.

This research is supported by The University of Kent, United Kingdom Abstract The UK’s flexible working strategy has developed progressively since 2000, reflecting changes in the economic, political and social climate. Research has shown employees to be concerned about the effects of flexible working on career success. This paper (N=266 & N=1093) examined male, female, managerial and non-managerial evaluations of employees who either used flexible working practices, worked long hours or worked regular hours. It also compared attitudes towards employees and their perceived success as a function of the reason as to why they used flexible working practices. Results showed individuals working the longest hours were perceived to significantly outperform all those who worked regular hours or who took part in some kind of flexible working option. Women with caring responsibilities were consistently viewed as less productive and more likely to leave their employment than men. Findings on promotion show overwhelming bias towards long hours being compatible with career success. Despite the positive effects on work performance, the effects of working long hours are recognised as detrimental to work-life balance. The implications for women in the workplace and work-life balance policy are further discussed.


Introduction
The UK's flexible working strategy has developed progressively since 2000, when the former Prime Minister, Tony Blair, launched the Government's campaign to promote a better work-life balance at a business breakfast at 10 Downing Street. Key areas highlighted by the government for action by employers are: knowing that both employers and employees benefit from work-life balance; understanding that part-time does not mean half hearted; wanting employees to enjoy their full holiday entitlement; encouraging them to work from home when appropriate; enabling employees to share jobs and adopt flexible hours in school holidays; and allowing their employees pain maternity and paternity leave (DTI, 2009). The aim of this campaign is to encourage employers to introduce flexible working, with the initial emphasis on work-life balance for all employees, not just those with caring responsibilities. However, legislative change has focused on care. UK employees can request a flexible working arrangement, and those with children up to the age of 16, disabled children under 18 or those with caring responsibilities for a spouse, partner or relative have a statutory right to request flexible working arrangements and to have this request considered seriously. If the employer refuses to grant their request, they have to demonstrate there is a reasonable business case for not being able to do so.
The promotion of work-life balance in the UK reflected changes in the economic and political climate as well as social changes. Within business, globalisation and the new economy, changes have resulted in customer demands and expectations for access to goods and services twenty-four hours a day. Increasingly this means that organisations must operate outside the traditional nine to five structure. Therefore organisations have to employ people who are prepared to work flexibly outside traditional working hours. Within families the traditional model of fathers who work and mothers who remain at home to care for children and/or the elderly are decreasing. One key aspect of women's increasing economic participation has been the employment rates of women with preschool children which almost doubled from 28% in 198028% in to 53% in 199928% in (McRae, 2003. Another is the dramatic growth in the number of one-parent households from 9% in 1971 to 25% in 2001. The increase in women's work participation is predominantly reflected in part-time work; 42% of women work part-time compared to just 9% of men (LFS, 2005).
The tensions between employers' need for greater flexibility and individuals' desire for work-life balance are reflected in employees' attitudes and work behaviour. Within this context, work-family research has its conceptual roots in role conflict theory (Katz & Kahn, 1978) whereby participation in one role is made more difficult by virtue of participation in another (Dierdorff & Ellington, 2008). Employees in the United Kingdom are amongst those who work the longest hours of any European country (FedEE, 2008) with long hours particularly common amongst men who have partners and children (Houston & Waumsley, 2003;Somech & Drach-Zahavy, 2007). Examination of dual-earner couples showed that, for those working in the same organisation, men experienced higher job prestige and less turnover than women (Moen & Sweet, 2002). When examining employee turnover, Boshoff and Mels (2000) found commitment to the profession and the organisation to be important antecedents of intention to resign. In addition, Van Dick et al., (2004) used a social identity approach to show organisational identification feeds into job satisfaction, which in turn, explained turnover intentions. Key reasons for women wanting to move from working in organisations to self-employment are greater freedom, autonomy, balance and the ability to live by a personal value system threatened by the organisation (Mallon & Cohen, 2001).
