Professionalising teachers in career dialogue: an effect study

ABSTRACT As a result of the changing notions of work, schools are increasingly acknowledging that they have a strong responsibility in guiding students not only in their academic growth, but also in their career development. This paper presents the results of a study about the effects of teacher training on career dialogues promoting career competency development in students. For the quantitative part of the study, a quasi-experimental research design was used to measure effects among 2291 students. Video recordings of conversations were used to do qualitative research. The results show that only when off-the-job training is followed by on-the-job coaching, the professionalising of teachers proves to be effective as measured on student level. Students notice that the guidance conversations are more appreciative, reflective and activating and that they focus more on self-image development, work and career action. The observations made with regard to the guidance conversations also show that these conversations become more career related.


Introduction
Although concepts such as the 'boundaryless career' (Arthur & Rousseau, 1989) and the 'protean career' (Hall, 2002) are contested (see Arnold, 2011;Rodrigues & Guest, 2010;Van Hoof, 2007), most career researchers agree that careers are more unpredictable nowadays than 30 years ago. As a result, individuals are expected to demonstrate more self-directedness in the labour marketpeople have to 'design' their lives themselves (Savickas et al., 2009). Life designing, however, is a complex processit is cognitive and intuitive at the same time (Krieshok, Black, & McKay, 2009) and requires experiential as well as reflective learning . Life designing is realised mostly by means of a dialogical co-construction of a career narrative by guidance counsellors and their clients (Savickas, 2011).
In many countries, trained guidance counsellors 1 are responsible for career guidance in education (Hooley, Watts, & Andrews, 2015). Recently, however, there is a trend towards making teacherswho often do not have specific training in the fieldresponsible for delivering career and employment guidance to students (Hearne & Galvin, 2015;Hooley et al., 2015). Whether this is a positive or negative development remains to be seen. It is clear that the current socio-political climate of education in Western societies favours an approach to teaching and learning in which test preparation and scripted curricula are the preferred methods (Hillocks, 2002;Marshall, 2009) which leaves little room for learning processes that would foster 'life designing'. In fact, a focus on high-stakes testing and standardisation has led to a rather narrow view of what counts as teaching and learning (Franciosi, 2004;Hargreaves, 2003;Lipman, 2004;Ravitch, 2010) and the policies associated with this entrench teachers almost solely in efforts to transmit knowledge. There is little space for narrative approaches like 'life designing' and as a result teachers have few encounters with their students that might foster this type of learning. This in turn leaves teachers with little to no experience in this largely verbal process that entails 'a collaborative relationship' (McIlveen & Patton, 2007, p. 10).
In addition, research shows that the learning processes aimed at 'life designing' require a learning environment where experiential and meaning-oriented learning are possible (Kuijpers, Meijers, & Gundy, 2011;Lengelle, Meijers, Poell, & Post, 2014). More specifically, learning processes aimed at life designing begin with a focus on actual experiences with real-life problems, but these experiences do not yet produce a sense of identity and direction. In order for that to happen, a dialogue is needed which enables students to attach meaning to their experiences. With this in mind, it is clear that teachers must be able to have a meaningful career dialogue with students. After all they are responsible for guiding the experiential learning that takes place and also havein particular in vocational educationtheir own experiences in the world of work. Another reason to focus on teachers is that there is considerable evidence that young people frequently seek out career support from teachers (Hutchinson & Bentley, 2011;Munro & Elsom, 2000;Nugent, Barker, & Welch, 2014).
