1 Introduction

“You have to compare”, says 12-year-old Ingemar in the film My Life as a Dog. This chapter’s main argument is the possibility of applying results from measurements of quality attributes in the various Nordic countries for the purpose of improvement. This is because differences in the Nordic countries’ teacher education provide empirical indications claiming that some countries’ policy-making functions better than that of other countries. This chapter deals with quality aspect comparisons related to teacher education in the Nordic countries. Different institutions carry out measurements that make it possible to compare quality aspects, partly on behalf of central education (for example, the Eurostudent survey) and partly as one permanent task vis-à-vis a public agency. Examples of the assessments include surveys from institutions such as the Norwegian Agency for Quality Assurance in Education (NOKUT). This agency runs the Study Barometer for Norway, the Swedish Higher Education Authority (Universitetskanslerämbetet), which conducts the Student Mirror, and the Ministry of Higher Education and Science, which conducts the Learning Questionnaire (Læringsbarometeret) in Denmark (2018). Finland and Iceland, as well as the autonomous territories of the Faroe Islands and Greenland, do not have these types of nationwide surveys for higher education. Notably, these surveys provide essential management information for each country’s education authority, and the measurements are used in public sector debates concerning teacher education and within the individual teacher education environment. This information can be used by the national authorities of the individual countries and networked improvement communities working to combine and spread ideas that provide innovation and learning and make teacher education more effective and efficient. To do this is Nordic collaboration model that includes coordination across nations, needed.

Teacher education has a learning-to-improve problem. In this chapter, the issue of whether it is possible to improve the quality of one teacher education institution is introduced, based on comparative improvement work. The basis of the problem is called improvement science (Bryk et al., 2015). The overall goal of improvement science is to ensure that quality work is based as much on evidence as it is on the best practices the quality work seeks to implement.

This chapter compares quality aspects in teacher education across Nordic countries. However, this endeavour is merely a starting point; a more robust methodology must be developed for continuous, systematic evidence-based improvement. Sustained efforts are necessary to advance improvements at large. However, this type of comparison for benchmarking processes has become commonplace in connection with large-scale surveys within school education, with the Nordic countries often compared in particular (Ludvigsen, 2016; Wester, 2018). The premise for such comparisons is that the Nordic countries have so many contextual similarities that it makes particular sense to compare them. The variation in contexts is, however, a challenge. Research designed to juxtapose relatively similar contexts is called “the most similar systems design” (Anckar, 2008). When the contexts have reasonably strong similarities, there is an analytical benefit to comparing them.

With the exceptions of TEDS-M and the Eurostudent project, systematic comparative measurements of the quality dimensions of different countries’ teacher education do not exist. Instead of comparative measurements of study quality between countries, the Scandinavian countries have chosen in-country comparisons of quality aspects of different types of studies in higher education. In Norway, this measuring instrument was named the Study Barometer. In Sweden, it is called the Student Mirror, and the Danish termed it the Learning Questionnaire. The questions used in these surveys have some similarities, but they are still so different that comparing them does not make sense. We argue that it would be sensible to compare the quality attributes of the Nordic countries’ teacher education programmes. However, for this argument to be valid, one must use the same measuring instrument. For this to be possible, cooperation is needed between the institutions of the Nordic countries that carry out such measurements as the aforementioned Study Barometer, Student Mirror and Learning Questionnaire. Such collaboration does not exist today. In the absence of such a standard Nordic measuring instrument, we present some tentative measurements utilising the same instrument in each Nordic country.

These in-country surveys can be valuable; however, for a country’s authorities, it should be of particular interest to compare relevant quality dimensions across countries and especially with countries with which it is natural to compare. Here, comparison across Nordic countries presents an interesting alternative. Failures emerge from the systems’ functioning; for instance, extreme fragmentation in Swedish campus teaching in teacher education programmes (Linderoth, 2016). Such fragmentation seems less apparent in, for example, Danish teacher education programmes. It is possible to reduce these harmful systemic variations by using consistent and sustained inquiry to drive improvement processes at a system level (e.g. forms of partnership arrangements between teacher education institutions and practicing schools or school mentor education). Nevertheless, the most meaningful level of improvement work is in the individual institutions that offer teacher education: the fabric of daily work is an arena of iterative refinements. The national comparison can, of course, provide an empirical basis for improvement, but we believe that ideas for improvement should also be spread across countries. In this chapter, results from a Nordic comparative study on selected teachers present the quality attributes in teacher education.

2 A Cross-Sectional Survey of Student Teachers in All Nordic Countries

The purpose of this chapter is to present the results of measurements of Nordic student teachers’ evaluations of the relevance of teacher education in courses like educational theory (pedagogikk) and subject didactics, along with selected dimensions of quality in practical training (practicum experiences). We also study Nordic student teachers’ time on task. Finally, we analyse Nordic student teachers’ turnover intentions.

