Identification with the SocialWork Profession: The Impact of Education.

The aim of this study is to examine how education affects students' identification with the social work profession. In particular, we examine the impact of students' experiences of their interactions with teachers, peers and supervisors at placement. A longitudinal design is applied. Data were collected from students during their first and third (final) year in social work education from seven universities and university colleges in Norway, representing a sample (panel) of 390 students. To evaluate the degree of students' identification with the profession, comparisons with student nurses are conducted. The analyses indicate (i) that students' dedication to and identification with the social work profession are largely established at a very early stage of education and (ii) that education has a positive impact on students' identification with the social work profession. At the end of their social work education, students who experience support and feedback from teachers and have confidence in their supervisors' competence express a higher degree of identification with the social work profession.

Introduction

Professional identification refers not only to values, attitudes and norms that may vary between professions, but also to affective aspects that are common for most professions. In this article, we emphasise the affective aspects and address students' dedication to and identification with the social work profession. Our intention is to examine whether social work education has an impact on professional identification. For the purpose of evaluation, social work students are compared with student nurses.

Social work education serves as a gateway to the profession. Even though social work education is expected to be important in professional socialisation, empirical research on this subject is limited (Barretti, 2004; Lishman, 2012). Some studies have examined the impact of social work education on students' attitudes, values and preferences, and the general picture is mixed—some indicate a substantial impact, but others do not (Pike, 1994; Barretti, 2004; Weiss ; Limb and Organista, 2006; Kaufman ). This contradictory result is explained by the following: a large variation exists in the research designs that have been used, none of the studies has been replicated, the sample sizes have been small and the studies have been often limited to a single school of social work (Pike, 1994; Barretti, 2004). One criticism put forward by an author of a meta-study is that empirical enquiry directed at the influences of role models, such as teachers and supervisors, and of student culture has been lacking (Barretti, 2004). Most studies have examined students at a single point in time, and longitudinal data and design are not common (Weiss ).

The aim of this longitudinal study, based on data gathered from seven schools of social work, is to examine how education affects students' dedication to and identification with the social work profession. This is a matter of professional socialisation and an aspect of students' desire to become a social worker. The role of education in professional qualification and socialisation is contested. The assumption of this study is that the degree of students' identification with the social work profession at graduation is affected by their interactions with teachers at college, peers and supervisors at placement. During education, students have to develop an understanding of the role of a social worker within a context of different traditions and approaches. An examination of how social work education affects students' professional identification is a contribution to the understanding of professional socialisation.

Identification with the profession

The concept of professional identification is frequently used, particularly in research on teacher education and nursing education, but without an explicit and common definition (Beijaard ; Lamote and Engels, 2010; Sims, 2011; Johnson ; Trede ). In this study, professional identification refers to students' degree of dedication to and identification with the social work profession.

According to social identity theory, individuals tend to classify themselves into various categories and social identification, which refers to an individual's affinity to certain human values. As such, social identification provides a partial answer to the question of who am I? and the answer is given in relation to a category (Ashforth and Mael, 1989). The decision to begin a certain education programme might be interpreted as a desire to become a member of a certain category of people—initially as a student and later as a member of a profession.

In addition, scholars have argued that identification with a profession refers to an ongoing process that is dynamic and relational. It is not only the answer to the question of who am I? but also the answer to the question of who do I want to become? (Beijaard ). It refers less to something that professionals have and more to a process that involves making sense of oneself as a professional. The process of identification occurs through interactions with role models, such as college teachers and placement supervisors, and with peers. It seems to be a question about professional socialisation—of the internalisation and creation of an identity.

However, all professions are characterised by an inherent tension. Externally, their legitimacy is linked to being an occupational group that is different from the others, whereas, internally, they consist of diverse approaches, traditions and methods (Abbott, 1988). On the one hand, social work is understood as a unified competence that is different from the competence of other occupations. This idea conjures up an assumption of the existence of common knowledge, skills and values, which constitute the competence of social work. On the other hand, what it means to act as a decent social worker is more contested. This situation opens up the field to consist of different approaches and methods. Thus, being dedicated to a profession and identifying with it involve aspects that are not only common and shared but also contested.

