had one or two children (86,27%), and were married for 6 to 10 years (40,20%). The participants worked in the food (42,16%), aviation (14,91%) medical (12,75%), educational (12,75%), mining (7,84%), financial (4,90%) engineering (2,94%) and motor industry (1,96%). They were performing administrative related work (85,29%), professionally employed (11,76%) or performing managerial functions (2,94%). Most of the participants (52,94%) experienced an average degree of role conflict with regard to the different roles they were responsible for. measuring instruments Occupational Personality Questionnaire (OPQ32n) The Occupational Personality Questionnaire was developed to ensure that the scales of the OPQ32 are relevant and acceptable to the workplace, as well as being comprehensive in terms of the personality variables covered (Saville & Holdsworth, 1999a). Its item contents are also appropriate for use with people from different ethnic and gender groups (Saville & Holdsworth, 1999a). The OPQ32 describes 32 dimensions or scales of people’s preferred or typical styles of behaviour at work. The OPQ32 measures three domains that encompass several dimensions each (Saville & Holdsworth, 1999b). The three domains are: l The relating domain (Persuasive, Controlling , Outspoken, Independent minded, Outgoing, Affiliative, Socially confident, Modest, Democratic, Caring) l The thinking style domain (Data rational, Evaluative, Behavioural, Conventional, Conceptual, Innovative, Variety seeking, Adaptable, For ward thinking, Detail conscious, Conscientious, Rule following) l The feelings and emotions domain (Relaxed, Worrying, Tough minded, Optimistic, Trusting, Emotionally controlled, Vigorous, Competitive, Achieving, Decisive) The OPQ 32 requires respondents to consider a statement and then to indicate their responses on a five-point scale ranging from strongly agree to strongly disagree. There are 230 statements to consider (Saville & Holdsworth, 2004). The internal consistency reliabilities ranged from 0,67 to 0,88, with a median of 0,81. In terms of gender, women’s internal consistency reliabilities ranged from 0,66 to 0,88 (Saville & Holdsworth, 1999b). The OPQ questionnaire has good face validity for occupational use. The inductive approach and job analytic techniques used in developing the OPQ questionnaire are conducive to content validity. This high content validity justifies the use of the questionnaire in the work environment. Construct validities ranged from -0,44 to 0,40 on the various scales of the questionnaire (Saville & Holdsworth, 1999b). Orientation to Life Questionnaire (OLQ) The Orientation to Life Questionnaire is used to measure sense of coherence. The questionnaire consists of 29 five-facet items. There are 11 items measuring comprehensibility, 10 measuring manageability and 8 measuring meaningfulness. Respondents are provided with 29 self-reporting items on various aspects of life and each item has seven possibilities. The respondents are requested to mark the numbers which are most applicable to them, with 1 and 7 being the extreme answers (Antonovsky, 1987). Antonovsky (1993) concluded that the OLQ is a reliable instrument with sound construct validity for measuring sense of coherence. Strümpfer and Wissing (1998) also confirmed the reliability and validity of the OLQ in various South African studies. In accordance with these findings, Coetzee and Rothmann (1999), Naudé and Rothmann (2000) and Pretorius and Rothmann (2001) found alpha coefficients of 0,89, 0,88, and 0,93 respectively for the composite Orientation to Life Questionnaire. In another South African study, Breed, Cilliers and Visser (2006) obtained alpha coefficients of 0,88 and 0,83 for two samples of students. Coping Orientations to the Problems Experienced Questionnaire (Dispositional COPE Questionnaire) Despite many years of theory and research and the development of a variety of self-report coping instruments, researchers still do not have a comprehensive understanding of the structure of coping (Vollrath, 2001). The structure of coping and the best way to measure it remains unresolved. Whether researchers prefer broader or narrower dimensions often depends on the research question at hand and the methodology to be employed (Suls, David & Harvey, 1996). Three problems were identified by Carver et al. (1989) in surveying existing measures of coping processes: l None of the existing measures sampled all the domains that the developers of the COPE Questionnaire regarded to be of theoretical interest. l The measuring instruments appear to suffer to a greater or lesser degree from a lack of clear focus in some items. l Existing scales have been derived empirically, rather than theoretically. In response to the problems identified, Carver et al. (1989) developed the COPE Questionnaire. Several versions of the questionnaire have been used by researchers, but the 61- item version was used in the present study. It includes scales that measure problem-focused coping styles, emotion-focused coping styles and lesser used coping styles, namely focus on and venting of emotions, behavioural disengagement and mental disengagement. The problem-focused coping styles are active coping, planning, suppression of competing activities, restraints coping and seeking social support for instrumental reasons. The emotion-focused coping styles are seeking social support for emotional reasons, positive reinterpretation and growth, acceptance, turning to religion and denial. The scales require respondents to indicate the extent to which they use individual coping strategies to cope with a named event, such as the role conflict experienced by the working mother. Scores range from 1 “I usually don’t do this at all” to 4 “I usually do this a lot”. Four items make up each sub-scale or coping strategy, which implies that the scores for each sub-scale can range from 4 (indicating a minimal use of a particular coping strategy), to 16 (indicating a maximum use of a particular coping strategy). According to Carver et al. (1989), the internal consistencies of the COPE Questionnaire scales all exceeded 0,6 with the exception of the mental disengagement scale. Test-retest reliabilities showed reasonable stability and