This article presents the results of two separate studies regarding a training programme for teachers in Dutch pre-vocational education which was aimed at getting teachers to start career dialogues with their students. Until now no research has been available regarding the training of teachers in narrative approaches in careers guidance; available literature focuses on teachers trained in trait-and-factor approaches (Watts & Sultana, 2004). Studies show that both on a qualitative and a quantitative level, the latter type of training is generally mediocre (McCarthy, 2004;Oomen, Van den Dungen, Pijls, & Egelie, 2012). Research by Winters (2012), Baert (2009), Winters, Meijers, Lengelle, &Baert (2012) and Winters et al. (2013) however, has shown that when career teachers develop a dialogical instead of a monological style, students respond with more reflexive answers. It is for this reason that we examined the effects of a training programme that aimed to help teachers conduct a career dialogue with their students. A career dialogue is in form appreciative, reflective and activating and in content aimed at the development of a selfimage (i.e. identity) and career-oriented actions.

Career dialogue: some considerations
Because careers are less predictable or even 'chaotic' (Pryor & Bright, 2011), there is an argument to be made for career guidance that is less focused on 'making the right choice' and more on developing career competencies.  and Kuijpers, Schyns, and Scheerens (2006) identify five distinctive career competencies: capacity reflection (observation of capabilities that are important for one's career), motivation reflection (observation of wishes and values that are important for one's own career), work exploration (researching work and job possibilities), career directedness (making thoughtful decisions and taking actions that allow work and learning to correspond with one's capabilities and motivation and challenges at work), and finally, networking (building and maintaining contacts focused on career development). In a large-scale study among students of Dutch pre-vocational, secondary and higher vocational education, Kuijpers et al. (2011) andKuijpers andMeijers (2012a) showed that a learning environment that stimulated real-life experiences with work and a dialogue about these experiences, did indeed contribute to the development and use of career competencies. Even when personality traits of students and their differing educational programmes were taken into account, the characteristics of the learning environment influenced the degree to which and the kind of career competencies learned and used by students. In particular, the career dialogue in schools and the conversation students have in the workplace proved to be crucial for applying career competencies and developing a career identity. Both contributed to career reflection, career forming (work exploration plus career directedness) and networking. In fact this dialogue was more strongly correlated with the development of career competencies than personality traits. Current forms of career guidance, such as personal development plans and portfolios, and more traditional ones such as conversations with the guidance counsellor and tests, did not noticeably encourage students to use career competencies. The main reason for this is the absence of a dialogue about the content of personal development plans and portfolios. Research shows both teachers and students focus on the formal requirements of personal development plan and portfolios resulting in strictly instrumental use of this tool (see also Mittendorff, den Brok, & Beijaard, 2010;Winters et al., 2013). This finding corresponds with other research: a career dialogue is needed to (re)formulate dreams and goals regarding a career (Pizzolato, 2007), to create vocational hope, that is, the feeling that a career is possible (Diemer & Blustein, 2007) and to help students develop a career narrative (Lengelle et al., 2014;. A career dialogue enables students to apply career competencies and in doing so, to develop a career identity in steps (Savickas, 2011). A career identity expresses itself in the form of a story (Savickas, 2005) and in a complex and dynamic world, this 'career story' helps people define who they are and how they should act within a career context. It does so by creating and providing meaning and direction (Wijers & Meijers, 1996) and by constructing a sense of causality and continuity about one's career path (LaPointe, 2010).
A career story takes shape in a dialogue in which students and their teachers or guidance counsellors work together to give meaning to the experiences students have around work. In such a collaboration, students are encouraged to construct a career story based on emotionally salient events, with respect for the feelings associated with those events, where they alternately visit both the past and future. According to Savickas (2011) three steps are taken in shaping a career story. First 'small stories' are constructed to give meaning to actual experiences. Second, the connections between the small stories and the resulting reinterpretations become a 'larger story'. In a third step, based on this larger story, actual action steps (i.e. work exploration, career shaping and networking) can be formulated. This approach of giving meaning to specific experiences as they relate to a more generalised career story and formulating action has much in common with the model of change developed by Prochaska, Diclemente, and Norcross (1992) and the career-learning model of Law (1996).