Dual earner couples have become the norm, but women's participation in, and commitment to, the workforce continues to be limited by the presence and age of a dependent child. The birth of children continues to perpetuate traditional divisions of work and caring roles in most couples, despite an expressed desire to share these roles more equally (Houston & Marks, 2002Houston & Waumsley, 2003). Most women with children move into part-time work, but part-time working and career breaks negatively impact upon women's lifetime earnings and the gender pay gap (Anderson, Forth, Metcalf & Kirby, 2001;Manning & Petrongolo, 2004;Walby & Olsen, 2002). The employment disadvantage experienced by many female part-time workers is illustrated in findings from a survey of 643 qualified National Health Service nurses (Lane, 2004) which shows the career progression of part-time nurses continues to lag behind that of their full-time counterparts. A key reason for this is the low status of part-time work. Contrary to the belief that part-time workers are uncommitted to their careers, prioritising family responsibilities over work, they reported high levels of under achievement. Managers report no formal policy not to employ part-time nurses in higher grades but usually the higher grades are only advertised as full-time posts (Lane, 2004).
In contrast to seeking to understand the disadvantages of women in the workplace, recent research highlights the possibilities of women being able to balance family and career success (Ezzedeen & Ritchey, 2009;Jones & Gittins, 2009). Within this remit, evidence was still found that women face certain obstacles when trying to reach the top in organisations, exacerbated in male-dominated cultures (Jones & Gittins, 2009). Women's success appeared to be founded on the right mental approach to balancing family and work, a choice many men do not have to face (Ezzedeen & Ritchey, 2009;Jones & Gittins, 2009). In addition, successful women were prepared to work long hours if necessary (Jones & Gittins, 2009). Social support was found to be an important resource in women's advancement and well-being (Ezzedeen & Ritchey, 2009) illustrating women as more communal (i.e. showing warmth and selflessness) than agentic (i.e. showing assertiveness and instrumentality) and men more agentic than communal. Outsourcing domestic responsibilities was also found to be common amongst successful career women. Thus, where career and mothering roles are maintained, domestic responsibility is subcontracted (Ezzedeen & Ritchey, 2009).
In contrast, however, strategies for coping with work-family conflict further highlighted women's struggle to combine work with family responsibilities, with results showing greater work family conflict for those women whose commitment to work was hindered by family demands (Somech & Drach-Zahavy, 2007). These findings suggest little change within the last decade in terms of women trying to combine work and family life (Duxbury, Higgins & Lee, 1994;Somech & Drach-Zahavy, 2007). This is further echoed in the findings of Linehan and Walsh (2000;2001) who suggest work-family conflict prevents many female managers from progressing to senior management because they typically have to choose between a career and family. More female than male managers have to decide between an international career and family commitments because of the relative inflexibility of organisations, the assumptions by senior management and society that the primary role of a woman is that of mother and for responsibility for home and family and not as an international manager. Career success is still based on a male career model which ignores the influence of marriage, pregnancy, children and household duties. Moreover, if women do achieve leadership roles, they are more likely to receive greater scrutiny and criticism than men and are evaluated less positively (Ryan & Haslam, 2005). Further, the leadership positions that women occupy are likely to be less promising than those of their male counterparts, an effect known as "the glass cliff effect" (Ryan & Haslam, 2005, p.87) Whilst flexible working policies were meant to provide employee choice and release employee pressure from long working hours, tensions are further reflected in the ways in which employees demonstrate conflicting attitudes towards flexible working. In a survey of over 1500 members of the Amalgamated Engineering and Electrical Union, Houston and Waumsley (2003) found that while the majority of men and women in this study thought that flexible working was beneficial to both employers and employees, with women more interested in uptake than men, there was a high level of concern that flexible working would damage their promotion prospects and their relationship with their colleagues. Furthermore, fifty per cent of respondents agreed that to be viewed favourably by management they have to put their jobs ahead of their personal life.
The Houston and Waumsley (2003) findings are in many ways consistent with those from The Third Work-Life Balance Employee Survey (Hooker, Neathey, Casebourne & Munro, 2007). While this report presents a very positive perspective on flexible working, thirty-two per cent of respondents agreed that people who work flexibly are less likely to get promoted and forty-two per cent felt that leaving work on time would negatively affect their career, as would taking leave to look after children or other dependants (37 per cent), working different working patterns (37 per cent), or working from home (25 per cent).