Does education matter?

When students start at university, they may have a more or less clear opinion of what it means to be a social worker. Some express distinct motivation and dedication to become simply a social worker, whereas others might express more diffuse enthusiasm and identification. During education, they might justify—or even change—their opinion of what it means to be a social worker. Thus, education is a meeting between students' pre-existing attitudes and the objectives of an education programme guided by teachers and supervisors.

In theories of professions, education has been generally ascribed an important role in professional socialisation and in the shaping of students' identification with future occupations (Merton, 1957; Parsons, 1978; Freidson, 2001; Barretti, 2004). Professional socialisation is traditionally understood as a process of internalisation of professional identity. According to Freidson, it ‘contributes to the development of commitment to the occupation as a life career and to a shared identity, a feeling of community or solidarity among all those who have passed through it’ (Freidson, 2001, p. 84).

The importance of education has particularly been stressed in structural functional perspectives, in which professional socialisation is understood as the internalisation of values and the development of an identity as a member of a profession. The major goal of education is understood to be the adoption of a professional role. This approach has also been termed the induction approach because it sees students as being induced into a role (Barretti, 2004).

Another approach, which is developed from perspectives in symbolic interactionism, tends to understand professional socialisation as the creation of a professional role. It refers to a process characterised by conflicts and contradictions rather than by harmony and uniqueness. Professional socialisation is understood as a recurring process of adjustment to various expectations and demands. The link between education and future professional performance has been played down. Rather than assuming that students assimilate a repertoire of professional roles and values, scholars have focused on investigating how students adapt to various situations in education (Vågan, 2009). This approach has also been termed the reaction approach because students are interpreted as active and conscious agents whose professional identity is being developed in their interactions with teachers and fellow students (Barretti, 2004). A crucial issue is how such potential tensions influence the creation of the identification with social work as an occupation and a profession.

The design of this study is inspired by the criticism that research on the impact of role models has been limited (Barretti, 2004). Social work education consists of various subjects and learning contexts, which are taught by teachers and supervisors with varied professional training and practical experience. The expectation in this study is that students' experience of their interactions with college teachers, placement supervisors and peers will affect their dedication to and identification with the social work profession at the end of their education.

Teacher–student interaction

Empirical research on student achievement and learning has indicated that how teachers organise and act is of great importance (Pascarella and Terenzini, 2005). This concerns factors such as instructional clarity, ability to give feedback and appreciation of suggestions from students. Because the growth in student enrolment has created more heterogeneous student groups, it can no longer be assumed that students share a common understanding of the purposes and the nature of studying at higher levels (Kyvik, 2009). This new composition probably places a particular emphasis on the role of the academic staff in supporting students to meet learning expectations (Ramsden, 2003; Gosling, 2009). The assumption in this study is that students who experience teachers as supportive and encouraging express a higher degree of identification with the social work profession.

Supervisor–student interaction

Numerous studies have described supervisor interaction as an important part of socialisation, involving the move from a position as a legitimate peripheral participant to that of a more experienced professional (Lave and Wenger, 1991; Wenger, 1998; Rønnestad and Orlinsky, 2005; Pungur, 2007). Placement, which is a larger or smaller part of most professional education programmes, is often students' first encounter with the professional field in their role as future professionals. An important aspect of the placement involves meeting the supervisor, a potential role model for a student's own entry into the profession. The supervisor introduces the student to clients, as well as to other members of the profession. The supervisor explains the range of challenges in daily routines, shows students how to prepare for work and how to handle different tasks, and advises students on how to learn from and deal with their experiences and how to prepare themselves better to face future challenges. Educational supervision, and particularly supportive supervision, is important for developing ‘attitudes and feelings conducive to the best job performance’ (Kadushin and Harkness, 2002, p. 277). The assumption in this study is that students, who have confidence in the placement supervisors' competence and experience them as role models, express a higher degree of identification with the social work profession.