In the process of developing career competencies, teachers play a supporting role; they do not offer explanations or solutions, but they help students to articulate their experiences so that the underlying perspectives related to those experiences become apparent (Savickas, 2011). That means that in the conversation with students, teachers must not only make space for discussing actual experiences but also make room for students' emotions. In other words, teachers must show appreciation (affect) before stimulating reflection and action.
The context of the research and the research questions The Dutch educational system has two main tracks after primary education. Almost 60% of all students enter the vocational track, starting with pre-vocational education (age 12-16), followed by secondary vocational education (age 16-20). The other 40% of students enter general secondary education after primary school and choose between education leading to higher vocational and/ or pre-university education. The vocational track has less status than the general track and especially in pre-vocational education, children from disadvantaged groups (i.e. traditional working class and minority groups) are over-represented. Although students formally have an equal opportunity to enter higher vocational education and even university, only 15% do so (Central Bureau of Statistics, 2009). For most children who enter pre-vocational education, formal education ends at age 17 (when compulsory education ends) or 18 (after finishing secondary vocational education).
Schools do not have the financial means to provide students with career guidance given by professional counsellors. Therefore, since the 1960s, teachers in the Netherlands have had the important role of providing career guidance to students. Career guidance happens in 'career lessons' during group and individual conversations with students and its frequency varies from weekly to once every three months (Meijers, Kuijpers, & Bakker, 2006;Oomen et al., 2012). However, teachers find it quite difficult to provide career guidance and more specifically have difficulty conducting career conversations with students (Kuijpers, 2011;Kuijpers & Meijers, 2012b). Currently, there is no research that has examined whether and to what degree teachers can learn to have useful career conversations.
In the current study, we examined to what degree training teachers in pre-vocational education contributed to them having good career dialogues with their students. The overall research question was: What is the impact of the training programme on the form and the content of the career dialogue between teacher and student? In order to answer this question, two studies were conducted. The first was a quantitative project in which we studied whether teachers changed their guidance-conversation routines after training in such a manner that students experienced a different form and different content during those conversations. The research question was: (a) To what extend are career conversations between teachers and students in form appreciative, reflective and activating and in content aimed at the development of self-image (identity) and career-oriented actions, and (b) do, according to students, the conversations change in form and content after teachers have taken the training? This study involved 2211 students whereby each was given a questionnaire before and after the teacher training and thereby asked to evaluate the career conversations they had had with a teacher both with regard to form and content. Of these students 1673 completed the questionnaire fully both times (for more details see below).
The second study was a qualitative examination of recorded guidance conversations and looked at whether teachers were able to change their routines in guidance conversations based on the input of the training. Here the question was: How do career conversations change in content and form as a result of the training? To be more objective in our study, we did not use self-reports of teachers or students, but instead we analysed 50 video recordings of career conversations done before and 32 video recordings after the training.

The training programme: professionalising teachers in career dialogue
In the period between September 2010 and March 2011, as part of a project of the Stichting Platforms VMBO (www.platformsvmbo.nl) financed by the Dutch Department of Education, a training programme was provided for teachers from pre-vocational education which was aimed at teaching them to have effective career dialogues with students. The training programme had a two-phase and two-layered structure. The two-phases consisted of an off-the-job and an on-the-job portion. In the off-the-job part, which took a total of three days with periods of two to four weeks in between, the emphasis was on explaining the theory and putting the theory into practice in career conversations within a safe environment. In the on-the-job portion, the emphasis was on the translation of what was learned to the school environment. For the dissemination and anchoring of this knowledge within the organisation, a two-layered structure was used, in which national experts trained two teachers from each participating school ('school coaches') in a four-day training programme, so that these school coaches could in turn coach teams within their own schools during the in-school training. In other words, 'school coaches' became the designated experts who would later train other teacher teams within their own schools. After the four-day training of the school coaches, the complete teacher team (including the school coaches) was trained for two days offthe-job. Before the actual course, teachers were given training in basic conversation techniques if they wished. After being trained for two days off-the-job, all teachers in the team did a foursession in-school training programme (two individual and two team sessions) with their school coaches, using video-recorded guidance conversations as a starting point for learning. The recordings were made by the teachers themselves and were of guidance conversations they had had with their students. National trainers trained both the school coaches and the teacher teams off-the-job and they also supported the part of the training that took place at the schools. After the on-the-job portion, a fifth training day for school coaches took place in which the aim was to anchor and disseminate the knowledge and insights acquired during the training.