ranged from 0,42 to 0,89 for the different scales. Convergent and discriminant validity was also in line with expectations (Carver et al, 1989). However, Carver et al.’s (1989) findings were not supported by other research results. The COPE in its current form raised some concerns with regard to its reliability and validity. Since the origin of the instrument, factor analyses of the items resulted in varied underlying structures. When Pienaar and Rothman (2003) compared factor analytic results across various studies with those of Carver et al. (1989), they found that varying numbers of factors were being reported and also that the factor structures did not replicate consistently across the studies. They subsequently subjected the COPE to confirmatory factor analysis and internal consistency analysis. However, the expected factor structures of the COPE could not be confirmed. Also, many of the alpha coefficients were lower than the cut-off point of 0,70 (Nunnally & Bernstein, 1994). In their exploratory factor analysis COPING OF WORKING MOTHERS 61 on the items of the COPE, Pienaar and Rothman (2003) obtained four factors, namely active coping, avoidance coping, seeking emotional support and turning to religion. Their conclusion was that more exploratory work on the COPE was required. Procedure Responsible managers in identified companies within specific industries were contacted to establish their willingness to participate in the study. Thereafter letters were sent to the managers, indicating the details of the study. On approval, logistic arrangements were made in order to administrate the measuring instruments to identified participants. ReSUlTS Validity and reliability of the CoPe In view of the concerns regarding the reliability and validity of the COPE, we decided to investigate its internal consistency reliability, and also its construct validity by means of an exploratory factor analysis (EFA) of the items. Before conducting the EFA, diagnostic tests were performed to ascertain whether the intercorrelation matrix contained an adequate number of substantial pairwise correlations and also low partial correlations to justif y conducting the factor analysis. The Kaiser Meyer – Olkin measure of sampling adequacy (MSA) was unsatisfactory at 0,59. Furthermore, many of the items yielded MSA’s smaller than 0,60. These items were systematically dropped until all the MSA’s of the remaining items were larger than 0,60. Forty four items remained. The Kaiser Meyer – Olkin MSA was then equal to 0,76. These results provided sufficient justification to proceed with the factor analysis. A principal components analysis (PCA) was subsequently performed using the intercorrelations between the 44 items as its input. Computer software developed by Watkins (2000) that generates eigenvalues by means of Monte Carlo simulations of Principal Components Analyses was then employed to determine the number of factors to retain for rotation. These eigenvalues, created through 100 replications, were used as the criteria for Horn’s Parallel Analysis (Horn, 1965). Five eigenvalues of the initial PCA proved to be larger than the Monte Carlo PCA eigenvalues and we therefore decided to retain five factors which accounted for 53,18% of the variance in the data. The intercorrelation matrix was then analysed by means of the principal axis factoring technique, extracting five factors. The resulting factor matrix was rotated to a simple structure using Kaiser’s varimax rotation procedure. The rotated factor matrix, which also includes the communalities associated with the scales, is presented in Table 1. The variables were ordered and grouped by size of loading to facilitate interpretation. The first factor was labelled as Venting of Emotions. Items loading on this factor relate to the free expression of emotions as a result of experiencing stressful events. Factor 2 was identified as Active Coping. Its items refer to actively working at trying to find a solution. The focus is on finding a strategy by means of which to address the problem. The third factor was labelled Seeking Social Support, because its items deal with involving another person or persons to get advice, clarity or support in handling the problem. Factor 4 was named Appraisal of Event. The items loading on this factor deal with interpreting the problem whilst maintaining a positive outlook on the problem and taking care to handle the stressful event in a mature manner. The fifth factor was labelled Turning to Religion and referred to praying and religious behaviour. Two items yielded negative factor loadings on Factor 4, namely Items 46 and 60. These items were subsequently reverse-scored for the calculation of scale reliabilities and for further analyses. The items with the highest factor loadings on each of the five factors were then used to define the coping sub-scales of the COPE Questionnaire. It is evident that there are similarities between these results and those found in a study conducted by Pienaar and Rothmann (2003). They identified four factors, namely active coping, avoidance coping, seeking social support and turning to religion after factor analysing a 53-item version of the COPE. Pienaar and Rothman (2003) also indicated that 12 of the 53 items did not load on the identified factors and had to be omitted from further analyses. Table 1 RoTaTed facToR maTRix and communaliTies foR PRinciPal axis facToRing and VaRimax RoTaTion of The coPe iTems item factors Commu- nalities 1 2 3 4 5 49 0,83 0,05 0,23 0,01 -0,05 0,74 35 0,79 0,05 0,31 0,14 -0,09 0,75 61 0,78 0,10 0,23 0,05 -0,06 0,67 16 0,76 -0,02 0,22 -0,07 -0,02 0,63 31 0,63 -0,15 0,03 -0,31 -0,05 0,51 9 0,60 0,11 0,51 -0,10 -0,07 0,64 42 0,57 0,08 0,17 0,10 0,05 0,38 28 0,43 0,31 0,17 0,34 0,19 0,46 54 0,24 0,75 0,06 0,09 0,15 0,65 52 -0,01 0,63 0,03 0,18 -0,04 0,43 40 0,02 0,68 0,11 0,17 -0,02 0,41 7 -0,05 0,61 0,14 0,27 0,15 0,48 14 0,15 0,60 0,01 0,11 0,07 0,40 23 -0,02 0,57 0,24 0,10 0,19 0,43 22 0,07 0,52 0,06 0,36 0,12 0,42 21 0,02 0,48 0,24 0,19 0,25 0,38 53 0,07 0,45 -0,12 0,14 -0,06 0,25 6 -0,10 0,35 0,23 0,11 0,02 0,20 3 -0,04 0,30 -0,08 0,27 0,16 0,19 25 0,20 0,12 0,80 0,07 0,10 0,71 47 0,26 0,20 0,71 -0,17 0,07 0,65 56 0,46 0,02 0,67 -0,02 