Peer interaction

The importance of peers in the creation of the identity of students as future professionals has been emphasised in several studies (Freidson, 2001; Kaufman and Feldman, 2004; Rønnestad and Skovholt, 2013). Freidson (2001) described the effect of ‘a group of like-minded people’ and made a reference to Collins's (1979) concept of consciousness community regarding peer groups. If peers serve as role models, they could theoretically guide fellow students in different directions. Thus, peer interaction does not necessarily reinforce the ethos of the formal programme. Whether peer interaction contributes to strengthening or undermining the identification with the profession probably depends on the characteristics of the student group and of the context in which the interaction takes place (Pascarella and Terenzini, 2005). In addition, peer interaction consists of a social and an academic dimension, both of which we expect are important for the formation of professional identification. However, the assumption in this study is that students who experience peers as academically supportive express a higher degree of identification with the social work profession.

Strategy and data

Longitudinal design

Even though the choice of professional education can be interpreted as a choice of occupation, students' degree of certainty at entry varies. Dedication to a certain profession might be developed during adolescence; accordingly, some novice students are strongly dedicated to becoming social workers, whereas others are less so. Thus, when the intention is to learn more about the effect of education, a design that takes into account dedication to and identification with the profession at entry is desirable. In this study, data collected in the second and sixth semester are available, making it possible to compare students' scores on professional identification measured rather early and late in the education programme. However, it makes it impossible to differentiate between the impact prior to beginning the education programme and during the first semester.

Social work education

Social work education programmes in Norway are offered at universities and university colleges. University colleges in Norway share many similarities with universities; they are regulated under the same legislation, are administered under the same ministry and have a similar structure of degrees. The university college sector in Norway has been characterised as the most research-intensive college sector in Europe (Heggen ).

At the bachelor's degree level, there exist three different social work education programmes in Norway. All of them qualify for membership in the Norwegian Association of Social Workers. Most social workers in Norway have a bachelor's degree, which is offered by thirteen university colleges and three universities. In addition, some university colleges and universities offer master's and doctoral degrees in social work.

In this article, social work education refers to the three-year education programme at the bachelor's degree level (180 European Credit Transfer and Accumulation System (ECTS)), which is governed by national framework regulations. Approximately half of the programme covers theory and practice in social work, and the other half covers social sciences, social/health policy and law. Two-thirds of social work education instructors are educated in social work and have at least a master's degree, and one-third have at least a master's degree in related fields such as sociology, psychology, political science and law (Terum and Nesje, 2014). Placement is an important aspect and covers approximately 20–25 per cent of a three-year education programme.

Comparative design

An important limitation associated with examining only one group of students is the lack of an evaluation standard; without a comparison, it might be difficult to interpret figures. To interpret social work students' degree of identification with the profession, a comparison with student nurses is conducted.

However, these education programmes have similarities as well as differences, which might form a basis for developing assumptions about effects and correlations. Both are three-year education programmes, in which placement is a part of the education; however, placement represents a far larger proportion (50 per cent) of the nursing education programme compared with that of the social work education programme (20–25 per cent). In addition, student nurses are usually organised in groups of students in placement, whereas social work students are individually deployed. In both programmes, a large proportion of the instructors are educated in their respective professions, but the proportion is higher in nursing education.

Data

The data used in the article were drawn from StudData, a database of the Centre for the Study of Professions (CSP) at Oslo University College. Data were collected in 2005 and 2007 from seven different universities/university colleges in Norway for social work education and from two universities/university colleges for nursing education. The Norwegian Data Protection Official for Research approved the development of StudData. Students had the opportunity to refuse to participate in the survey, and respondents gave permission to connect data from the first and second collections.

