It is well documented that there is a problem recruiting and retaining women in technology. People talk about a 'leaky pipeline' with girls and women being lost at all stages - the proportion of females gradually reducing through school, university and during work. This is a problem for a country committed to social justice and fair access to opportunities, and for the technology industry, as it is not capitalising on the talents of women.
We already know a fair bit about what's going on and why this happens. First, women don't get the same levels of early exposure to computers that boys do. Computer games, which are a key route into developing an interest in the field, are developed by, aimed at and played by men and boys. Boys too seem to get more access to computers at school and at home. Second, there is a masculine culture associated with technology-related environments. The stereotypes (which do, to an extent reflect the reality) indicate that technology is a male dominated field, filled with coding-obsessed geeks, where women just don't belong. This is off putting to many women both as they are contemplating their own future careers, and when they actually start their first jobs. Finally, it's confidence. Despite the reality that women and girls are perfectly adept at coding and well able to hold their own academically in this field, women believe that they are less capable than men in this arena.
A lot of research has been done in this field, but despite all of this understanding and the vast sums of money that have been poured into initiatives to change things, there has been little shift in the proportion of women choosing and sticking in technology. Clearly we are missing something.
My colleague Anke Plagnol and I decided to take another look, a deeper one, to explore the experiences of female computer science students studying in the UK. We found 200 students (male and female) to fill in a survey, and then followed up with in-depth interviews with 20 female students.
Some of what we found echoes findings we have seen elsewhere: the women have lower confidence than the men; they feel that they are less likely to fit in within the industry, and this seems to make them feel less sure about their career plans. But we were really struck by the extent and strength of the messages that these women had been getting from all angles, all their lives, reinforcing the idea that they are just not cut out for technology. The participants had been told by their parents, their school teachers, university professors, employers, colleagues, fellow students and even each other, that women's brains just don't work in the right way to succeed in this field. Every one of the women we interviewed had experienced and witnessed multiple incidents where their achievements were undervalued ('you only got that job because you are a girl'; 'they let girls into these courses with lower grades'); their ability questioned ('you won't be able to manage this'; 'perhaps you could ask your [male] colleague to help?') or their ideas dismissed. As one of the women summed up 'ultimately, they just think you're dumb'.
No wonder this relentless barrage takes its toll on women's confidence and choices.
The other really interesting thing to emerge from the interviews was that there might be a problem (for some women) with the way that computer science is taught in universities. People, in general, are split into those who are motivated by agency (which means they like to learn independently aiming to master a topic on their own, and are stimulated by competition), or commune (which means they prefer to work collaboratively, learning through being taught and supported by a tutor). In general, men are more likely to be motivated by agency and women by commune. But the computer science classroom is highly agentic. Students are expected to learn things on their own, and are not encouraged to team up, to help each other understand the more complicated ideas, or ask for help from their tutors. Departments too encourage students by setting up competitions. This approach just doesn't suit communal learners (mostly although not exclusively women) who enjoy and gain confidence from working collaboratively, and who need more input from their tutors to feel sure that they are doing the right thing. And the agentic teaching style adopted in computer science classrooms seems to make women feel less enthusiastic about their subject and less confident about their ability to succeed.
But where should this lead us? You might argue that we simply need to make the classrooms more communal, offering more support, and more opportunities for genuine collaboration. But it's not quite as simple as that. Technology evolves at such a pace that the industry needs its workers to be self-directed learners. Making the classrooms less agentic could actually end up doing the students and the industry a disservice.
But I wonder if it might be possible to find some kind of compromise? One in which students are taught to be self-directive, rather than tutors assuming from the star that they are comfortable with this approach to learning. Students could be given more support at the start, but encouraged to own their own learning, and taught strategies for working things out by themselves. And surely developing skills in collaborative working would be valuable for the industry? So perhaps close collaborations (not just group projects where everyone takes on their own tasks) could be encouraged alongside some of the existing focus on individual competition.
This suggestion that the pedagogical culture might not suit women as well as men is, I think, a new one and I think is definitely worth pursuing. It would be interesting to see if we could measure the degree of agency within a computer science classroom, and to find out whether we can generalise these findings - is it generally true that computer science classes in fact serving the agentic learners better? And how easy would it be to teach people who are naturally communal learners to be more confident about learning independently?
Sunday, 30 June 2019
Saturday, 22 June 2019
Relational Frame Theory: a simple summary
I've become a bit interested in Acceptance and Commitment Therapy (ACT) over the last few months, as I think it might have a lot to offer the careers world. Underpinning ACT is a theory by Steven Hayes, called Relational Frame Theory (RFT). I don't know how vital it is to understand RFT in order to make good use of ACT, but I thought I would have a go.