0,01 0,67 39 0,14 0,18 0,64 0,06 -0,09 0,48 17 0,47 -0,07 0,63 -0,12 0,13 0,66 1 0,17 0,02 0,61 0,00 0,02 0,40 32 0,40 -0,01 0,58 -0,13 0,12 0,53 26 0,05 0,16 0,20 0,70 0,20 0,59 60 -0,34 -0,20 0,02 -0,56 0,07 0,48 19 0,19 0,32 -0,18 0,55 0,01 0,48 46 -0,40 -0,15 -0,10 -0,53 0,02 0,47 20 -0,09 0,25 -0,10 0,52 -0,04 0,35 10 -0,01 0,43 0,27 0,50 0,00 0,50 50 0,13 0,38 0,06 0,48 0,01 0,39 34 0,19 0,21 -0,01 0,48 0,06 0,31 4 -0,09 0,25 -0,22 0,46 -0,07 0,33 38 -0,06 0,19 0,04 0,45 0,01 0,24 8 0,01 0,29 -0,07 0,40 0,00 0,25 36 0,21 0,17 -0,06 0,37 0,19 0,25 24 -0,14 0,26 0,15 0,37 0,06 0,25 55 0,06 0,32 -0,16 0,37 -0,01 0,27 58 -0,03 0,06 0,07 0,12 0,91 0,85 27 -0,01 0,10 0,04 -0,03 0,91 0,84 43 0,02 0,14 0,02 -0,02 0,87 0,79 11 -0,10 0,07 0,04 0,19 0,78 0,66 Factor loadings ≥ 0,3 are in bold type Factor labels: F1: Venting of Emotions, F2: Active Coping, F3: Seeking Social Support, F4: Appraisal of Event, F5: Turning to Religion HERBST, COETzEE, VISSER62 The means, standard deviations, mean inter-item correlations, and Cronbach alpha coefficients of the five COPE scales are given in Table 2. Table 2 means, sTandaRd deViaTions, mean inTeR-iTem coRRelaTions, and cRonbach alPha coefficienTs of The fiVe coPe scales Sub-scale (factor) mean Stan- dard devia- tion num- ber of items item mean inter- item corre- lation Cron- bach alpha Venting of Emotions 20,94 5,75 8 2,62 0,50 0,89 Active Coping 32,94 5,61 11 2,99 0,33 0,84 Seeking Social Support 17,90 5,25 7 2,56 0,53 0,89 Appraisal of Event 41,05 6,68 14 2,93 0,27 0,83 Turning to Religion 13,34 3,23 4 3,34 0,76 0,93 Inter-item correlation coefficients were used to determine whether the internal consistencies of the constructs were not too high, thereby affecting the validity. Clark and Watson (1995) suggested that mean inter-item correlations between 0,15 and 0,50 are acceptable. The mean inter-item correlations of the COPE sub- scales ranged from 0,27 to 0,76 for the shortest sub-scale, namely the four-item Turning to Religion scale. The internal consistency reliabilities of the five sub-scales were acceptable, because the Cronbach alphas ranged from 0,83 to 0,93. The item means of the five sub-scales varied between 2,56 for Seeking Social Support and 3,34 for Turning to Religion. The means and standard deviation cannot be compared with results of other studies, because the scales are made up of different items. Nevertheless, the obtained alpha coefficients compared favourably with reliabilities obtained by Pienaar and Rothman (2003), who reported reliability coefficients of between 0,80 and 0,92 for the sub-scales that they identified. descriptive statistics and reliabilities of the oPQ32n and olQ Descriptive statistics and internal consistency reliabilities of the OPQ32n and OLQ are reported in Table 3. Saville and Holdsworth (1999b) reported mean scores for the 32 OPQ32n scales of between 15,49 and 25,65, with standard deviations ranging between 3,71 and 5,73. Similar results were found in the current study, because the mean scores were between 15,92 and 28,01, and the standard deviations between 3,16 and 5,94. The reliability coefficients for the 32 scales ranged between 0,60 and 0,87. The OPQ32n may therefore be regarded as an internally consistent measuring instrument for personality dimensions/traits. Similar findings were obtained by Saville and Holdsworth (1999b) who reported reliability coefficients between 0,63 and 0,87. When respondents’ scores on the three sub-scales of the OLQ were added, scores for a composite sense of coherence measure were obtained. The mean for the 120 respondents was 134,26 and the standard deviation was 23,35. Wissing and Van Eeden (1997) obtained a mean score of 136,52 with a standard deviation of 21,68, whereas Coetzee and Rothman (1999) obtained a mean score of 143.11 with a standard deviation of 21.42. The internal consistency reliability coefficient of the OLQ was equal to 0,89. This finding is similar to the coefficients ranging from 0,82 to 0,95 reported by Antonovsky (1993), and that of 0,89 reported by Coetzee and Rothman (1999). The reliabilities of the three sub-scales, comprehensibility, manageability, and meaningfulness, ranged from 0,77 to 0,82. These findings indicate that the composite scale as well as the sub-scales of the OLQ yielded internally consistent measures. Table 3 descRiPTiVe sTaTisTics and cRonbach alPhas of The oPQ32n and olQ (n=102) Scale mean Standard deviation Skewness Kurtosis Alpha oPQ32n Persuasive 19,48 5,40 0,05 -0,68 0,82 Controlling 21,89 4,58 -0,56 -0,14 0,81 Outspoken 20,71 4,70 0,01 -0,39 0,73 Independent Minded 21,18 4,33 -0,12 -0,17 0,71 Outgoing 21,22 5,33 -0,12 -0,85 0,83 Affiliative 21,79 5,94 -0,56 -0,27 0,85 Socially Confident 21,40 5,02 -0,27 -0,92 0,85 Modest 19,08 4,50 0,17 -0,88 0,81 Democratic 24,83 4,09 -0,93 0,83 0,67 Caring 26,80 4,86 -0,63 0,35 0,78 Data Rational 22,14 4,84 -0,82 0,38 0,81 Evaluative 23,86 3,88 -0,14 -0,39 0,63 Behavioural 25,56 4,77 -0,66 0,83 0,82 Conventional 20,39 4,73 -0,11 -0,63 0,76 Conceptual 21,61 4,11 0,14 -0,45 0,66 Innovative 22,08 4,34 -0,26 -0,04 0,77 Variety Seeking 21,95 4,87 0,12 -0,82 0,73 Adaptable 20,21 4,44 -0,18 -0,36 0,76 Forward Thinking 22,30 4,55 -0,35 -0,64 0,79 Detail Conscious 26,05 4,84 -0,51 0,17 0,75 Conscientious 26,46 3,16 -0,24 -0,22 0,60 Rule Following 23,82 5,07 -0,58 -0,41 0,85 Relaxed 19,54 5,19 -0,09 -0,95 0,86 Worrying 22,50 4,85 -0,43 -0,68 0,87 Tough Minded 17,38 4,81 0,20 -0,59 0,83 Optimistic 25,85 5,28 -0,26 -0,48 0,83 Trusting 18,98 5,08 -0,07 -0,48 0,76 Emotional Control 20,32 4,92 -0,11 -0,66 0,83 Vigorous 28,01 4,22 -0,48 0,18 0,77 Competitive 15,92 4,52 1,03 1,09 0,83 Achieving 24,99 5,21 -0,42 0,10 0,78 Decisive 19,67 4,03 0,19 -0,48 0,77 olQ Comprehensibility 44,41 10,90 -0,35 0,07 0,82 Manageability 48,46 9,20 -0,83 1,31 0,77 Meaningfulness 41,39 7,96 -1,02 2,18 0,78 Sense of Coherence 134,26 23,35 -0,88 2,14 0,89 Correlations of personality variables and sense of coherence with coping styles Product-moment correlation coefficients were computed between the personality dimensions and coping styles, and also bet ween the sense of coherence scales and coping styles. These correlations are indicated in Table 4. Statistical significance was indicated