The overall response rate in the sixth semester of education was 72 per cent, and the sample included 623 social work students and 343 nursing students. Of this gross sample of 966 respondents, 555 respondents answered a similar questionnaire in their second semester of education, which means that there exist panel data for 390 social workers and 165 nurses. The tests conducted indicate that the analysis based on the gross sample shows more or less the same results as the analysis on the panel data, which is particularly relevant for the model used in the regression analysis. This finding indicates that the effects shown are reasonably robust.

Variables and indexes

The dependent variable of dedication to and identification with the profession (shortened to professional identification in the tables) is measured in the second and sixth semesters. Items 1 and 2 in the index are very similar to items on the affective occupational commitment scale (Meyer ), whereas Items 3 and 4 are similar to items in the career commitment measure (Blau, 1988). All items referring to the affective aspects of students' relationship to social work—telling whether they are proud to inform others of the choice of education/profession, whether they would make the same choice again, whether they identify with the social work profession and whether they can hardly think of a better profession. In a factor analysis of three different samples, these four items were found to load on a single factor. The variable is based on a five-step scale (Cronbach's alpha = 0.72).

Teacher–student interaction (shortened to teacher interaction) is constructed of five attitude items (see Table 1). The five items explain ‘teacher–student interaction’ from the students' point of view—that is, whether teachers clarify the objectives of each course, whether they listen to and are interested in students' input, and whether they support students through feedback and supervision. The variable is based on a seven-step scale (Cronbach's alpha = 0.78).

Table 1

Operationalisation of indexes

Professional identification I am proud to tell others that I have made up my mind to become a social worker [nurse] I do not really identify with the social work [nursing] profession I cannot think of a better occupation than to work as a social worker [nurse] If I could make the choice again, I would not choose to become a social worker [nurse]

Teacher–student interaction Right from the start, the teachers clarify the objective for each course The students' ideas and suggestions are valued in the teaching The teachers are interested in what students think about issues relating to the profession The teachers challenge and support the students through supervision The students receive constructive feedback from teachers on work they have submitted

Supervisor–student interaction I received regular and systematic supervision. The supervisor were good role models. The supervisor gave helpful feedback on my own efforts. The supervisor encouraged me to review my own practice. The supervisor have helped me to integrate theory and practice. I had confidence in my supervisors' competence. I learned something from the supervision.

Peer interaction The teaching is organised in such a way that I learn a great deal from my fellow students There is a poor social environment amongst the students The students have a good opportunity to cooperate in groups There is a supportive climate amongst us students

Values on items are recoded from negative to positive for: ‘Professional identification’ Items 2 and 4, and ‘Peer interaction’ Item 2.

Supervisor–student interaction (shortened to placement supervisor interaction) is constructed from seven items connected to the placement. The seven items explain ‘supervisor–student interaction’ from the students' point of view—that is, whether they perceive supervisors as good role models, whether they have confidence in their supervisors' competence and whether they receive feedback during work placement and are stimulated to integrate theory and practice. This variable is based on a five-step scale (Cronbach's alpha = 0.88).

Peer interaction is constructed from four items and cover both a social dimension and an academic dimension of view; the variable is based on a seven-step scale (Cronbach's alpha = 0.60). The level of Cronbach's alpha for ‘peer interaction’ might indicate more than one dimension. Item 2 has a distinct social dimension, whereas Item 4 is a bit more ambiguous (Table 1). Item 4 may refer to both the social and academic climate. Hence, Item 4 has significant correlations to both Items 1 and 3, as well as to Item 2. Significant correlations exist between Items 1 and 3, between 1 and 4, and between 3 and 4, as well as between Items 2 and 4. The correlations between Items 1 and 2 and between 2 and 3 are weaker.

A factor analysis has confirmed that the convergent validities for ‘professional identification’, ‘teacher–student interaction’ and ‘supervisor interaction’ are satisfactory, which indicates that the items in each index refers to the same phenomenon. In addition, the factor analysis shows that these three concepts are not observed to be related to each other, which indicates satisfactory discriminant validity. Even though the convergent validity for ‘peer interaction’ is less satisfactory, we would theoretically argue that an examination of professional identification ought to include both a social dimension and an academic dimension.