Teachers were trained partly through role playing activities and partly by analysing recorded career conversations they had had with their own students. They were also provided with some recent theoretical insights about career dialogues aimed at life designing. They were instructed to use a meaningful experience of a student as the starting point for a career conversation. This meant getting students to tell a small story about an actual experience and help them to assign personal meaning to it by combining various small stories into a larger story. Subsequently, teachers were taught to encourage students to use career competences by asking them specific questions. Finally, much attention was paid to affective components of the conversation: teachers were taught to give appreciative feedback (i.e. to engage with the emotions of students instead of ignoring these) and explicitly expressing particular qualities they felt they had identified in students (for an elaborate description of the training see Den Boer, Kronenberg, Kuijpers, Meijers, & Vogels, 2014). The content of both the training programme and questionnaire was based on insights acquired in earlier research. The pre-test answers of the students were not used to develop the training programme.

Method and instrument
As mentioned above, the training programme was developed as part of an initiative financed by the Dutch Department of Education. Each of the more than 200 schools for pre-vocational education could register a team to take part. In the end, 20 teams registered in this first training initiative (a total of 230 teachers). Later a second training initiative was implemented. The current research is based on the schools participating in the first initiative. A little more than half (59%) of the 230 teachers who participated in the research were female; one third had less than four years of work experience, one third had 4-10 years and one third had more than 10 years of work experience.

Study 1: quantitative analysis
Both before and after the training sessions, questionnaires were administered to an experimental group of 2211 students. These were students of 191 teachers who were part of the total training group of 230 teachers who participated in the programme. Of the 2211 students, 1673 completed the questionnaire fully both times. The same questionnaire was used with a control group of 593 students (students of colleague-teachers within the same school who did not participate in the training). In the end, 25 students could not be slotted because they had not filled in the names of their teachers. For every class that participated in the research, a comparable group within the same school was planned as a control. However, in several schools this was impossible because the school was too small, therefore a control group was not available. For the analyses of this article, the group of students of teachers who attended the full training programme (838 students) was compared with the group of students of teachers who did not attend the training at all or attended only the off-the-job part (1428 students). Although the size of the experimental and control groups differed considerably, both groups were comparable because of the similarity of the distributions with regard to the output variables (difference in skewness from .002 on content about self-image to .068 on appreciative form of conversation) and deviation (difference of .2205 on content about self-image to .5038 on appreciative form of conversation). To minimise potential bias, the analyses checked for personal and situational-based variables to control for differences between schools. Most students were in their third and fourth year and came from all sectors in pre-vocational education (i.e. healthcare, economics, technical education, agriculture and multi-sectoral education). The experimental and control groups were comparable with respect to sector, school level and study year. However, in the experimental group there were some more students from the sectors' economics and education, and more from the lowest school level. Their average age was 15% and 20% of students were from the immigrant population. The student group was divided equally along gender lines.
To measure the effects of the training done, there were two variants of the experimental/control group: (1) students of the teachers' group who participated in the two-day off-the-job training programme with a control group of teachers who had no training at all; (2) students of teachers who had participated in the full-training process off the job, and also had at least two individual coaching sessions, and two team-coaching sessions on the job, compared with a control group of students of teachers who did not participate in the (full) training programme.