These recent findings demonstrate that there are still significant tensions between a desire for work-life balance and career achievement, despite legislative and culture change. This remains consistent with research from the 1990s. Kossek & Ozeki (1999) conducted a review of studies that examined the effects of family-friendly policies on a variety of outcomes. They concluded that work-family policies result in higher levels of actual individual productivity and positive attitudes to the employer/organisation. However, they also found that perceptions of turnover and organisational commitment were more strongly linked to traditional notions of long hours demonstrating commitment.
Government initiatives designed to change both legislation and work place culture are clearly limited by the culture within a specific workplace and the management of flexible working. Whilst improvements in turnover, emotional well-being and commitment have been shown by making flexible working arrangements widely available for part-time employees who are only able to invest limited time in their careers (Baruch & Winkelmann-Gleed, 2002) and while the availability of flexible working practices is an essential precondition to their uptake by employees, this does not necessarily lead to uptake. Thompson, Beauvais, and Lyness (1999) and Allen (2001) found that both perceived supervisor support and perceptions of family support within the organisation were positively related to overall use of flexible working arrangements. Thompson et al., (1999) provided evidence to suggest that the type and number of work-family programmes offered is not as important as the culture of an organisation, which, in itself, is crucial for determining not only whether people will use benefits, but also their general attitudes towards the organisation (Galinsky & Stein, 1990;Lewis & Taylor, 1996;Thompson et al., 1999). Bond and Wise (2003) found that the management of family leave policies was one of the increasing number of HR practises that are devolved to line managers and that managerial discretion was often a critical part of the formal provision within an organisation. However, they also found that most operated with little training or liaison with HR specialists and that there was a low level of awareness of recent statutory changes in relation to parental leave. Decisions about the implementation of flexible working have considerable impact on individual managers. Powell and Mainiero (1999) found that flexible working arrangements make managers' jobs more complex and difficult by placing demands that are over and above traditional supervisory demands. In the context of research evidence that shows that managers are primarily rewarded for the results of their work, rather than for any concern shown to their employees, they argued that it was not surprising that managers tend to focus on their own short-term best interests when making decisions about the implementation of the organisation's flexible working policies. More recently, and in keeping with differences in attitudes between managers and employees, Turnipseed and Rassuli (2005) found that managers place more emphasis on measured performance whilst 'helping others in the workplace' is seen as a significant contributor to performance by employees.
In the following two studies we investigated the effects of flexible working patterns on a number of aspects of work performance. A series of vignettes were created describing men and women who worked different patterns of hours for a fictitious 'successful' company. No information was given about their work performance, only about their working patterns and reasons for using these patterns. With the exception of the long hours worker, all worked the same number of actual hours. Participants were asked to evaluate a selection of employees. The aim was to control the type and amount of information available to participants in order to gain a clearer sense of how employees are perceived as a result of differing work patterns and interests/responsibilities outside work.
In the context of the literature described above, the first hypothesis was that individuals who worked long hours would be perceived to have better work performance when compared to those who either worked regular hours or who worked flexible hours. Following Hogarth, Hasluck, Pierre, Winterbotham, and Vivan's (2000) findings that a substantial number of employers felt flexible working practices were unfair to some staff, the second hypothesis predicted that those who worked regular hours would be perceived with greater approval as work colleagues than those who worked long hours or some kind of flexible working option. In the context of the mixed results of the review by Kossek and Ozeki (1999), the relationship between working patterns and perceptions of turnover intention was also investigated. In addition, it was considered that there might be a hierarchy of justifications for working flexibly and to this end, the study also examined relative evaluations in the context of reasons for working flexibly. Finally, it was also expected that the raters would recognise the negative effects of working long hours and rate these employees as having low levels of work-life balance. As flexible working has a higher uptake and is perceived more positively by women (Houston & Waumsley, 2003), the effects of gender of both participant and target on the evaluations were also examined.