Analysis

The analysis was performed in two steps. First, students' scores on dedication to and identification with the profession are examined, measured in early and late stages of the education programmes. Scores at the group level and the individual level are measured and compared. At the individual level, changes from early to late stages of the education programmes could be showed as increasing scores, decreasing scores and no change in scores.

Second, a regression analysis is performed, with variation of professional identification measured in the sixth semester as a dependent variable. Three independent variables—teacher interaction, supervisor interaction and peer interaction—are introduced. Regression analysis is performed to determine the impact of students' experiences of these interactions on the identification with the social work profession. Professional identification, measured in the second semester (early), and gender are the control variables.

Table 2 presents the mean and standard deviation values for identification with the profession for student social workers and student nurses, measured in early and late stages of the education programmes.

Table 2

Professional identification measured in the second and the sixth semesters of education

	Second semester		Sixth semester
	MV	SD	MV	SD	Difference	n
Social workers	3.78	0.72	3.85	0.76	+0.07	390
Nurses	3.62	0.76	3.68	0.86	+0.06	165
All	3.73	0.73	3.80	0.80	+0.07	555

Mean values (MV), standard deviation (SD) and number (n). The differences between professions are significant at the start (p = 0.017/F = 5,732) and at the end (p = 0.020/F = 5,480) (one-way analysis of variance (ANOVA) with the least significant difference (LSD) post hoc test). The difference from entry to end of study is significant (p = 0.035) for student social workers, but not for nurses (paired samples t-test). Scale: 1–5.

As shown in Table 2, the mean values in the sixth semester are slightly higher, compared with the second semester. In addition, the mean values are highest for student social workers. The standard deviation indicates equal variance for student nurses and student social workers. Changes in professional identification at the group level are not big.

The results in Table 2 are based on panel data, which contain information from students who answered questionnaires in both the second and the sixth semesters of education. This is a selected population because it does not include students who have dropped out or for other reasons did not answer the questionnaire in the sixth semester. A comparison of the mean values between the panel sample and the gross sample, measured in the second semester, could indicate whether those who dropped out/did not answer the questionnaire had higher or lower scores.

For student social workers, small differences are registered between the gross sample and the panel sample (3.74 and 3.78, respectively). For student nurses, more distinct differences are observed. The mean value was higher in the gross sample (3.77 compared with 3.62 in the panel data), which may indicate that the most dedicated students dropped out/did not answer the questionnaire at the end of the education programme.

As shown in Table 3, the pattern of changes in the identification is more multifaceted at the individual level. For social work students, 45 per cent had a higher score on professional identification at the end of the education programme, 20 per cent had the same score and 35 per cent had a lower score; student nurses showed a similar pattern. These results indicate that two-thirds of student social workers have equal or higher scores on identification with the profession measured at the end of the education programme. Further analyses of student social workers indicate that the proportion with higher score was about at the same level, independently of the level of the entrance score. In other words, the likelihood of getting higher scores was to a small degree influenced by the level of entrance score. The next step is to examine whether students' experiences of the interactions with teachers, supervisors and peers affect professional identification. As a preliminary step of the regression analysis, the correlations between these variables are calculated. Table 4 shows bivariate coefficients between professional identity early in education, teacher interaction, supervisor interaction, peer interaction and professional identity late in education.

Table 3

Changes in professional identification scores from the second to the sixth semesters in education

	Lower score (%)	Same score (%)	Higher score (%)	n
Social workers	35	20	45	390
Nurses	39	17	44	165
All	36	19	45	555

Significance level: not significant (Pearson).