RFT is a cognitive behavioural theory, which means that it is interested in the links between thoughts and behaviour. RFT has a reputation for being terribly difficult to understand, but I think that at least some of its key tenets are actually quite intuitive. One challenge is that the literature about RFT uses complex, technical and inaccessible language. It describes itself as an account of arbitrary derived relational responses non-arbitrarily applied. Well, that's clear, isn't it?
Let's start with the idea of relational responses. This is really based on the old notion of associations, but it takes it a bit further. So, an association is when you have an experience which links two things in your mind. At a basic level, you can imagine that the first time you tasted apple pie, you thought it tasted great. You thus established a link between apple pie and deliciousness. But you can go further. Perhaps you spent time during your childhood visiting your Granny in Woking and eating her fabulous apple pie; this may have established a link in your mind, between Woking and the loveliness of apple pie. And even though the chain of associations goes: Woking...Granny...Apple pie...delicious taste, the link between Woking and delicious taste becomes direct for you; Woking forever becomes a place which warms your heart. And you remember about reinforcement? The more times you had delicious apple pie in Woking, the stronger and more enduring that link between Woking and happy times becomes.
RFT talks about arbitrary relational responses, which are associations (relational responses) which are not linked to physical properties (ie are arbitrary). This can be illustrated with language as an example: there is 'cat' the animal, 'cat' the collection of sounds that we can produce and hear as the word 'cat', and the written word 'cat', spelled C-A-T. These are three separate things, and the links between them are entirely arbitrary. There is no God-given reason why the letter C should be pronounced the way we say it, or that the animal should be given the name 'cat' - the links are arbitrary, human inventions. But for us, the links have been very well reinforced, and it's almost hard for us to remember that the animal, the written word and the spoken word are three separate things.
So the idea of relational responses is about two or more things that in your mind are related to each other. The two things linked together are known as a frame, and the relationship between them is defined in different ways, for example, the two things might be considered as equivalent, one might contain the other or one might cause the other.
The idea of derived relational responses means that you can make links between things you haven't actually experienced. For example, you might have been told about a link between two things, and even if you hadn't seen the link for yourselves, you could still internalise the association and make one of your own relational frames. For example, if you are told that tigers are dangerous because they bite, you might be afraid of a tiger, even if you have never actually seen a tiger bite. You have thus established an association (a relational frame) between tiger and fear, even though you had never seen a tiger do anything to frighten anyone.
As well as these vicarious associations, which you can learn from others. you can also create associations between things based on two different sets of associations. So, if you went to a job interview in Tooting, and the interview went badly, you might develop two associations: job interviews make me feel bad; and; job interviews happen in Tooting, and this could lead you to believing that therefore Tooting makes me feel bad.
This could then become even more obscure as you might then establish a close link in your mind between Balham and Tooting (two stations next to each other on the tube line) and this then extends the relational frame in this way: job interviews make me feel bad; job interviews happen in Tooting; Tooting is very similar to Balham and therefore Balham makes me feel bad. This is called transformation of stimulus which refers to the process by which one association (Tooting and feeling bad) is applied elsewhere (Balham is similar to Tooting, so Balham also makes me feel bad).
RFT goes into a fair bit of detail about the different kinds of relationships that can be established between things, and talks about how they are built up. I'm not sure that level of detail is important for those engaging with RFT in order to understand ACT, but it's useful to know they are part of the overall RFT theory.
One final important aspect of RFT is the idea of cognitive fusion in which thoughts get tied up (fused with) reality and you start believing that your thoughts are in fact literally the same as the truth - the idea that thinking the thought or using the words 'I'm rubbish at interviews' leads you to believe that this is a true, literal fact. One thing that ACT is really good at, is making people understand that thoughts and words aren't the same as facts: the words are just words, the thoughts are just thoughts - and both these are distinct from the reality of your interview performance.
References
Blackledge, J. T. (2003). An introduction to relational frame theory: Basics and applications. The Behavior Analyst Today, 3(4), 421.
RFT is a cognitive behavioural theory, which means that it is interested in the links between thoughts and behaviour. RFT has a reputation for being terribly difficult to understand, but I think that at least some of its key tenets are actually quite intuitive. One challenge is that the literature about RFT uses complex, technical and inaccessible language. It describes itself as an account of arbitrary derived relational responses non-arbitrarily applied. Well, that's clear, isn't it?