The self-evaluation based questionnaire used for collecting data from the students is an adaptation of the questionnaire used by Kuijpers et al. (2011) and Kuijpers and Meijers (2012a). The questionnaire measures on a four-point-scale two aspects of the form and two aspects of the content of a career dialogue. The form aspects are (1)  , study year, study level and sector were included as control variables. The variable 'full training program' is a digitome variable and was assigned when the teacher of the student completed the off-the-job as well as the on-the-job part of the professionalising project.
To determine whether students experienced different outcomes when engaging in conversations with trained versus untrained teachers, students' responses to the measures in the questionnaire at the end of the teacher training were compared using regression analyses. To prevent personal or situational variables from differing too much between the experimental and the control groups, the analyses were controlled for these variables. The influence of career guidance conversation was analysed regardless of the differences between the groups. Due to the hierarchical nature of the data (students within classes within schools), multilevel regression analysis was conducted, using the MLWIN v.2.0 program (Rasbash, Charlton, Browne, Healy, & Cameron, 2005). All continuous variables were normalised to a mean of zero and to a standard deviation of one. All hypotheses were tested two-sided and p-values less than .05 were considered to be significant, unless indicated otherwise.

Study 2: qualitative analysis
For the qualitative part of the study, teachers who participated in the training made video recordings of their career conversations with students before and after the training programme. The instruction given to the teachers was that they had to record a video of a career conversation they had with a student just before their training and soon after the last on-the-job part of the training. These types of career conversations were part of their regular work activities. At the start of the training, 133 teachers had handed in such a recording. Of each of the 20 participating teacher teams, 2-3 recordings were selected from the teachers who had, obtained in the second measurement of Study 1, the highest scores on all dimensions of career guidance conversations as well as from the teachers who had the lowest scores, (taking into consideration also whether the recording made had sufficient sound quality to be analysed). In all, the recordings of 50 teachers were analysed. After the training programme, a total of 35 of these 50 teachers handed in a second recording, of those the quality of 32 recordings was good enough to be analysed.
For the qualitative analyses of the recorded conversations, we used the instrument developed by Winters et al. (2009). This instrument makes a distinction between the form and content of a career dialogue. With regard to form, the instrument firstly looks at whether the teacher is appreciative during the conversation, that is whether there is a focus on abilities, active listening or sharing relevant experience, as well as asking about feelings and validating the student. Secondly, the instrument looks at whether teachers stimulate reflection, for instance: do they make students think, solve a problem and let them put experiences and their meanings into words. Finally, the instrument measures whether the teacher helps activate (i.e. encourages action in) the student, which is seen in a teacher challenging students to engage in further orientation, in experimenting with what students want to learn and in collecting evidence of students' abilities. With regard to content, the instrument looks at whether the following topics are discussed: course of studies (i.e. study progress), internship (i.e. progress with work placement), private life (e.g. home situation and experiences outside school), choices (e.g. future study choices and occupational choices) and so-called critical experiences (i.e. experiences that are important and/or emotionally relevant for students). The instrument also looks to see to what degree self-image is talked about and how it is talked about.
Finally, the instrument examines to what extent the five career competencies are spoken about and at what level this happens. The first competency is reflection on qualities, whereby the first level involves reflecting on skills, the second reflecting on one's own performance and the third reflecting on one's own person ('I am someone who … '). The second competency involves a reflection on motives, whereby the first level asks 'what attracts me?', the second 'why am I attracted by this?' and the third 'where do I go/stand for?'. The third competency involves work exploration, whereby the first level is focused on concrete work activities, the second on dilemmas and developments at work, and the third on work (activities) in association with one's own qualities and motives. The fourth competency is career shaping, whereby the first level is focused on action for study, the second on organising one's own course of studies and the third on collecting information/evidence for one's future career. The fifth competency is networking, whereby the first level is who you know, the second how to use a network and the third how to expand your network for your career. In the analysis of career dialogues, the research first looked at whether attention was being paid to the development of these five career competencies and in the second place at which of the three levels these conversations were being carried out.