with asterisks, whereas practical significance implying medium or large effects, was indicated by means of addition signs. COPING OF WORKING MOTHERS 63 Table 4 coRRelaTions of PeRsonaliTy dimensions and sense of coheRence wiTh coPing sTyles Venting of emotions Active Coping Seeking Social Support Appraisal of event Turning to Religion oPQ 32n Persuasive RP1 0,09 0,26** -0,06 0,19 0,10 Controlling 0,09 0,23* -0,16 0,20* 0,12 Outspoken 0,26** 0,24* -0,01 0,05 -0,10 Independent Minded -0,21* 0,19 -0,31**+ 0,13 0,01 Outgoing 0,42**+ 0,13 0,20* 0,01 -0,11 Affiliative 0,36**+ 0,00 0,27** -0,11 0,03 Socially Confident 0,19 0,28** 0,02 0,19 -0,10 Modest -0,34**+ 0,03 -0,22* -0,05 0,15 Democratic 0,22* 0,13 0,30**+ 0,14 0,01 Caring RP10 0,14 0,21* 0,11 -0,05 0,28** Data Rational TS1 -0,18 0,15 -0,28** 0,25* 0,13 Evaluative -0,13 0,16 -0,17 0,22* -0,20* Behavioural 0,15 0,26** 0,13 0,17 -0,08 Conventional 0,03 0,01 0,26** 0,03 0,02 Conceptual -0,02 0,18 -0,10 0,04 -0,11 Innovative 0,15 0,20* -0,04 0,11 -0,22* Variety Seeking 0,07 0,10 -0,06 0,05 0,04 Adaptable -0,04 -0,18 0,02 -0,26** -0,01 Forward Thinking 0,03 0,18 -0,02 0,14 0,11 Detail Conscious 0,15 0,34**+ 0,00 0,26** 0,00 Conscientious -0,06 0,11 -0,16 0,11 0,19 Rule Following TS12 -0,13 0,13 0,00 0,11 0,25* Relaxed FE1 0,13 0,07 0,03 0,31**+ 0,07 Worrying -0,17 -0,13 0,06 -0,24* -0,02 Tough Minded -0,07 0,09 -0,11 0,30**+ 0,00 Optimistic 0,23* 0,12 -0,04 0,11 0,00 Trusting 0,15 0,07 0,18 0,01 -0,03 Emotional Control -0,60**++ 0,09 -0,29** 0,22* 0,06 Vigorous 0,02 0,16 -0,16 0,03 -0,11 Competitive 0,01 0,24* -0,20* 0,21* -0,02 Achieving 0,06 0,28** -0,08 0,20* 0,03 Decisive FE10 -0,04 0,00 -0,19 -0,07 0,06 olQ Comprehensibility 0,11 0,10 0,03 0,18 -0,02 Manageability 0,13 0,14 0,08 0,41**+ 0,09 Meaningfulness 0,21* 0,32**+ 0,09 0,40**+ 0,19 Sense of Coherence 0,17 0,21* 0,08 0,38**+ 0,09 *Statistically significant: p ≤ 0,05 * *Statistically significant: p ≤ 0,01 + Practically significant: r ≥ 0,30 (medium effect) ++ Practically significant: r ≥ 0,50 (large effect) Practically significant correlations were obtained between the coping style Venting of Emotions and four OPQ subscales, namely Outgoing, Affiliative, Modest, and Emotional Control (large effect). No practically significant correlations were obtained between Venting of Emotions and the OLQ sub-scales. In addition, Outspoken, Independent minded, Democratic, Optimistic, and OLQ Meaningfulness, yielded statistically significant correlations with Venting of Emotions. The highest correlation was between Emotional control and Venting of Emotions (r = -0,60). Active Coping yielded correlations representing medium effects with Detail Conscious and OLQ Meaningfulness only. Statistically significant correlations were obtained with nine more variables, namely Persuasive, Controlling, Outspoken, Socially Confident, Caring, Behavioural, Innovative, Competitive, and Achieving. The coping factor, Seeking Social Support, correlated practically significantly with only two OPQ variables, namely Independent Minded and Democratic. For seven more OPQ variables statistically significant correlations were obtained. They were Outgoing, Affiliative, Modest, Data Rational, Conventional, Emotional Control, and Competitive. Practically significant correlations were obtained between the coping style Appraisal of Event and two OPQ sub-scales, namely Relaxed and Tough Minded. In addition, the OPQ scales Controlling, Data Rational, Evaluative, Adaptable, Detail Conscious, Worrying, Emotional Control, Competitive, and Achieving yielded statistically significant correlations with this coping factor. The highest correlations for Appraisal of Event was its practically significant correlations with OLQ Manageability (r = 0,41) and OLQ Meaningfulness (r = 0,40). There were no practically significant correlations between Turning to Religion and the OPQ or OLQ variables, but statistically significant correlations were obtained for four OPQ variables, namely Caring, Evaluative, Innovative, and Rule following. Regression analysis of the coping styles on personality dimensions and sense of coherence In the present study five multiple regression analyses, one for each of the five coping styles, were conducted to determine the proportion of the total variance of the various coping styles that were explained by personality and sense of coherence. The OPQ dimensions and the three OLQ sub-scales were the independent variables, and the five coping styles were the dependent variables for the respective regression analyses. In order to reduce the numbers of independent variables, only variables with statistically significant correlations with the dependent variables were entered in the regression, that is, variables marked by means of asterisks in Table 4. Standard multiple regression analyses were conducted, because our interest was not in finding prediction equations for predicting the coping styles for our sample. Rather, our interest was in assessing the magnitude of the correlations between the dependent and independent variables (using Table 4) and in assessing the magnitude of the overall relationship between the coping styles and the independent variables. Standard multiple regression also enabled us to assess how much of the overall relationship was contributed uniquely by each independent variable, because the independent variables are evaluated in terms of how much they add to the prediction of the coping styles that differs from the percentage of variance accounted for by the combination of all the other independent variables. The data was examined to determine whether the assumptions underlying multiple regression were met for the five regressions to be performed. Firstly, the scatterplots between the independent and dependent variables were examined to establish whether the relationships were linear. No sign of marked nonlinearity was observed. Secondly, the residual plots of the standardised residual values against the standardised predicted values were examined to determine whether the error values were independent and yielded equal variances. There was no indication of correlations between these errors, and a fair degree of homoscedasticity appeared to be present. Thirdly, the