Let's start with the idea of relational responses. This is really based on the old notion of associations, but it takes it a bit further. So, an association is when you have an experience which links two things in your mind. At a basic level, you can imagine that the first time you tasted apple pie, you thought it tasted great. You thus established a link between apple pie and deliciousness. But you can go further. Perhaps you spent time during your childhood visiting your Granny in Woking and eating her fabulous apple pie; this may have established a link in your mind, between Woking and the loveliness of apple pie. And even though the chain of associations goes: Woking...Granny...Apple pie...delicious taste, the link between Woking and delicious taste becomes direct for you; Woking forever becomes a place which warms your heart. And you remember about reinforcement? The more times you had delicious apple pie in Woking, the stronger and more enduring that link between Woking and happy times becomes.
RFT talks about arbitrary relational responses, which are associations (relational responses) which are not linked to physical properties (ie are arbitrary). This can be illustrated with language as an example: there is 'cat' the animal, 'cat' the collection of sounds that we can produce and hear as the word 'cat', and the written word 'cat', spelled C-A-T. These are three separate things, and the links between them are entirely arbitrary. There is no God-given reason why the letter C should be pronounced the way we say it, or that the animal should be given the name 'cat' - the links are arbitrary, human inventions. But for us, the links have been very well reinforced, and it's almost hard for us to remember that the animal, the written word and the spoken word are three separate things.
So the idea of relational responses is about two or more things that in your mind are related to each other. The two things linked together are known as a frame, and the relationship between them is defined in different ways, for example, the two things might be considered as equivalent, one might contain the other or one might cause the other.
The idea of derived relational responses means that you can make links between things you haven't actually experienced. For example, you might have been told about a link between two things, and even if you hadn't seen the link for yourselves, you could still internalise the association and make one of your own relational frames. For example, if you are told that tigers are dangerous because they bite, you might be afraid of a tiger, even if you have never actually seen a tiger bite. You have thus established an association (a relational frame) between tiger and fear, even though you had never seen a tiger do anything to frighten anyone.
As well as these vicarious associations, which you can learn from others. you can also create associations between things based on two different sets of associations. So, if you went to a job interview in Tooting, and the interview went badly, you might develop two associations: job interviews make me feel bad; and; job interviews happen in Tooting, and this could lead you to believing that therefore Tooting makes me feel bad.
This could then become even more obscure as you might then establish a close link in your mind between Balham and Tooting (two stations next to each other on the tube line) and this then extends the relational frame in this way: job interviews make me feel bad; job interviews happen in Tooting; Tooting is very similar to Balham and therefore Balham makes me feel bad. This is called transformation of stimulus which refers to the process by which one association (Tooting and feeling bad) is applied elsewhere (Balham is similar to Tooting, so Balham also makes me feel bad).
RFT goes into a fair bit of detail about the different kinds of relationships that can be established between things, and talks about how they are built up. I'm not sure that level of detail is important for those engaging with RFT in order to understand ACT, but it's useful to know they are part of the overall RFT theory.
One final important aspect of RFT is the idea of cognitive fusion in which thoughts get tied up (fused with) reality and you start believing that your thoughts are in fact literally the same as the truth - the idea that thinking the thought or using the words 'I'm rubbish at interviews' leads you to believe that this is a true, literal fact. One thing that ACT is really good at, is making people understand that thoughts and words aren't the same as facts: the words are just words, the thoughts are just thoughts - and both these are distinct from the reality of your interview performance.
References
Blackledge, J. T. (2003). An introduction to relational frame theory: Basics and applications. The Behavior Analyst Today, 3(4), 421.
Tuesday, 18 June 2019
Where do clients get stuck? Theories of career decision making difficulties
I spend much of my working life thinking about the training of professional career practitioners. I base the content of my sessions, books or articles on the things that I think career practitioners will find useful, but as someone who claims to be fully committed to strengthening the evidence base underpinning our work, I wonder if I need a stronger theoretical basis to help me work out what topics to cover?
One useful starting point would be to find out what career dilemmas clients face. If we have a clear understanding of the career decision making difficulties career clients bring to their careers interviews, then we can start to ensure that the practitioners are equipped with the right range of techniques to help them.
This field is fairly under-researched (Kelly & Lee, 2002), with no real agreed definition of career decision making difficulties, and limited inclusion of the construct in career theories. The literature is also a bit confusing (ha! there's a surprise) but are a few different models I have come across.