Participants
Participants were 266 final year undergraduate students. The sample consisted of 97 males and 169 females, with a mean age of 20 years. Average work experience was between one and two years of part-time work.

Materials
A booklet was designed which contained a description of a 'successful' UK company and descriptions of six employees within the organisation. Each employee was described as having a different working pattern: long hours; working 9-5, flexible hours to look after a child, flexible hours to look after an elderly parent, flexible hours to play sport, and flexible hours to play music. Information was only given about their working patterns and their hobbies and commitments outside work, no information was given about their work performance and all employees were described as having outside interests regardless of working pattern. With the exception of the long hours worker, the regular hours worker and all the flexible workers had the same number of actual hours.
Participants were asked to rate each employee on a series of 7-point Likert scales designed to measure work performance, turnover intention, colleagality, and work-life conflict. These scales were based on items employed in previous research (Houston & Marks, 2003;Houston & Waumsley, 2003).
Turnover Intention was measured using 2 items, (e.g. Jane will leave the company within the next two years), Cronbach's Alpha = .75.
Work-Life Balance was measured using 3 items (e.g. Sarah is able to balance work and life outside work) Cronbach's Alpha = .73.

Procedure
Participants took part in the study in a large group in a lecture theatre. All were asked to read the description of the organisation and of all six employees and then to rate each employee. In order that any perceived differences by male and female participants toward male and female targets using different working styles could be measured, two questionnaires were used, counter-balancing gender within scenarios.

Results
The data were analysed using a Participant Gender (Gender) x Target Gender (Target) x Work Pattern (Pattern: Regular, Long Hours, Child Care, Elder Care, Hockey, Music) MANOVA on the four measures relating to Performance, Turnover, Colleague Approval and Work Life Balance. Target was a between participants factors and Pattern was a within-participants variable. As shown in Table 1, there were no significant multivariate main effects or interactions among Gender and Target. However, the multivariate effects of Pattern, Pattern x Target and Pattern x Target x Gender were significant. We therefore turn to examine the significant univariate effects. Table 1 about here Table 2 shows the means and standard deviations as well as significant pairwise differences for the main effects of Pattern, which were significant (p <.001) for all four measures, and showed large effects in the case of both Performance and Work Life Balance. Table 2 about For performance, participants believed that people who work long or regular hours outperform all those who use flexible working, and that those using flexible working for elder care perform the least well. Specifically, the Long Hours worker is viewed as performing significantly better than all others (all p <.001), and the Regular worker is viewed as performing significantly better than all remaining others (all p <.001). The Musician is viewed as performing significantly better than all remaining others (p's all <.05). The Child Carer and Hockey player do not differ significantly from one another, but the Elder Carer is viewed as performing significantly worse than the Child Carer (p < .005).
For turnover intention participants believed that people who used flexible working for elder care, sport or musical pursuits were more likely to leave their organisation than people who worked long or regular hours or used flexible working for child care reasons. Pairwise differences within these two subsets were all non significant, whereas pairwise differences between the subsets were all highly significant (all p <.001).
For approval from colleagues, pairwise differences among patterns revealed that regular working was regarded as significantly more collegial than all other patterns, including long hours working (p<.02), child care (p<.01) and music (p <.001). All of these were judged to attract more approval than hockey and elder care which also did not differ from one another.
Participants believed that people who worked long hours or who used flexible working for elder care or child care were viewed as having less good work life balance than those who worked regular hours or worked flexibly to pursue sport or musical interests. Pairwise comparisons reveal that the Long Hours pattern was regarded as having significantly less balance than all others (all p <.001), and the Elder Care worker was perceived as having significantly less balance than all remaining others (all p <.001). The Child Carer was perceived as having significantly less balance than all remaining others (all p <.001), whereas the Regular worker, Hockey Player and Musician were seen as relatively more balanced and did not differ significantly from one another.