This instrument was tested for its usefulness and reliability by analysing three conversations done by two independent researchers. In response to their analyses, the instrument was slightly adjusted. Quotes from the pre-test and post-test conversations were classified based on aspects of form and content. For every five minutes of conversation, a researcher tallied the relevant aspects. We compared the degree of career orientation in the pre and post-test conversations in general. Also, we studied differences in pre-and post-test results within teacher cases. The cases that showed key differences between the pre-and post-test conversations (8 out of 16 conversations) were fully transcribed.

Study 1: quantitative analysis
Based on the results of the first study, it is clear that students consider their career conversations with teachers to be appreciative (mean 2.78) and not very reflective or activating (mean 2.38). According to students, little time is spent speaking about their qualities and motives (mean 2.08), and even less attention (mean 1.95) is paid to work and activities that would be suitable and that they could do. In the post-test measurement, students of teachers who participated in the full training programme found that their guidance conversations had become more career related (as opposed to school-focused) over time as compared with students of teachers who did not do the full training programme. Table 1 shows that when we compared the training group to the control group, participating in the full training programme contributes to the scores of the outcome variables (i.e. post-test scores on form and content of guidance conversations). The guidance conversations after the training are more appreciative and reflective/activating, and they are more about selfimage, work and career actions of students than before the training. That is the case when we take the scores of the pre-test into account, and control for the personal and situational factors. The model improved significantly when we added the training programme into the model in which personal and situational factors were taken into account. Conversations became more career oriented with respect to form, which expressed itself in more appreciative (χ2 6.4, df 1) and reflective/activating (χ2 7.8, df 1) conversations, and they also changed with respect to content, which expressed itself in more attention paid to self-image (χ2 6.7, df 1) and work/job actions (χ2 4.5, df 1). The training is the only variable that explains the variation in the scores regarding the form and content of the conversations, except with regard to talking about work and career competencies. On this latter point, boys show a much greater and significant improvement in the conversations as compared with girls (z-score .063). Apart from the increase regarding work and career action from the perspective of male students, there is no significant impact regarding ethnicity, year of study, level of study or sector on the increase of career-related form and content after the training period. Students of teachers who participated in the training, experienced their guidance conversations as being more appreciative, reflective and activating since the pre-test than students in the control group did. Moreover, the former group talked more about self-image, work and career actions. When we conducted the analysis using the off-the-job training programme as the experimental condition instead of the full training programme, we found no significant effects in the development of the form and content of career conversations, as measured at a student level (not shown in table).
So, the results of the quantitative study indicate that training programmes should not only be off the job, but should include an on-the-job portion as well. Training teachers in how to be appreciative, reflective/activating and how to talk about self-image, work and career actions results in changes in students' perspective with regard to the career conversations. Study 2: qualitative analysis

Form of the conversation
The results of the qualitative study confirm the results of the quantitative study. This indicates that the training programme contributed to changes in teachers' routines with regard to their guidance conversation with students. We see that teachers were able to change their guidance conversations from talking to students about their study progress to talking with (this includes listening) students about their career development. Table 2 shows that the conversations are more career related after training than before. This holds true for teachers who scored in the second measurement of the quantitative study 'above average' as well as for the 'below average' group. Conversations after training also took more time as compared with the conversations before training. The pre-test conversation took an average of 11 minutes and the post-test conversation 19 minutes; it was clear that the teacher and student talked more fully about the career of the student. In addition, the development of the self-image of students was taken into account more frequently. Even work and career actions became topics of conversation, whereas in the pre-test conversations, these subjects were barely addressed.