Table 4

Correlation matrix of the variables

	Professional identification (early)	Teacher–student interaction	Supervisor interaction	Peer interaction
Teacher–student interaction	0.15**
Supervisor interaction	0.20**	0.20**
Peer interaction	0.12**	0.36**	0.11**
Professional identification (late)	0.62**	0.24**	0.20**	0.20**

n = 555. ** means level of significance <0.01.

Table 4 shows a significant correlation between professional identification measured early and late in the education programme, which indicates that identification with the profession was largely established in the second semester. The correlation between teacher, supervisor and peer interactions were on a moderate level, and highest between teacher interaction and peer interaction.

The dependent variable in the regression model is professional identification measured in the sixth semester. Controlling for professional identification measured at the early stage makes it possible to identify the net effect of the interactions with teachers, supervisors and peers.

Model 1 in Table 5 establishes ‘the baseline’; as expected, professional identification measured early in the education programme has a strong effect on identification with the profession measured late in the education programme. This is the case for both student social workers and student nurses, but gender shows no effect.

Table 5

Effect of education on professional identification

	Social workers		Nurses
	Model 1	Model 2	Model 1	Model 2
Gender	−0.04	−0.06	0.04	−0.04
Professional identification (early)	0.62**	0.58**	0.55**	0.49**
Teacher–student interaction		0.15**		−0.07
Supervisor interaction		0.11*		0.14*
Peer interaction		0.05		0.23**
N	333	333	138	138
R2	40.5	45.1	30.7	36.9
R2 change (F-test)	74.6	83.7**	19.8	24.1**

Significance levels: * p = 0.05, ** p = 0.01 marked with bold type. Model 2 increases the explanation of the total model. F increases from 92.6 to 104.9 (significance level: 0.01). Dependent variable: professional identification (late); control variables: gender and professional identification (early). Linear regression analysis for student nurses and student social workers—two models. Beta values. Model excludes cases pairwise.

In Model 2, when interactions with teachers, supervisors and peers are included, the explained variance (R2) increases significantly. For student social workers, a model improvement of 4.6 per cent is identified, which means that education has a significant effect on students' dedication to and identification with the profession. Model improvement is also the case for student nurses. (A similar analysis, but with change in professional identity between early and late stages as the dependent variable, shows the same pattern; however, the effects are generally weaker.)

Interaction with supervisors at placement shows significant effects for both student social workers and student nurses. These findings indicate that students who have confidence in the supervisors' competence and consider them to be good role models experience a stronger dedication to and identification with the profession.

For student social workers, interaction with teachers on campus shows the strongest effect, which indicates that students who experience to receive feedback and encouragement from teachers express a stronger identification with the social work profession. For student nurses, peer interaction shows the strongest effect, which indicates that the experience of a climate of co-operation and peer learning positively affects student nurses' identification with the profession.

Discussion

The aim of this study is to examine whether education has an impact on students' dedication to and identification with the social work profession. The assumption is that affective outcomes, interpreted as students' dedication to and their identification with the social work profession, are important for professional practice. Being proud to be a social worker expresses something not only about organisational affiliation, but also about motivation and the desire to do a decent job. Thus, it is important to examine whether education affects students' professional identification.

The analyses indicate (i) that professional identification seems to have been established at an early stage of education, (ii) that the level of professional identification at entrance largely predicts the level at the end of education and (iii) that, in addition, education has an effect on professional identification.

Because the first data collection was completed early in the second semester, it is impossible to distinguish attitudes established prior to beginning the education programme from those established in the first semester. If social work education represents a ‘gateway’ to the profession, there is reason to believe that an application for admission to a social work education programme expresses a desire to become a social worker. If so, many students have developed the desire prior to beginning the social work education programme. This implies that students at entrance have some assumptions of what it means to be a social worker. In that case, education represents an encounter between students' pre-existing perceptions of what it means to be a social worker and the ways this manifests by teachers and supervisors. They are disseminators and role models of what it means to be decent social workers, even though the links between education and the professional role are complex.