The overall competence that a student teacher must attain involves several areas that are often separated in teacher education: education in subjects, educational theory, subject didactics and guided practice in schools. The weights assigned to these four components may vary along with the terms used to describe them, but all teacher education programmes include them (Hopmann, 2006). We aim to determine whether it is worth comparing different quality aspects of teacher education in the Nordic countries in the context of ‘benchmarking’. We collected data from all Nordic countries using similar surveys that were translated into the various Nordic languages ​​by native speakers. All student teachers who participated in the study had completed ‘long-term practice’ and were in the final phase of the teacher education programme. Therefore, we assume that they had the experience needed to perceive quality-related aspects of both campus teaching and field experience in practical schools.

A key question is whether the samples accurately represent each country. Details about sampling are found in Elstad et al. (2021a, b), Björnsdóttir et al. (2022), and Juuti et al. (2018). The questionnaires obtained from Norway, Iceland and Denmark were completed by student teachers from several teacher education institutions with relatively broad support and a high response rate. Thus, we believe that these samples reflect the essential characteristics of how student teachers in these countries perceive the quality aspects of teacher education; indeed, the Icelandic sample is almost a complete population survey. An obvious limitation of the samples from Sweden and Finland is that questionnaire data were collected from only one teacher education institution in each country. On the other hand, bias is reduced because very few people who were asked to participate declined to fill out the questionnaire.

Parts of the surveys administered in Iceland and Norway were distributed electronically, while the other countries’ students were given paper-based questionnaires. In the former case, students were asked to participate by email, with potential participants’ email addresses were provided by their teacher education institution. In the latter case, compulsory teaching sessions were conducted in which the paper-based questionnaires were distributed. A representative of the research project was present when the paper-based questionnaire surveys were completed. In our experience, no student teacher conspicuously refused to participate in the survey. While we cannot know for certain whether some student teachers chose not to submit the questionnaire, we believe that this was very unlikely. However, the question of whether the sample of each Nordic country is representative of its population must be investigated in future studies.

The purpose of this chapter is to perform exploratory rather than confirmatory research. We compare the central tendencies and spread of key concepts and study the empirical relationships between certain important variables. Quality dimensions are measured by several indicators (a multi-item approach) to identify common underlying concepts (Kline, 2004). We have begun the process of developing relevant instruments for use as benchmark indicators in teacher education programmes in the Nordic countries. However, we recognize the need to calibrate these instruments more finely for future benchmarking processes to provide a firmer empirical basis for comparison and benchmarking (Haladyna & Rodriguez, 2013).

Previous studies have compared teacher education between some or all of the Nordic countries (e.g., Afdal & Nerland, 2014; Canrinus et al., 2017, 2019; Elstad et al., 2021a; Hammerness & Klette, 2015; Hammerness et al., 2016; Klette et al., 2017). However, to the best of our knowledge, quantitative analysis of the quality aspects that comprise our focus have not been performed for all Nordic countries.

3 Nordic Student Teachers’ Evaluations of Educational Theory, Subject Didactics and Practice Training

Table 12.1 shows Nordic student teachers’ evaluations of educational theory, subject didactics and practice training. A seven-point Likert-type scale was used in the questionnaire (1 = strongly disagree, 4 = neither agree nor disagree, 7 = strongly agree). The individual questions associated with each of the aggregate variables in the table are presented in Appendix.

Table 12.1 Nordic student teachers’ evaluations of educational theory, subject didactics and practice training
Full size table

The central conclusion we draw from the results is that there are astonishingly similar empirical patterns among the Nordic countries in terms of both central tendency and spread; with only a few exceptions, the differences are relatively small. The results of this study do not provide empirical evidence that teacher education in a given country is qualitatively better in terms of any dimension than in any other country, in contrast to prior analyses of intentions related to teacher education in different countries (see, e.g., British Educational Research Association, 2014; Darling-Hammond, 2017). This leads us to believe that teacher education environments in the Nordic countries face fairly comparable challenges and dilemmas created by relatively comparable situations. Because the similarities are so great, we believe that a detailed comparison of all Nordic countries is justified.

All Nordic countries reveal a perceived greater relevance of subject didactics than educational theory. We interpret this primarily as a consequence of the situational logic of teaching (Popper, 1976). In educational theory teaching, examples that are relevant to teaching practice are taken from several subjects, but this is not the case with subject didactics teaching. However, it is inappropriate to state that educational theory teaching is consistently poorer with regard to relevant experience than subject didactics teaching; educational theory simply has different specific conditions and serves a broader set of purposes.

Educational theory in teacher education also has similarities across the Nordic countries. Therefore, it may be relevant to examine, for example, what makes educational theory more relevant in Danish teacher education than in other Nordic countries. We believe that measurements are a better starting point for comparison than intentional documents and interviews (see, e.g., Andersen et al., 2017), as they can serve as a base to try to go beyond numbers (e.g., by analysing interview data) to find plausible explanations for measured differences.