As to the form of the post-training conversations, we see that students had more input than before. First, (not shown in Table 2) unlike in the pre-test, the goal and content of the conversation was determined beforehand in the post-test conversation (i.e. a clear contact and contract). Examples of this are: 'In this conversation we will talk about your wishes for the future and how to get there', 'Today we will talk about the choice for your internship' and 'You came to me because you wanted to talk to me about your application for your future vocational education, right?' In the post-test conversation, teachers displayed more active listening, they validated students' input more often and they asked about feelings more frequently. More often than in the pre-test conversation, appreciative remarks such as 'I think you are really good at that' were made by teachers and students were asked about their qualities more frequently. In general, in the post-test conversation there was more focus on abilities instead of inabilities, as illustrated in 'What are you really good at?', 'Yes, it is difficult to say what you are good at, but anything you say is all right. You say you are good at talking with others, what else are you good at?', 'Do you know what you are good at; what would make you a good hairdresser?' and 'Could you mention some of your talents that might be useful for the work you desire?'

Content of the conversation
With regard to the content of the conversation, we saw that students were challenged more in thinking further and more deeply about their careers and orienting themselves. There was a strong increase in talk about the students' occupational choice and a small increase in conversations about (progress with regard to) the internship of students and students' private lives. In the pretest conversation, students were mainly asked to talk about the things they already knew. A teacher for instance asked in the pre-test conversation: 'What do you do in your spare time?', and the student answered: 'Soccer, four times a week [ … ] and I always do my homework. Do you also want to know something about why I like to stay with my grandparents?' This conversation ended with the conclusion that the student has a good relationship with his grandparents. In the post-test, the same teacher asked: 'If your mother would name two qualities you have, which ones would she mention?', 'What does that mean?' and 'Can you use these qualities in another context?' In other words, the student is challenged to discover a talent he is not yet aware of. Together the teacher and student conclude that the student is able to make people feel better using humour.
In the pre-test conversation, a teacher inquired about a student's possible future work in a pre-test conversation: 'Do you already know what you want to become in the future?' The student responded and then teacher and student moved on to the next topic of conversation. In the post-test conversation the same teacher asked: 'If you could be anyone for one day, who would you be and why?' Later on in that conversation, the student mentioned that she would like to work with animals. The teacher then asked her to elaborate by asking: 'Have you explored what kind of work you could do with animals?' and 'What kind of work-related activities would you like to do in working with animals?' Students are rarely challenged to practice what they want to learn or to prove where they are already capable. Furthermore, students are rarely challenged to formulate their thoughtsoften the teacher fills these in for them. After the training programme, the topics discussed in the guidance conversations were often more career related. For example, a teacher talking in the pre-test conversation about a student's poor grades in English, asked the questions: 'What subject do you like?' and 'Are you having fun in the classroom?' and took 6 of the 7.31 minutes to have that conversation. In the post-test conversation, which took 12.36 minutes, the same teacher asked questions like: 'What do you like about working with people?', 'What do you need to do to accomplish that?', 'Which steps need to be taken?', 'What skills and qualities do you think you need for work like that?' and 'What do you need to do now to find out whether this is a right choice?'.
When the conversation is about the student's course of study, the post-test conversation is more focused on the attitude of students and less on their marks and problems. More talking is done about their private lives, their extracurricular experiences, and the experiences students have that are related to their career. A teacher in the post-test conversation asked: 'If you consider [hobby] and [desired future work], is there some relationshipthings you are good at that you could use in future work?' Reflection on qualities and motives are addressed on a higher abstraction level. Such conversations provide input for self-image development and can promote self-confidence. More often than in the pre-test conversation, teachers talked with the students about work and about the actions students could undertake to realise their vocational wishes. These subjects were also more frequently addressed in relation to the student's future. A teacher who, in the pre-test conversation, talked only about the motivation of a student about an upcoming test, talked about learning from a future perspective in the post-test. The teacher asked questions like: 'Where do you see yourself in the future?' and about the internship of the student: 'Why do you see yourself in that specific place?' In response to an answer like 'this is a place where I can learn a lot' the teacher asked: 'What do you hope to learn?', 'Are these the things that you can use in your future?' and 'How?' In this way the teacher challenged students to think about what they can now do (and learn) which will help them build their desired future. Furthermore, work was not only addressed in the future sense, but also as it related to present experiences: 'Can you tell me about experiences you have in the music world? Did you explore work activities and demands?', 'What different kinds of things have you done?' and 'What do you learn from doing that?' Networking was not a topic of discussion in the pre-test conversation. However, in almost all posttest conversations networking was introduced. Networking is more about who students know, not about how to approach people or how they can expand and maintain their network.