Interactions Involving Participant Gender and Target Gender
Performance, Turnover and Colleagues' approval were all affected by a Pattern x Target interaction. For all three variables there was a significant difference in judgments about male versus female targets who worked flexibly to engage in childcare.
For performance, F(1,262) = 8.99, p <.005, partial 2 = .033, male targets who worked flexibly to do child care (M = 4.45, SD = 0.64) were judged to perform better than female targets who did so (M = 4.22, SD = 0.69). For turnover, a similar effect appeared F(1,262) = 9.50, p <.005, partial 2 = .035, male targets who worked flexibly to do child care (M = 3.25, SD = 0.90) were judged to be less likely to leave than female targets who did so (M = 3.53, SD = 0.95). For approval from colleagues, F(1,262) = 6.95, p <.01, partial 2 = .026, male targets who worked flexibly to do child care (M = 4.98, SD = 0.84) were judged to attract greater approval than female targets who did so (M = 4.74, SD = 0.79). Thus male workers who work flexibly to look after children are perceived to be more productive, less likely to leave and more likely to be approved by colleagues than are female workers in an identical situation.
Judgements about turnover and colleagues' approval were further affected by a Pattern x Gender x Target interaction. These interactions indicate that Gender and Target interact to affect judgments for some patterns but not others. Therefore we investigated the simple interactions involving Gender and Target within each Pattern.
For colleague approval none of the simple interaction effects were significant, but some were marginally significant. Examination of the simple main effects of Gender and of Target within each pattern revealed that males judged the approval of a flexible worker looking after children differently if the worker was male (M = 5.03, SD = 0.92) versus female (M= 4.67, SD = -.70), F (1,262) = 6.94, p <.01, partial 2 = .025. In addition, a marginal Gender x Target interaction on the Long Hours pattern, F (1,262) = 3.62, p <.06, partial 2 = .014, was due to the fact that females working long hours were judged more positively by males (M = 5.14, SD = 1.28) than by females (M = 4.72, SD = 1.02), F (1,262) = 4.82, p <.05, partial 2 = .018 whereas males working long hours were judged similarly by both genders (M = 4.74, SD = 0.95, M = 4.83, SD = 1.01, judged by males and females, respectively).

Participants
Four thousand questionnaires were posted to the home addresses of members of a large UK trade union. Cooperation with the trade union made it possible to send equal numbers to men and women and to managerial and non managerial employees. Participants were asked to complete the questionnaire and return it in a post-paid envelope. A thirty percent response rate was achieved. Once partially completed questionnaires were eliminated from the sample, 281 male managers, 375 female managers, 210 male non-managers and 227 female non-managers responses were analysed.

Materials
Participants were given the same booklet used in Study One, containing descriptions of a UK company and its employees and were asked to rate the employees on the same dimensions of work performance, turnover intention and colleague's approval. In addition participants in Study Two were asked to rate the employees on an established measure of work-life and life-work conflict and to evaluate promotion prospects.
Work-life and Life-work Conflict were each measured using 2 items selected from Netemeyer, Boles, & McMurrian (1996) family-work conflict scale in order to represent 'life' rather than specific family items (e.g. Matt's job produces strain that makes it difficult to fulfil home responsibilities and the demands of Kate's home life interfere with her responsibilities at work).
Promotion prospects were measured by asking participants to state which employee would be most likely to be promoted and which would be the least.

Procedure
In order that any perceived differences by male and female participants toward male and female targets using different working styles and any differences between work categories could be measured, four questionnaires were used, crossing gender and work category within scenarios. Questionnaires differed only in that questionnaire one depicted male managerial employees, questionnaire two depicted female managerial employees, questionnaire three depicted male administrative employees, and questionnaire four depicted female administrative employees. Male and female managerial participants were asked to complete the male and female managerial employee questionnaire. Male and female non-managerial participants were asked to complete the male and female administrative employee questionnaire.

Results
The data were analysed using a Participant Gender (Gender) x Target Gender (Target) x Role Status (Managerial vs Non Managerial) x Pattern ( Regular, Long Hours, Child Care, Elder Care, Hockey, Music) MANOVA on the five measures relating to Performance, Turnover, Colleague Approval, Work Life Conflict and Life-Work Conflict. Gender, Target and Role were between participants factors and Pattern was a within-participants variable. As shown in Table 3, there were small but significant multivariate main effects of Gender, Target and Role, but no significant multivariate interactions among these variables. Table 3 also shows a very large significant multivariate effect of Pattern, and significant multivariate interactions between Pattern x Gender, Pattern x Target and Pattern x Role.  Table 3 about here The relevant means and standard deviations, as well as significant pairwise differences, are provided in Table 4. None of the three or four way interactions were significant.
What is immediately striking from Table 3 is that the large multivariate effect of Pattern is attributable mainly on the performance, work life and life work measures. For the first two of these, all pairwise differences were significant, mostly p <.001. For the last, only Hockey and Music did not differ significantly. The highest performance is attributed to long hours and regular workers, and similarly these have the lowest life-work conflict scores.
The interaction between pattern and participant gender also showed a consistent picture. Males believed that all of the flexible working arrangements would result in worse performance than did females. Males, more than females, also believed that long hours and regular working were associated with lower turnover intention, but that child and elder care were associated with higher turnover intention. Males rated work-life and life-work conflicts as higher than females for all patterns.
The interaction between pattern and target gender showed a slightly smaller multivariate effect size. However, this showed that flexible workers doing child or elder care were seen as more likely to leave the organisation if they were female than male.
The interaction between role and pattern was also quite clear and was significant for all patterns. Managers believed long hours and regular working affected performance and turnover intention less positively than did non-managers. Managers believed flexible working affected performance and turnover intention more positively than did non managers. Managers also perceived less collegial approval of longer hours and regular working, but also of flexible working for child care than did non managers. However managers thought there would be more approval for flexible working to play music than did non managers. Managers rated higher work-life scores for long hours working and lower scores for all forms of flexible working than did non-managers. Managers also rated lower life-work scores for all forms of flexible working than did non-managers.