Broadly speaking, it seems that there are two categories of career decision making difficulties: cognitive issues and affect issues (ie what is going wrong, and how people are feeling) (Chartrand, Robbins, Morrill, & Boggs, 1990). The bulk of the literature seems to focus on the cognitive issues, and aims to find scales which practitioners can use to help work out where their clients are struggling. This all feels very positivist to me and I think that perhaps career issues are a bit too complex to narrow down in this way. But I can see that it might form a useful start to a conversation.
1. Career Decision Making Difficulties (Gati et al., 1996)
One of the most widely cited (cognitive) frameworks of Career Decision Making Difficulties was written by Gati et al., in 1996. (Gati's name is all over the place within the career decision making difficulties literature). Gati identified three overarching difficulties, and separated that into 10 different strands:
2. Career Thought Index (Sampson et al., 1996)
Gati's framework overlaps a fair bit with another one called the Career Thought Index (Sampson et al., 1998) which is linked to the Career Information Processing theory which you may have come across before. This one is also divided into three areas:
These two approaches (Gati's taxonomy and Sampson's Career Thought Index) were compared and Kleinman et al. (2004) found that they basically measured the same thing, and suggested that it didn't much matter which one people chose either for research or for practice.
3. Emotional and Personality Related Aspects of Career decision making difficulties (Saka, Gati & Kelly, 2008)
This framework was developed by Saka, Gati and Kelly (2008) and focuses on the person rather than the content, suggesting that the personality and emotional state of the individual has a significant impact on their career decision making process. Their framework has three overarching factors: pessimism, anxiety and self-concept and identity, made up of 11 separate aspects.
I really like that this one focuses on the individual and their subjective experience - that seems important and is an obvious gap in Gati's taxonomy. But on its own, without the more objective, cognitive aspects of Gati's taxonomy, it seems equally limited.
4. Cluster Analysis of Career Decision Making Difficulties (Kelly & Lee, 2002)
Some other academics have had a go at looking at the similarities between some of the different career decision making frameworks. Kelly and Lee (2002) looked at the Career Decision Scale, the Career Decision Difficulties Questionnaire and the Career Factors Inventory, and identified six different factors: lack of information; need for information, trait indecision, disagreement with others, identity diffusion and choice anxiety.
There are a few interesting things to note about this framework. First that there is a distinction between lack of information and the need for information. It seems that, for some people, knowing that they have a gap in their knowledge isn't the same as wanting to find things out. The second thing is that the students in question didn't really focus on the lack of information about themselves. I wonder if this is a result of students who are not self-aware not realising that they aren't self-aware? This highlights that you might get different answers to these questions, depending on who you ask. A career counsellor might see clearly that an individual needs information and lacks self-awareness, but this view may not be consistent with the person's own analysis.
I like that this one brings in both affect factors (anxiety and identity) and cognitive factors, but it seems to miss some of Gati's useful things such as lack of motivation, lack of information about the self, dysfunctional myths and internal conflicts, and also only considers choice anxiety - Saka's model acknowledges other types of anxiety.
4. Combining Cognitive and Affective Models
Having had a look at all these different models, I wonder if I am most drawn to a combination of numbers 1 (Gati's cognitive taxonomy) and 3 (Saka's emotional and personality factors). This would give us a six factor model that looks like this:
To me this seem to provide a pretty comprehensive account of the difficulties that I see in my clients, and seems to cover everything that I have read in the literature. Its comprehensive nature means that we are left with 21 different aspects, which seems quite a lot, but then career decisions are complex, so they can only be simplified to a certain degree.
One point to note is that Gati's notion of 'dysfunctional beliefs' is not correlated with any other measures of career decision making difficulties. This is an interesting one because I think it implies that clients who have an unrealistic understanding of the career choice process actually feel fine. That makes sense, doesn't it? But I wonder how career practitioners should best deal with this one?
And so what?
Well, my starting point is about the training of career practitioners. I think this list of career decision making difficulties could be used as a sort of check list for practitioner training. We should be making sure that our practitioners are equipped with range of skills and techniques which are needed to help clients to deal with these kinds of scenarios. And I'm not sure that at the moment, we are.
References
Chartrand, J. M., Robbins, S. B., Morrill, W. H., & Boggs, K. (1990). Development and validation of the Career Factors Inventory. Journal of counseling Psychology, 37(4), 491.
Gati, I., Krausz, M., & Osipow, S. H. (1996). A taxonomy of difficulties in career decision making. Journal of counseling psychology, 43(4), 510.
Kelly, K. R., & Lee, W. C. (2002). Mapping the domain of career decision problems. Journal of Vocational Behavior, 61(2), 302-326.