That the level of professional identification early on predicted the level at the end of education expresses a twofold challenge for the social work education: to maintain and further develop the identification and dedication among those who at entrance have high scores, and manage to increase the professional identification among those who enter with lower scores. In addition, it is important to emphasise that education matters.

The literature on the impact of social work education on students' attitudes and preferences is ambiguous and contradictory. This study indicates that education has a positive impact on professional identification for almost half of the students. The impact of education might have been even more significant if data had been collected when students started their education programme. Thus, the empirical evidence is in accordance with previous research that identified a positive impact of education on students' attitudes and preferences (Barretti, 2004; Weiss ; Limb and Organista, 2006).

For social work, this study shows that teacher–student interaction and interaction with supervisors at placement affects professional identification. Students who experience support and feedback from teachers and who have confidence in the supervisors' competence express a stronger dedication to and identification with the profession. This indicates, in accordance with previous research, that feedback from teachers and supervisors and the experience of good role models are important in professional socialisation (Lave and Wenger, 1991; Wenger, 1998; Rønnestad and Orlinsky, 2005; Pungur, 2007).

An interesting difference between nursing education and social work education is that school-based teaching is a more dominant learning arena in the social work programme. In addition, Table 5 indicates that students who experience college teachers who listen, give support and feedback increase the level of professional identification. A similar effect is not displayed for student nurses. This might explain why teacher interaction affects professional identification for student social workers.

In nursing education, placement covers a greater part of the education programme and, in placement, student nurses are usually organised in groups of students. Table 5 indicates that students who experience both the social and academic dimension of peer interaction increase the level of professional identification. A similar effect is not displayed for student social workers. This might explain why peer interaction for student nurses affects professional identification.

The regression analysis indicates effects, but the direction of the impact is not obvious. In this article, we emphasise ‘interactions’ because students are interpreted as active agents in the formation of professional identification. Not all students perceive supervisors as good role models or experience teachers with the ability to provide good and relevant feedback. This is not only an effect of how teachers and supervisors act and perform, but also an effect of how students act and react. Hence, students interpret the same teachers and supervisors differently.

Table 4 shows positive correlation between students' scores on professional identification measured early and interaction with teachers, supervisors and peers measured late. Because the effects assume a certain time order, and cannot go back in time, this indicates that students who are motivated and dedicated ‘read’ and interpret the interaction with teachers differently from students who are less motivated and dedicated. If so, self-reinforcing processes take place. Students who are initially dedicated feel they receive more feedback and support from teachers and more often perceive supervisors as role models; their feelings and perceptions in turn reinforce their commitment and identification with the profession.

The underlying assumption in this article is that professional identification is an important element of the social work education programme. The lesson to learn from this study is to have greater awareness of how to increase professional identification among students with low motivation and dedication at entrance, and still being able to maintain and cultivate professional identification of students who at entrance have high dedication and motivation.

Conclusion

Generally, students' dedication to and identification with the social work profession seem to a great degree to be established early in education. Social work education also fosters professional identification, particularly when students receive feedback and encouragement from teachers and consider supervisors to be good role models. Central to the professional socialisation is the interaction between teachers/supervisors and students, as well as the interaction among students. A challenge facing social work education is developing knowledge of how students' perceptions of teachers and supervisors are created and of how students' motivation and dedication can be cultivated.

Social work education, like most other types of professional education, is not paradigmatic. Social work education involves the organisation and teaching of a vast number of subjects by teachers and supervisors who represent different perspectives, traditions and approaches. Tensions might exist within subjects as well as between teachers and supervisors at placement. In this situation, students have to develop a professional identification as social workers through interaction with people who represent slightly different approaches and traditions. To handle this situation, students need to be introduced to openness about various approaches within the profession. Acting as a professional is basically about managing uncertainty, and a more explicit introduction to the diversity involved might make students better prepared to live with a certain amount of confusion and controversy. This might lead to the construction of a framework that is necessary for the creation of an identity as a social worker.