Recommendation for Promotion
The proportion of participants who recommended promotion for each target varied significantly, Chi-square (5) = 684.118, p <.001, eta squared = .511. Pairwise comparisons revealed highly significant differences (p's <.001) between recommendations between all patterns in the following order of preference: Long hours working (42.6%), regular working (25.6%), flexible working with children (16.1%) flexible work for music (7.3%), and both flexible working for hockey (4.7%) or elder care (3.8%). The latter two did not differ significantly.
Of further interest is whether these patterns varied as a function of Gender, Target or Role. Given the absence of two and three way interactions in the preceding analysis it was not surprising that hierarchical loglinear analyses revealed no significant higher order interactions between promotion recommendations for each type of pattern and these three variables. However there were significant simple associations between each variable and promotion recommendation for each pattern.
The results of Chi-Square tests comparing the proportions that would recommend promotion for each pattern are provided in Table 5. Table 5  This shows that males are more likely to promote longer hours workers and less likely to promote those who work flexibly to care for children. Participants are more likely to promote male targets than female targets who work flexibly to care for children. Managerial participants are less likely than non managerial participants to recommend promotion in recognition of long hours working, and more likely to recommend promotion for someone working flexibly to care for children, or playing sport.

Discussion
Across both the undergraduate and working samples there is clear evidence that flexible working is seen as detrimental to work performance and career progression when compared to long or regular hours of work. There is also clear acknowledgement of the work-life benefits of flexible working. Both the undergraduate and the working samples show strong indication that some reasons for working flexibly are more 'acceptable' than others; where hobbies are driving flexible working patterns this is taken as a sign of increased turnover intention.
The effect of gender of target and gender of participant produced some perhaps surprising results. In the undergraduate sample, men who were working flexibly in order to look after children were rated more positively on work performance and colleagues' approval, as well as being seen as less likely to want to leave the organisation, than women who were doing the same thing. Women working long hours were seen as having greater colleague approval by males than females, whereas men working long hours were rated similarly by both genders. Gender differences in attitudes to working hours are also reflected in the findings. Female participants were more likely than male participants to view a woman working long hours as likely to leave, whereas male participants were more likely to think that women working regular hours, or flexible working to cover childcare would leave. From this it can be concluded that young men believe that women are more likely to leave their jobs than men, unless they are demonstrating high work commitment through long hours.
There were also important findings in relation to gender in the working sample. Male participants indicated that all of the flexible working arrangements would result in worse performance than did female participants. Male participants also perceived long hours and regular working to be associated with lower turnover intention, and child and elder care to be associated with higher turnover intention, more than did female participants. Male participants also perceived both work-life and life-work conflicts to be higher than did females for all patterns. These findings concur with those of Moen and Sweet (2002) who found men experienced less turnover than their female partners working in the same organisation. In addition, perceptions of women struggling to combine work and childcare is in keeping with the findings of Linehan and Walsh (2000;2001), Ryan and Haslam (2005) and Somech and Drach-Zahavy (2007).
In general managers presented more positive views about flexible working than non managers. It is not clear whether this is due to positive experiences of managing flexible working or a greater understanding of the rationale behind work-life balance policies.
The findings on promotion show an overwhelming bias toward long hours and traditional working patterns, as well as strong gender differences. Male participants were more likely to promote long hours workers and less likely to promote those doing flexible working in order to cover childcare. However male targets using flexible working for childcare were more likely to be nominated for promotion than females using the same pattern. These findings support those of Houston and Waumsley (2003), Lane (2004) and Hooker et al., (2007) where a high level of concern was evident over flexible working damaging promotion prospects. However, the current findings add to these the negative demise of women struggling with child care and career commitments.