Kleiman, T., Gati, I., Peterson, G., Sampson, J., Reardon, R., & Lenz, J. (2004). Dysfunctional thinking and difficulties in career decision making. Journal of Career assessment, 12(3), 312-331.
Saka, N., Gati, I., & Kelly, K. R. (2008). Emotional and personality-related aspects of career-decision-making difficulties. Journal of Career Assessment, 16(4), 403-424.
Sampson, J. P., Jr., Peterson, G. W., Lenz, J. G., Reardon, R. C., & Saunders, D. E. (1998). The design and use of a measure of dysfunctional career thoughts among adults, college students, and high school students: The Career Thoughts Inventory. Journal of Career Assessment, 6, 115-134
One useful starting point would be to find out what career dilemmas clients face. If we have a clear understanding of the career decision making difficulties career clients bring to their careers interviews, then we can start to ensure that the practitioners are equipped with the right range of techniques to help them.
This field is fairly under-researched (Kelly & Lee, 2002), with no real agreed definition of career decision making difficulties, and limited inclusion of the construct in career theories. The literature is also a bit confusing (ha! there's a surprise) but are a few different models I have come across.
Broadly speaking, it seems that there are two categories of career decision making difficulties: cognitive issues and affect issues (ie what is going wrong, and how people are feeling) (Chartrand, Robbins, Morrill, & Boggs, 1990). The bulk of the literature seems to focus on the cognitive issues, and aims to find scales which practitioners can use to help work out where their clients are struggling. This all feels very positivist to me and I think that perhaps career issues are a bit too complex to narrow down in this way. But I can see that it might form a useful start to a conversation.
1. Career Decision Making Difficulties (Gati et al., 1996)
One of the most widely cited (cognitive) frameworks of Career Decision Making Difficulties was written by Gati et al., in 1996. (Gati's name is all over the place within the career decision making difficulties literature). Gati identified three overarching difficulties, and separated that into 10 different strands:
Most of the factors are fairly self-explanatory, but 'dsyfunctional myths' refers to an incorrect understanding of the process of either choosing or getting a job - the clients who think that you can tell them what to do, for example, or who think that making a living from writing novels is easy.
2. Career Thought Index (Sampson et al., 1996)
Gati's framework overlaps a fair bit with another one called the Career Thought Index (Sampson et al., 1998) which is linked to the Career Information Processing theory which you may have come across before. This one is also divided into three areas:
- Decision-making Confusion (thoughts and emotions which make it impossible to make career decisions)
- Commitment Anxiety (an inability to commit to one choice through anxiety)
- External conflicts (usually a conflict between what the client wants and what their loved ones want for them).
These two approaches (Gati's taxonomy and Sampson's Career Thought Index) were compared and Kleinman et al. (2004) found that they basically measured the same thing, and suggested that it didn't much matter which one people chose either for research or for practice.
3. Emotional and Personality Related Aspects of Career decision making difficulties (Saka, Gati & Kelly, 2008)
This framework was developed by Saka, Gati and Kelly (2008) and focuses on the person rather than the content, suggesting that the personality and emotional state of the individual has a significant impact on their career decision making process. Their framework has three overarching factors: pessimism, anxiety and self-concept and identity, made up of 11 separate aspects.
I really like that this one focuses on the individual and their subjective experience - that seems important and is an obvious gap in Gati's taxonomy. But on its own, without the more objective, cognitive aspects of Gati's taxonomy, it seems equally limited.
4. Cluster Analysis of Career Decision Making Difficulties (Kelly & Lee, 2002)
Some other academics have had a go at looking at the similarities between some of the different career decision making frameworks. Kelly and Lee (2002) looked at the Career Decision Scale, the Career Decision Difficulties Questionnaire and the Career Factors Inventory, and identified six different factors: lack of information; need for information, trait indecision, disagreement with others, identity diffusion and choice anxiety.
There are a few interesting things to note about this framework. First that there is a distinction between lack of information and the need for information. It seems that, for some people, knowing that they have a gap in their knowledge isn't the same as wanting to find things out. The second thing is that the students in question didn't really focus on the lack of information about themselves. I wonder if this is a result of students who are not self-aware not realising that they aren't self-aware? This highlights that you might get different answers to these questions, depending on who you ask. A career counsellor might see clearly that an individual needs information and lacks self-awareness, but this view may not be consistent with the person's own analysis.
I like that this one brings in both affect factors (anxiety and identity) and cognitive factors, but it seems to miss some of Gati's useful things such as lack of motivation, lack of information about the self, dysfunctional myths and internal conflicts, and also only considers choice anxiety - Saka's model acknowledges other types of anxiety.