normal probability plots of the residuals against the expected normal values were examined to establish whether the error values yielded normal distributions. Some deviation from normality was observed. Since regression is fairly robust HERBST, COETzEE, VISSER64 to moderate deviations from normality, we decided to proceed as if the assumption of normality was met. When testing the assumptions, the results for Factor 5 (Turning to Religion) consistently appeared to be borderline. The fact that the factor was measured by four items only, probably affected the stability of the measure. The data was also examined by means of an outlier analysis to determine whether there were influential cases that could unduly affect the regression results. Cook’s distance was calculated. None of the cases in any of the variables presented a Cook’s distance greater than one and so none of the cases had an undue influence on the model. The regression results are given in Table 5. For Venting of Emotions, the R for regression (R = 0,66) was statistically significant (F(9,91) = 7,70, p < 0,0001). Only one regression coefficient differed significantly from zero, namely that for Emotional control (t = -4,77, p<0,000001). The implies that only one variable (Emotional control) offered a unique contribution that was statistically significant for the prediction of Venting of Emotions once the contributions of the remaining independent variables had been accounted for (Part r2 = 0,14 or 14%). Altogether, 43% (38% adjusted) of the variability of Venting of Emotions was predicted by the nine variables that were entered into the regression. In addition to the 14% of variance contributed uniquely by Emotional control for the prediction, the nine variables in combination shared another 29% of the variance of Venting of Emotions. With regard to Active Coping, the R for regression (R = 0,54) was also statistically significant (F(11,90) = 3,32, p = 0,0007). Two independent variables, Competitive and Detail conscious, contributed significantly to the prediction of Active Coping, each sharing 4% of unique variance with Active Coping. The 11 independent variables used in the regression in combination shared another 21% of the variability of Active Coping. Altogether, 29% (20% adjusted) of the variability of Active Coping was accounted for by 11 independent variables. For Seeking Social Support, none of the nine independent variables yielded statistically significant unique contributions toward predicting Seeking Social Support, despite the correlations of two variables with the dependent variable representing medium effect sizes. The multiple correlation was equal to 0,55 and was statistically significant (F(9,92) = 4,35, p = 0,0001). The nine independent variables explained thirty percent of the variance of Seeking Social Support (20% adjusted). The multiple correlation for Appraisal of Event (R = 0,67) was statistically significant (F(13, 88) = 5,41, p<0,00001). Two independent variables, namely Emotional control and OLQ Meaningfulness, shared statistically significant unique portions of variance with Appraisal of Event. Their contributions amounted to 12% and 3% respectively. In combination, the 13 independent variables used in the regression shared another 29% of the variability of Appraisal of Event. Altogether, 44% (36% adjusted) of the variability of Appraisal of Event was accounted for by the 13 independent variables. For Turning to Religion, four independent variables were used in the regression. The multiple correlation (R = 0,41) was statistically significant (F(4, 97) = 5,02, p = 0,001). Only one independent variable, Caring, yielded a regression coefficient that differed significantly from zero and presented a unique contribution (5%) to explaining the variance in Turning to Religion. In total, 17% (14% adjusted) of the variability in Turning to Religion was predicted by the four independent variables. discussion The purpose of the study was to examine the relationships between personality dimensions, sense of coherence and coping styles in a sample of working mothers. A secondary objective of the research was to investigate the psychometric properties of the COPE questionnaire within the context of this study. Although the COPE is an instrument that is widely used to assess coping strategies, researchers have found contradictory results regarding the dimensionality of the COPE (Pienaar & Rothmann, 2003). As a result, it is problematic to compare results of studies using the COPE in different settings and for different samples, or to generalise the findings of these studies. An exploratory factor analysis of the COPE in the current study yielded five factors with acceptable levels of internal consistency, namely Venting of Emotions, Active Coping, Seeking Social Support, Appraisal of Event and Turning to Religion. These results appear somewhat similar to the factors found by Pienaar and Rothmann (2003). For instance, three of their identified four factors were named active coping, seeking social support and turning to religion. They identified a fourth factor namely avoidance coping that appears different from the other two factors identified in this study. However, despite the neat solution obtained, the limited sample size of the current study warrants that the results of the factor analysis should be interpreted with caution. All the measuring instruments used in the study, including those for the independent variables, namely personality traits (OPQ) and sense of coherence (OLQ), showed acceptable levels of internal consistency for this sample of working mothers. The results showed that personality traits mainly from the relating, and feelings and emotions domains of the OPQ correlated with venting of emotions as a coping strategy. It therefore appeared that working mothers who were outspoken, outgoing, affiliative, democratic and optimistic, but who tended to show a lack of emotional control and were also not modest or independent minded, would use venting of emotions as a coping strategy. Working mothers who used venting of emotions as a coping strategy also appeared to regard stressors as worth investing time and energy in (meaningfulness). As was expected, emotional