Discussion
An important conclusion of this study is that a two-day off-the-job training programme for teachers was insufficient to achieve significant changes in guidance conversations, measured at a student level. However, off-the-job training combined with individual coaching and team coaching on-the-job, proved to be effective in improving guidance conversations from a student perspective. An actual improvement in guidance behaviours requires being guided in applying the off-the-job training in the teacher's own context. Not only the quantitative study showed changes in guidance conversations after the training programme, but changes were also seen in the recorded conversations. In other words, the training contributed to a change in often long-entrenched teacher routines. Teachers asked more career-oriented questions and students gave more careeroriented answers.
Although the training programme seems to be effective in making career conversations more dialogical, some caution about conclusions is in order as the study had several limitations. First, the experimental group was not fully comparable with the students in the control group. Because teachers guide whole classes, it was not possible to take students out of one class and randomly assign them to the experimental or control group. This problem was addressed by using classes of teachers, who were due to start their training later, as part of the control group. However, some schools were so small that there was no comparable control group or the conditions prevented a similar group from participating as a control group (e.g. students at work placements not being present at school to participate). A second limitation was that teachers provided only one recording of a career conversation to be analysed, therefore it was difficult to determine whether these recordings were representative of a particular teacher or not. Nevertheless, new skills and topics could be distinguished in the recorded career conversations; these new behaviours are not easy to fake. Moreover, the results from the analysis of the recordings were comparable with those from the quantitative study: students experienced the changes in teacher behaviours. The third limitation of the study, is that it cannot be guaranteed that teachers who did the training over a five-month period did not influence the learning of colleagues who did not take the training. In other words, it is possible that the results found among members of the control group were influenced by the training as well. And lastly, it is not clear whether the training programme would continue to have positive effects long term. Longitudinal research to examine the effects of the training over time is therefore important.
Despite the positive effects of the training programme and the fact that many teachers asked for this type of training in conducting career conversations, it is generally not easy to motivate teachers to participate in such programmes. A need for professionalisation is not self-evident to teachers (Van Driel, 2008). They often prefer short off-the-job training courses where they learn how to work with specific instruments. The training described here is not only more time consuming (a combination of off-the-job and on-the-job training), it also contributes to changes in one's professional identity. The development of personality traits and qualities only takes place when those who are learning to find the content meaningful (and that is something quite different than content being considered 'necessary'; see Hensel, 2010). In order to work towards meaningful learning within the dominant educational culture, collective learning (Lodders, 2013) and transformative leadership (Geijsel, Meijers, & Wardekker, 2007) is essential. Such leadership, however, is rare in education. The research results presented here therefore do not answer the question whether it is good for students to make teachers responsible for delivering career and employment guidance. It does, however, show that teachers can be trained if they are willing to be trained; whether successfully trained teachers will be able to apply what they have learned in the settings in which they work remains to be seen. Note 1. We note that the term 'counsellor' slightly differs from country to country. In this article, we make a distinction between 'guidance counsellors' (those who are trained as counsellors) and teachers (who may have some counselling training but whose focus is mainly on teaching a subject).

Notes on contributors
Marinka Kuijpers is professor at the Welten Institute of the Open University, the Netherlands. E-mail: mact.kuijpers@ caiway.nl Frans Meijers is professor at the School of Education, The Hague University of Applied Sciences, the Netherlands. E-mail: fj.meijers@xmsnet.nl

Disclosure statement
No potential conflict of interest was reported by the authors.