4. Combining Cognitive and Affective Models
Having had a look at all these different models, I wonder if I am most drawn to a combination of numbers 1 (Gati's cognitive taxonomy) and 3 (Saka's emotional and personality factors). This would give us a six factor model that looks like this:
To me this seem to provide a pretty comprehensive account of the difficulties that I see in my clients, and seems to cover everything that I have read in the literature. Its comprehensive nature means that we are left with 21 different aspects, which seems quite a lot, but then career decisions are complex, so they can only be simplified to a certain degree.
One point to note is that Gati's notion of 'dysfunctional beliefs' is not correlated with any other measures of career decision making difficulties. This is an interesting one because I think it implies that clients who have an unrealistic understanding of the career choice process actually feel fine. That makes sense, doesn't it? But I wonder how career practitioners should best deal with this one?
And so what?
Well, my starting point is about the training of career practitioners. I think this list of career decision making difficulties could be used as a sort of check list for practitioner training. We should be making sure that our practitioners are equipped with range of skills and techniques which are needed to help clients to deal with these kinds of scenarios. And I'm not sure that at the moment, we are.
References
Chartrand, J. M., Robbins, S. B., Morrill, W. H., & Boggs, K. (1990). Development and validation of the Career Factors Inventory. Journal of counseling Psychology, 37(4), 491.
Gati, I., Krausz, M., & Osipow, S. H. (1996). A taxonomy of difficulties in career decision making. Journal of counseling psychology, 43(4), 510.
Kelly, K. R., & Lee, W. C. (2002). Mapping the domain of career decision problems. Journal of Vocational Behavior, 61(2), 302-326.
Kleiman, T., Gati, I., Peterson, G., Sampson, J., Reardon, R., & Lenz, J. (2004). Dysfunctional thinking and difficulties in career decision making. Journal of Career assessment, 12(3), 312-331.
Saka, N., Gati, I., & Kelly, K. R. (2008). Emotional and personality-related aspects of career-decision-making difficulties. Journal of Career Assessment, 16(4), 403-424.
Sampson, J. P., Jr., Peterson, G. W., Lenz, J. G., Reardon, R. C., & Saunders, D. E. (1998). The design and use of a measure of dysfunctional career thoughts among adults, college students, and high school students: The Career Thoughts Inventory. Journal of Career Assessment, 6, 115-134
Tuesday, 11 June 2019
Why, despite everything, women are happier at work than men
So here's an interesting thing. Despite being disadvantaged in the workforce, women report higher levels of job satisfaction than men. It's a strange paradox. We know that women have access to fewer opportunities than men - there are occupational fields they struggle to get into, and have a much harder time getting into senior posts than men do. They also get paid less, and are assumed to be less capable than their male counterparts, meaning that they have to work harder and perform better than men to achieve the same rewards. When children come along, their choices are yet more constrained as employers' perception of the value they can add in the work place goes down and their domestic responsibilities increase relative to men's. And yet, consistently, they report being happier at work than men, and women working part time report being the happiest of all.
There are a few different theories that have been put forward to account for this paradox. The first is the 'grateful slaves' view of Hakim (2000) which suggests that for women, the opportunity to work part time or flexibly is enough - they don't need high salaries or interesting roles to be happy at work, as long as they can pick their children up from school, nothing much else matters. The second theory is the 'making the best of a bad job' idea from Walters (2005) which suggests that women know that their opportunities are constrained, but they also know that their hands are tied, they aren't going to be able to do anything about it, so might as well just get on with it and be as positive as they can. The third theory is by Clark (1996) which suggests that job satisfaction is significantly related to what he calls 'relative income', described as the difference between what you expect to earn and what you actually earn. His theory is that women's expectations of pay are low and so they end up pleasantly surprised by their earnings and this increases their levels of satisfaction. Each of these theories has some empirical backing, and seems to account for some of the differences in levels of satisfaction between men and women, but by no means all of it.
A more recent study by Zou (2015) indicates that the key to solving the puzzle lies in what they call different work orientations.
Job satisfaction is made up of a range of different specific work features, and Zou's study suggests that women and men tend to derive their job satisfaction from different aspects of work. They describe five different work orientations:
Intrinsic orientation: getting your job satisfaction from the joy of the job itself
Extrinsic orientation: valuing pay and promotion
Effort orientation: work load and hours
Future orientation: personal and career development
Human orientation: interpersonal relationships
Their study found two interesting things. First, that people who are highly orientated towards intrinsic and human factors are most likely to have higher levels of job satisfaction, and then second, you've guessed it, women are more likely to have stronger orientations to intrinsic and human factors.