control made a unique contribution of 14% towards explaining the variance of venting of emotions as a coping strategy. Working mothers who are persuasive, like to be in charge, are outspoken, socially confident and caring, who try to understand the motives of others, are innovative, competitive, achieving and detail conscious, would be likely to engage in an active coping strategy. This finding appears to be in line with previous research by Roesch, Christian and Vaughn (2006) who also found that extraversion (related to being persuasive and outgoing on the OPQ as shown by Visser and Du Toit, 2004), conscientiousness (related to being controlling, detail conscious and achieving on the OPQ as shown by Visser and Du Toit, 2004) and openness (related to being innovative on the OPQ as shown by Visser and Du Toit, 2004) were positively related to active coping. Furthermore, working mothers who tend to engage in the active coping strategy, would have a strong sense of coherence and experience high levels of meaningfulness, as also previously shown by Rothmann (2004) and Redelinghuys and Rothmann (2005). However, being competitive and detail conscious showed unique contributions of only 4% each towards explaining the variance of active coping. Seeking social support as a coping strategy would more likely be used by working mothers who lack emotional control, who are not independent minded, modest or competitive, who prefer dealing with feelings and opinions, who are outgoing and democratic, and prefer a conventional approach. These findings confirmed results of Storm and Rothmann (2003) who also found that extraversion, openness to experience and conscientiousness were correlated with seeking social support. It appeared that the strength of sense of coherence of working mothers would not influence their use of this strategy. In addition, none of COPING OF WORKING MOTHERS 65 Table 5 sTandaRd mulTiPle RegRession analysis RegaRding PeRsonaliTy dimensions, sense of coheRence and coPing sTyles model Standardised coefficients non-standardised coefficients t p Part r2 (unique) Beta Se B Venting of emotions Intercept 30,60 4,48* 0,00002 Optimistic 0,04 0,09 0,05 0,46 0,649 0,001 Emotional control -0,49 0,10 -0,57 -4,77* 0,000000 0,14 Outspoken -0,11 0,11 -0,13 -0,95 0,345 0,006 Independent minded -0,07 0,10 -0,10 -0,74 0,460 0,003 Outgoing 0,15 0,11 0,17 1,39 0,168 0,01 Affiliative 0,06 0,10 0,06 0,63 0,530 0,003 Modest -0,10 0,09 -0,13 -1,11 0,269 0,008 Democratic 0,07 0,09 0,10 0,80 0,424 0,004 OLQ Meaningfulness 0,014 0,09 0,01 0,15 0,878 0,0001 R = 0,66 R2 = 0,43 Adjusted R2 = 0,38 Active Coping Intercept 5,15 1,02 0,309 Competitive 0,23 0,11 0,28 2,13* 0,036 0,04 Achieving 0,02 0,12 0,02 0,19 0,852 0,0003 Persuasive 0,09 0,11 0,10 0,88 0,384 0,006 Caring 0,12 0,10 0,14 1,13 0,260 0,01 Controlling -0,02 0,12 -0,03 -0,18 0,861 0,0002 Outspoken 0,02 0,11 0,02 0,16 0,875 0,0002 Socially minded 0,04 0,12 0,04 0,31 0,754 0,0008 Detail conscious 0,21 0,10 0,24 2,18* 0,032 0,04 Behavioural 0,17 0,10 0,20 1,59 0,116 0,02 Innovative 0,01 0,10 0,01 0,09 0,926 0,00007 MEAN 0,17 0,10 0,12 1,65 0,103 0,02 R = 0,54 R2 = 0,29 Adjusted R2 = 0,20 Seeking Social Support Intercept 19,68 2,84* 0,005 Emotional control -0,20 0,10 -0,22 -1,96 0,053 0,03 Competitive -0,11 0,10 -0,13 -1,16 0,249 0,01 Independent minded -0,02 0,11 -0,02 -0,17 0,864 0,0002 Outgoing 0,01 0,11 0,01 0,06 0,952 0,00003 Affiliative 0,13 0,11 0,12 1,20 0,232 0,01 Modest -0,11 0,10 -0,13 -1,11 0,269 0,009 Democratic 0,16 0,10 0,21 1,60 0,113 0,02 Data rational -0,19 0,10 -0,20 -1,95 0,055 0,03 Conventional 0,19 0,10 0,21 1,96 0,054 0,03 R = 0,55 R2 = 0,30 Adjusted R2 = 0,23 Appraisal of event Intercept 1,57 0,18 0,859 Relaxed 0,02 0,10 0,02 0,16 0,873 0,0002 Worrying -0,14 0,10 -0,19 -1,45 0,150 0,01 Tough minded 0,17 0,09 0,23 1,86 0,066 0,02 Emotional control 0,37 0,08 0,50 4,39* 0,00003 0,12 Competitive 0,11 0,09 0,17 1,28 0,203 0,01 Achieving 0,06 0,10 0,07 0,54 0,594 0,002 Controlling -0,07 0,11 -0,10 -0,65 0,517 0,003 Data rational 0,07 0,10 0,10 0,74 0,464 0,003 Detail conscious 0,09 0,09 0,12 0,98 0,328 0,006 Evaluative 0,13 0,09 0,23 1,45 0,150 0,01 Adaptable 0,003 0,09 0,01 0,04 0,966 0,00001 Manageability 0,22 0,12 0,16 1,78 0,078 0,02 Meaningfulness 0,24 0,12 0,20 2,01* 0,047 0,03 R = 0,67 R2 = 0,44 Adjusted R2 = 0,36 Turning to Religion Intercept 13,87 4,57* 0,00001 Caring 0,24 0,10 0,16 2,37* 0,020 0,05 Rule following 0,11 0,11 0,07 1,07 0,287 0,01 Evaluative -0,18 0,09 -0,15 -1,89 0,062 0,03 Innovative -0,19 0,10 -0,14 -1,88 0,063 0,03 R = 0,41 R2 = 0,17 Adjusted R2 = 0,14 *p ≤ 0,05 HERBST, COETzEE, VISSER66 the personality traits yielded a statistically significant unique contribution towards explaining the variance of seeking social support as a coping strategy. With regard to appraisal of event, it appeared that working mothers who would be most likely to use this coping strategy, are those who like to be in charge, base their decisions on facts and figures, critically evaluate information, who behave consistently across situations, are detail conscious and achieving, who find it easy to relax and experience little worry, who are not easily offended and tend to show emotional control. Working mothers who have a strong sense of coherence and who experience stressors to be manageable and meaningful, would also tend to use this coping strategy. These results confirm the results of Rothmann (2004) who also found that a strong sense of coherence correlated with positive reinterpretation. Meaningfulness uniquely accounted for 3% of the variance of appraisal of event as a coping strategy. In addition, emotional control explained another 12% of the variance. Personality traits that correlated with turning to religion, were mainly from the thinking style domain of the OPQ. Working mothers who tended to be rule following, helpful and supportive to others, but who did not focus on potential limitations and were not inclined to be creative and inventive, would tend to turn to religion as a coping strategy. No relationship was found between sense of coherence and turning to religion as a coping strategy. This is somewhat contradictory to the finding of Strang and Strang (2001) which indicated that the