Overall our findings are consistent with recent research by Heilman and Okimoto, (2008) which also asked undergraduate students and working adults to evaluate hypothetical workers described in vignettes. They found that anticipated job commitment and anticipated achievement striving were negatively affected by parental status, in both men and women. Both mothers and fathers were expected to be less dependable than non parents and mothers in particular were perceived to be less competent in work related behaviour.
Of interest in our study and that of Heilman and Okimoto (2008) are the undergraduate attitudes, which are not coloured by current or specific organisational culture but preconceptions that are likely to be carried forward into the workplace. These are not only likely to perpetuate the long hours culture in the UK but to be of continuing detriment to women. Whilst general attitudes towards an organisational culture have been shown to be important in terms of uptake of flexible working options (Galinsky & Stein, 1990;Lewis & Taylor, 1996;Thompson et al., 1999) this study has highlighted that traditional views about women, work performance and working patterns still persist. Successful career women, as demonstrated in recent studies by Jones and Gittins (2009) and Ezzedeen and Ritchey (2009) suggest this success is achieved by long hours work when necessary.
Our findings indicate that flexible working practices will not necessarily change attitudes towards the patterns of work currently found in the UK, despite the governments initiate on work-life balance. Since the majority of employees using flexible working practices are women with childcare responsibilities there is a danger of exacerbating a two-tier workforce and further widening the gender gap in the workplace: those who use flexible working practices (mainly women) and whose careers will not progress, and those who work long or regular hours, and who are professionally successful. Furthermore, the introduction of the statutory right to request flexible working hours from their employer for those with childcare responsibilities may only serve to widen the career gap between those who use flexible working (mostly women with childcare responsibilities) and others in the workplace.
Our study shows that working long hours is still viewed as one of the key ways of obtaining career success and flexible working hours are seen as detrimental to work performance and career progression. HR strategy should be to examine the impact of flexible working on individual performance, productivity and success. For many, work-life balance will only become a realistic option when we can demonstrate that working styles and patterns are unrelated to success, or, that flexibility is actually associated with these outcomes. Currently, it appears that flexible working, being female and having a successful career are still perceived by many to be incompatible.

Limitations and Future Directions
Our findings must be evaluated in light of limitations with regard to the generalisability of the vignette design. Hypothetical scenarios do not offer 'real' conditions and can, thus, skew attitudes and perceptions that may otherwise be different given inclusive circumstances of individual employees. The information provided about the target was purposely sparse and uncomplicated to enable us to clearly and precisely manipulate our independent variables. Conversely, though, this sparseness may have facilitated the use of stereotypes because of the absence of individuating information. Moreover, our studies all used a paper-and-pencil format, did not necessitate actual interaction between the subjects and those whom they were rating, and sometimes used single-item measures. These, too, are potential limitations of the research. It may be argued that the nature of the undergraduates work experience might limit the generalisations that we can make from the study 1 data. However, we put forward the argument that the attitudes shown by our results offer strong evidence that preconceptions about females in the workplace antedate graduates moving into full time employment. Given that similar attitudes were found in a 'real' workplace setting, it would also appear that pre-conceived workplace attitudes held by undergraduates, and attitudes held by more experienced employees in the workplace, are contrary to UK government initiatives.
All this strongly suggests that our concepts should be tested in ongoing workplace settings, perhaps by obtaining attitude ratings from males and females in different types of job categories and industries. Future research is needed to identify whether there are conditions under which flexible working can be undertaken without detriment to professional success, where professional success can be achieved without resorting to long hours of work, and where work-life balance can be obtained without compromise to either.   Note SDs are italicised. Pairwise comparisons between levels of a between participants factor: +.05>.001,*p<.001.