So, women work because they are interested in the jobs themselves, and because they like the people. Men also work for these reasons, but tend to think that pay and promotion and career development are more important. And it just so happens that intrinsic motivation and relationships are factors which link closely with job satisfaction.
These different orientations go some way to explaining career choices. Men are more likely to be orientated to pay and promotion, so perhaps push for higher salaries, choose higher paying occupations and focus more on getting promoted. Women are less interested in these things so don't go out of their way to make them happen.
But where does that leave us? Job satisfaction has always been my first goal as a career practitioner, and on this measure women are way ahead. But does that mean that we should stop trying to encourage women to go for promotions? I don't think so, although we don't want to be so keen to make women's lives better that we end up making their lives worse.
The Nordic model is interesting. Scandinavia has what is considered to be the most progressive regime for women, and though ideology and legislation has reached a point where women's participation in the labour market is fairly similar to men's. They have more women working full time than anywhere else in Europe, which is thought to be a real positive, but given that we know that women working part time are happier than women working full time, is this really the right aspiration for the rest of us?
One important issue that I can't see how to resolve is that of power in society. My instinct is to say that what we need to do is re-model our social values, and stop conceptualising pay, promotion and status as the signs of success. If we could start to value job and life satisfaction as much, or more, than money, then we could find ourselves in a society that accords women who work part time in female spheres and lower level jobs as equal to men who command higher salaries and have more senior roles. But how is this ever to be managed?
A less extreme range of salaries across the country could perhaps help. If senior roles were paid only a bit more than junior roles, then other factors would take on a more significant role in people's career development choices. At the moment, the increase in salary associated with promotion is so great that it overshadows other considerations - people go for promotions because of the money, regardless of whether they actually think they will be happier in these more senior roles. If the financial incentives were only marginal, then people might be more likely to take a balanced view of promotion, going for it only if it ticks a number of boxes for them. We might well then see more of a balance in the senior roles, as they will attract people who are motivated by a range of different factors, and not just those (disproportionately men) who are motivated by money.
On a more individual level, if we can find ways to encourage men to pursue options which don't just satisfy their monetary motivation, then they might end up being happier in the workplace. And this will allow for more opportunities for women to find a wider range of options which can meet their needs. Much of the push towards gender equality is focused on women - empowering them to make bolder choices. But I wonder if here, I am suggesting that actually it's men who need to be re-educated to work out what is going to make them happy?
Just as an aside, I think the literature on job satisfaction is a bit patchy and part of the reason for that is that there are so many things that seem to be linked to it, yet academics seem to hone in on just one or two, meaning that we don't have a particularly good overall picture. A lot of the large scale studies have been done by economists, for example, who seem convinced that income is the most important thing to study, which is a bit strange, given that we know that income has only a small, weakly significant correlation with job satisfaction (Judge et al.'s 2010 meta-analysis found a correlation of 0.15 p<0.05)
References
Clark, A. E. (1996). Job satisfaction in Britain. British journal of industrial relations, 34(2), 189-217.
Hakim, C. (2000). Work-lifestyle choices in the 21st century: Preference theory. OUp Oxford.
Judge, T. A., Piccolo, R. F., Podsakoff, N. P., Shaw, J. C., & Rich, B. L. (2010). The relationship between pay and job satisfaction: A meta-analysis of the literature. Journal of Vocational Behavior, 77(2), 157-167.
Walters, S. (2005). Making the best of a bad job? Female part‐timers’ orientations and attitudes to work. Gender, Work & Organization, 12(3), 193-216.
Zou, M. (2015). Gender, work orientations and job satisfaction. Work, employment and society, 29(1), 3-22.
There are a few different theories that have been put forward to account for this paradox. The first is the 'grateful slaves' view of Hakim (2000) which suggests that for women, the opportunity to work part time or flexibly is enough - they don't need high salaries or interesting roles to be happy at work, as long as they can pick their children up from school, nothing much else matters. The second theory is the 'making the best of a bad job' idea from Walters (2005) which suggests that women know that their opportunities are constrained, but they also know that their hands are tied, they aren't going to be able to do anything about it, so might as well just get on with it and be as positive as they can. The third theory is by Clark (1996) which suggests that job satisfaction is significantly related to what he calls 'relative income', described as the difference between what you expect to earn and what you actually earn. His theory is that women's expectations of pay are low and so they end up pleasantly surprised by their earnings and this increases their levels of satisfaction. Each of these theories has some empirical backing, and seems to account for some of the differences in levels of satisfaction between men and women, but by no means all of it.