meaningfulness component of sense of coherence was created by faith. Of the personality traits, only caring yielded a significant unique contribution of 5% toward explaining the variance in turning to religion as a coping strategy. Overall the results indicated that only a small percentage of variance in coping styles (4%-15%) could be accounted for by the combination of personality traits and sense of coherence. As reported in the introduction, previous research also found relatively modest links between coping, dispositions and traditional personality variables. This suggests that the choice of particular coping styles may be influenced by factors other than stable dispositions or personality traits of working women. Working mothers may therefore benefit from training regarding the various coping styles that are available and the appropriateness of each for coping with various stressors or scenarios. This in turn may enable them to extend their repertoire of coping skills to counter the effects of increased stress or role conflict, despite their own dispositions or personality traits. As also recommended by Pienaar and Rothmann (2003), future studies on coping should be done in conjunction with personality measures to further clarif y the relationships between personality variables and coping dispositions. Because the sample size of the present study was limited, the results of the factor analysis and some of the regression analyses should be interpreted with caution. Nevertheless, the results certainly pointed out that the COPE yielded yet another set of correlates that did not duplicate earlier research. Therefore, researchers should continue to try to identif y a standardised measure of coping strategies that can be used in comparing results across different studies. The five- dimensional structure of coping found in this study appeared to represent a promising and workable model, but attention should be paid to the turning to religion factor by adding relevant items to its scale in order to enhance its internal consistency. ACKnowledgemenT We gratefully acknowledge the contributions of Riët te Eiselen and Anneli Hardy of the Statistical Consultation Service of the University of Johannesburg and Tina Joubert of SHL (South Africa) who assisted with the computer analyses. RefeRenCeS Acker, J. (1992). The future of women and work: Ending the twentieth century. Sociological Perspectives, 35 (1), 53-68. Antonovsky, A. (1979). Health, stress and coping. San Francisco: Jossey-Bass. Antonovsky, A. (1984). The sense of coherence as a determinant of health. In J.D. Matarazzo, S.M. Weiss, J.A. Herd, M.E. Miller & S.M. Weiss (Eds.), Behavioural health: A handbook of health enhancement and disease prevention (pp. 114-129). New York: Wiley-Interscience. Antonovsky, A. (1987). Unravelling the mystery of health. London: Jossey-Bass. Antonovsky, A. (1993). Complexity, conflict, chaos, coherence, coercion and civility. Social Science and Medicine Journal, 37 (8), 969-981. Biernat, M. & Wortman, C.B. (1991). Sharing of home responsibilities between professionally employed women and their husbands. Journal of Personality and Social Psychology, 60, 844-860. Brannen, J. & Moss, P. (1991). Managing mothers: Dual earners households after maternity leave. London: Unwin Hyman. Breed, M., Cilliers, F., & Visser, D. (2006). The factor structure of six salutogenic constructs. South African Journal of Industrial Psychology, 32(1), 74-86. Carrim, S.O. (2000). Working women in their multiple role environment: A salutogenic perspective. Unpublished master thesis, University of South Africa, Pretoria. Carver, C.S. & Scheier, M.F. (1994). Situational coping and coping dispositions in a stressful transaction. Journal of Personality and Social Psychology, 66, 184-195. Carver, C.S. & Scheier, M.F. (1989). On the self-regulation of behaviour. New York: Cambridge University Press. Carver, C.S., Scheier, M.F. & Pozo, C. (1992). Conceptualising the process of coping with health problems. In H.S. Friedman (Ed.), Hostility, coping and health (pp. 167-199). Washington, DC: American Psychological Association. Carver, C.S., Scheier, M.F. & Weintraub, J.K. (1989). Assessing coping strategies: A theoretically based approach. Journal of Personality and Social Psychology, 56 (2), 267-283. Chusmir, L.H. & Durand, D.E. (1987). Stress and the working mother. Personnel, 64(5), 38-43. Clark, L.A. & Watson, D. (1995). Construct validity: Basic issues in objective scale development. Psychological Assessment, 7, 309-319. Coetzee, S.C. & Rothmann, S. (1999). Die verband tussesn koherensiesin en werkstevredenheid by bestuurders. Journal of Industrial Psychology, 25(3), 31-38. Cohen, J. (1988). Statistical power analysis for the behavioural sciences (Revised edition). Orlando, CA: Academic. Collins, R.L., Taylor, S.E. & Skokan, L.A. (1990). A better world or a shattered vision? Changes in life perspectives following victimization. Social Cognition, 8, 263-285. Douglas, S. & Michaels, M. (2004). The mommy myth: The idealization of motherhood and how it has undermined women. New York: Free Press. Eksi, H. (2004). Personality and coping: Multidimensional research on situational and dispositional coping. Educational Sciences: Theory and Practice, 4(1), 94-98. Eysenck, M.W. (1983). Anxiety and individual differences. In G.R.J. Hockey (Eds.), Stress and fatigue in human performance (pp. 273 - 298). New York: Wiley. Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30, 179-185. Howell, D.C. (1995). Fundamental statistics for the behavioural sciences (3rd ed.). California: Duxbury. Huysamen, G.K. (1993). Metodologie vir die sosiale en gedragswetenskappe. Johannesburg: Southern Boek- uitgewers. Klerman. J.A. & Leibowitz, A. (1994). The work-employment distinction among new mothers. Journal of Human Resources, 29 (2), 277-303. COPING OF WORKING MOTHERS 67 Koziara, K.S., Moskow, M.H. & Tanner, L.D. (1987). Working women: A summary of research knowledge and research needs. In K.S. Koziara, M.H. Moskow & L.D. Tanner (Eds.), Working women: Past, present, future (vii-xxii). Washington: Bureau of National Affairs. Lazarus, R.S. (1966). Psychological stress and the coping process. New York: McGraw –Hill. Lazarus, R.S. & Folkman, S. (1984). Stress, appraisal and coping. New York: Springer. Maconachie, M. (1990). Childcare: A study of white married couples with young children. South African Journal of Sociology, 21 (3), 197-177. May, C. E. (1999). An enquiry into the factors impeding the career advancement of women academics at Peninsula Technikon. Unpublished master’s thesis, University of the Western Cape, Belville, South Africa. Mattlin, J., Wethington, E. & Kessler, R.C. (1990). Situational determinants of coping and coping effectiveness. Journal of Health and Social Behaviour, 31, 103-122. McCrae, R.R. & Costa, P.T., Jr. (1986). Personality, coping and coping effectiveness in an adult sample. Journal of Personality, 54, 385-405. Mlonzi, E.N. & Strümpfer, D.J.W (1998). Antonovsky’s sense of coherence scale and 16PF second-order factors. Social Behaviour and Personality, 26 (1), 39-50. Moore, D.S. (1995). The basic practices of statistics. New York: W.H. Freeman. Mouton, J. & Marais, H.C. (1992). Basic concepts: Methodology of the behavioural sciences. Pretoria: HSRC. Naidoo, A.V. & Jano, R. (2002). Role salience of dual-career women managers. South African Journal of Industrial Psychology, 28 (3), 69-74. Nunally, J. & Bernstein, I.H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill. Ozer, E.M. (1995). The impact of childcare responsibility and self- efficacy in the psychological health of professional working mothers. Psychology of Women Quarterly, 19 (3), 315-335. Pearlin, L.I. & Schooler, C. (1978). The structure of coping. Journal of Health and Social Behavior, 19, 2-21. Pienaar, J. & Rothmann, S. (2003, June). Coping strategies in the South African Police Ser vice. Paper presented at the 6th Annual Conference of the Society for Industrial/Organisational Psychology. Johannesburg, South Africa. Redelinghuys, F.J. & Rothmann, S. (2005). Koherensiesin, coping, uitbranding en begeestering van predikante. Tydskrif vir Geesteswetenskappe, 45(4), 466-477. Richardson, G.A. (1991). A qualitative investigation into the meaning of work for the South African career woman. Unpublished D.Litt et Phil dissertation, Universit y of South Africa, Pretoria. Roesch, S.C., Christian, W. & Vaughn, A.A. (2006). Relations between the Big Five personality traits and dispositional coping in Korean Americans: Acculturation as a moderating factor. International Journal of Psychology, 41(2), 85-96. Rothman, S. (2004). Burnout, psychological strengths and coping strategies of senior managers in a manufacturing organisation. Management Dynamics, 13(4), 26-37. Rothmann, S. & Van Rensburg, P. (2002). Psychological strengths, coping and suicide ideation in the South African Police Services in the North West Province. South African Journal of Industrial Psychology, 28(3), 39-49. Ruderman, M.N., Ohlott, P.J., Panzer, K. & King, S.N. (2002). Benefits of multiple roles for managerial women. Academy of Management, 45 (2), 369-386. Saville & Holdsworth Ltd. (1999a). Occupational Personality Questionnaires course notes. Pretoria: SHL. Saville & Holdsworth Ltd. (1999b). OPQ32: Manual and user’s guide. Surrey: SHL. Saville & Holdsworth Ltd. (2004). The Occupational Personality Questionnaire in South Africa: Psychometric research perspectives. SHL. Schwartz, F.N. (1989). Management women and the new facts of life. Har vard Business Review, 67, 65-76. Sears, H.A. & Galambos, N.L. (1993). The employed mother’s well-being. In J. Frankel (Ed.), Springer series: Focus on women: Vol 14. The employed mother and the family context (pp 49 - 67). New York: Springer. Senior, L. (July 2003). Women in the workplace. Research Bulletin, 3, 4-19. Shaw, E. & Burns, A. (1993). Guilt and the working parent. Australian Journal of marriage and Family, 14 (1), 30-43. Shipley, P. & Coats, M. (1992). A community of dual-role stress and coping in working mothers. Work and Stress, 6 (1), 49- 63. Spade, J.Z. (1994). Wives’ and husbands’ perceptions of why wives work. Gender and Society, 8 (2), 170-188. Statistics South Africa. (2004). South African Statistics 2003. Statistics South Africa. Sternberg, R.J. (1994). In search of the human mind. New York: Harcourt Brace. Steyn, H.S. (1999). Praktiese beduidendheid: Die gebruik van effekgroottes. Potchefstroom: Potchefstroom University for Christian Higher Education. Storm, K. & Rothmann, S. (2003). The relationship between burnout, personality traits and coping strategies in a corporate pharmaceutical group. South African Journal of Industrial Psychology, 29 (4), 35-42. Strang, S. & Strang, P. (2001). Spiritual thoughts, coping and ‘sense of coherence’ in brain tumour patients and their spouses. Palliative Medicine, 15, 127-134. Strümpfer, D.J.W. (1990). Salutogenesis: A new paradigm. South African Journal of Psychology, 20 (4), 265-276. Sullivan, G.C. (1993). Towards clarification of convergent concepts: Sense of coherence, will to meaning, locus of control, learned helplessness and hardiness. Journal of Advanced Nursing, 18 (11), 1772-1778. Suls, J., David, J.P. & Harvey, J.H. (1996). Personality and coping: Three generations of research. Journal of Personality, 64 (4), 711-735. Visser, D. & Du Toit, J.M. (2004). Using the Occupational Personality Questionnaire (OPQ) for measuring broad traits. South African Journal of Industrial Psychology, 30(4), 65-77. Vollrath, M. (2001). Personality and Stress. Scandinavian Journal of Psychology, 42, 335-347. Vosloo, S.E. (2000). Rolkonflik by die werkende moeder – ‘n geestesgesondheidsperspektief. Unpublished D.Com dissertation, University of South Africa, Pretoria. Watkins, M. W. (2000). Monte Carlo PCA for parallel analysis [computer software]. State College, PA: Ed & Psych Associates. Welbourne, T.M. (1999). Wall Street likes women: An examination of women in the top management teams of Initial Public Offerings. Centre for Advanced Human Resource Studies. Cornell University. White, B., Cox, C. & Cooper, C. (1992). Women’s career development: A study of high flyers. Great Britain: TJ. Wissing, M.P. & Van Eeden, C. (1997, July). Facing the challenge to explicate mental health salutogenically: Sense of coherence and psychological well-being. Paper presented at the 55th Annual Convention of International Council of Psychology, Graz, Austria.