A more recent study by Zou (2015) indicates that the key to solving the puzzle lies in what they call different work orientations.
Job satisfaction is made up of a range of different specific work features, and Zou's study suggests that women and men tend to derive their job satisfaction from different aspects of work. They describe five different work orientations:
Intrinsic orientation: getting your job satisfaction from the joy of the job itself
Extrinsic orientation: valuing pay and promotion
Effort orientation: work load and hours
Future orientation: personal and career development
Human orientation: interpersonal relationships
Their study found two interesting things. First, that people who are highly orientated towards intrinsic and human factors are most likely to have higher levels of job satisfaction, and then second, you've guessed it, women are more likely to have stronger orientations to intrinsic and human factors.
So, women work because they are interested in the jobs themselves, and because they like the people. Men also work for these reasons, but tend to think that pay and promotion and career development are more important. And it just so happens that intrinsic motivation and relationships are factors which link closely with job satisfaction.
These different orientations go some way to explaining career choices. Men are more likely to be orientated to pay and promotion, so perhaps push for higher salaries, choose higher paying occupations and focus more on getting promoted. Women are less interested in these things so don't go out of their way to make them happen.
But where does that leave us? Job satisfaction has always been my first goal as a career practitioner, and on this measure women are way ahead. But does that mean that we should stop trying to encourage women to go for promotions? I don't think so, although we don't want to be so keen to make women's lives better that we end up making their lives worse.
The Nordic model is interesting. Scandinavia has what is considered to be the most progressive regime for women, and though ideology and legislation has reached a point where women's participation in the labour market is fairly similar to men's. They have more women working full time than anywhere else in Europe, which is thought to be a real positive, but given that we know that women working part time are happier than women working full time, is this really the right aspiration for the rest of us?
One important issue that I can't see how to resolve is that of power in society. My instinct is to say that what we need to do is re-model our social values, and stop conceptualising pay, promotion and status as the signs of success. If we could start to value job and life satisfaction as much, or more, than money, then we could find ourselves in a society that accords women who work part time in female spheres and lower level jobs as equal to men who command higher salaries and have more senior roles. But how is this ever to be managed?
A less extreme range of salaries across the country could perhaps help. If senior roles were paid only a bit more than junior roles, then other factors would take on a more significant role in people's career development choices. At the moment, the increase in salary associated with promotion is so great that it overshadows other considerations - people go for promotions because of the money, regardless of whether they actually think they will be happier in these more senior roles. If the financial incentives were only marginal, then people might be more likely to take a balanced view of promotion, going for it only if it ticks a number of boxes for them. We might well then see more of a balance in the senior roles, as they will attract people who are motivated by a range of different factors, and not just those (disproportionately men) who are motivated by money.
On a more individual level, if we can find ways to encourage men to pursue options which don't just satisfy their monetary motivation, then they might end up being happier in the workplace. And this will allow for more opportunities for women to find a wider range of options which can meet their needs. Much of the push towards gender equality is focused on women - empowering them to make bolder choices. But I wonder if here, I am suggesting that actually it's men who need to be re-educated to work out what is going to make them happy?
Just as an aside, I think the literature on job satisfaction is a bit patchy and part of the reason for that is that there are so many things that seem to be linked to it, yet academics seem to hone in on just one or two, meaning that we don't have a particularly good overall picture. A lot of the large scale studies have been done by economists, for example, who seem convinced that income is the most important thing to study, which is a bit strange, given that we know that income has only a small, weakly significant correlation with job satisfaction (Judge et al.'s 2010 meta-analysis found a correlation of 0.15 p<0.05)
References
Clark, A. E. (1996). Job satisfaction in Britain. British journal of industrial relations, 34(2), 189-217.
Hakim, C. (2000). Work-lifestyle choices in the 21st century: Preference theory. OUp Oxford.
Judge, T. A., Piccolo, R. F., Podsakoff, N. P., Shaw, J. C., & Rich, B. L. (2010). The relationship between pay and job satisfaction: A meta-analysis of the literature. Journal of Vocational Behavior, 77(2), 157-167.
Walters, S. (2005). Making the best of a bad job? Female part‐timers’ orientations and attitudes to work. Gender, Work & Organization, 12(3), 193-216.
Zou, M. (2015). Gender, work orientations and job satisfaction. Work, employment and society, 29(1), 3-22.
Subscribe to:
Posts (Atom)