It is of interest to look at the highest average values among the Nordic countries. If we assume that those countries are largely comparable, the highest average value of a quality dimension shows what could be achieved in the other countries and is thus an indicator of good practice. Regarding the perceived relevance of educational theory and subject didactics in teacher education, both average values ​​are highest in the Danish sample. These results align with evaluations performed by an expert group:

The quality and relevance of teacher education in a number of areas [is] satisfactory.[…] It […] succeeded in making teacher education more relevant to some of the key needs of customers, including in relation to the work of classroom management.[…] Teachers’ basic professional skills (educational theory and teacher professionalism as well as general education/Christian Knowledge, Life Enlightenment and Citizenship) constitute a new subject area in teacher education from 2013, although pedagogical subjects and subject areas have long been an integral part of education. It is a pervasive point in the evaluation that the professional colleges work actively to ensure a connection between the teacher’s basic subject and the teaching subjects. Both teachers and students feel that the context is being sought. (Danish Agency for Research and Education, 2018, pp. 10–12)

One unanswered research question we pose is whether Danish teacher educators teach educational theory and subject didactics in ways that are particularly interesting and inspirational.

Our starting point is the belief that everything can be improved, including teacher education programmes in the Nordic countries. For example, the Swedish School Commission (SOU) finds that there are reasons for ‘strengthening the integration between theory and practice in teacher education’ (SOU, 2017, p. 35). For Norwegian teacher education, the relevance of educational theory to students’ field experiences is emphasized in current policy: ‘There is [...] a great potential for creating a better connection between theory teaching and practice training in teacher education’ (Norwegian Ministry of Education and Research, 2014, p. 42) and ‘there are indications that [...] the practice relevance should be strengthened. In general, the distance between campus and practice is still large’ (Norwegian Ministry of Education and Research, 2017, p. 11). Our results show that the relevance experienced in educational theory teaching is clearly lower in Norway than Denmark: the average scale value is around 4.75. In addition, the results reveal major differences in teacher education between institutions in Norway. The critical question is also quite straightforward: Is it possible to learn from good practice through benchmarking processes? For example, what makes Danish student teachers find greater relevance in educational theory than Norwegian student teachers? This question might be investigated by focusing on the interaction between several factors, including teaching materials and how the teaching of educational theory is organized.

The biggest difference between average scale values is found between the Danish and Finnish data sets. There is a standard deviation in difference in favour of Denmark – and thus a significant difference – between the relevance of educational theory teaching experienced in Denmark and Finland. Since the relevance of experience in subject didactics education in Finland is roughly as high as in the other Nordic countries, this value for educational theory cannot be an expression of a general response modesty in the Finnish sample. On the other hand, only very good students are admitted to Finnish teacher education programmes, and these students can be thought of as discerning. Thus, contextual differences must be considered. One could ask whether it is fair to compare the relevance of educational theory between countries; this question should be addressed in future research.

The Finnish results can be interpreted based on the following information from the teacher education institution. In years 1–6 of the teacher education programme, student teachers have short courses on subject didactics for each subject. Professional student teachers will typically attend courses in two school subjects (e.g., Finnish and the natural sciences) and hence two subject didactics courses. In the teacher education programme, there are several courses focused on the subject of teacher education, which is called educational theory. These courses are the same for all student teachers in Finland and cover topics like general didactics and educational theory, developmental psychology and diversity in education (multicultural education, special education and general content regarding human skills). At the time of data collection, these students had taken educational theory courses focusing on educational philosophy, educational history and sociology, curriculum theory and argumentation in educational contexts. This information is important for understanding the Finnish results for educational theory. The sample includes students from both general and subject-specific teacher education.

If we assume that there are real measurable differences between student teachers’ perceptions of the quality of Finnish teacher education and teacher education in the other Nordic countries, we must say that teacher education in the Scandinavian countries faces significant challenges in terms of how society perceives education. Our measurements and other relevant research may enable appropriate communications about teacher education in the Nordic countries, and a knowledge base can be developed to contribute to an open and informed conversation about the strengths and challenges of teacher education.

Measurements of the relationship between theory and practice are fairly similar for all the Nordic countries. The indicator ‘In the practice periods, I discuss practical experiences with the supervisors/practice teachers in light of what we have learned’, for example, has an average value for all Nordic countries slightly above the neutral midpoint (four) but is highest in Sweden and Iceland. Although it is not easy to compare this study and Canrinus et al. (2017), both identify similar differences between Norwegian and Finnish teacher education.

One quality aspect that scored below four in all Nordic countries is the extent to which the practice supervisor is well acquainted with what student teachers learn about educational theory and subject didactics and thus the extent to which the supervisor clarifies the relationship between educational theory and the student teacher’s experiences in the classroom. In this study, this aspect is referred to as ‘practice supervisor’. The degree to which the practice supervisor has the opportunity to become acquainted with, for example, the campus curriculum is important. If practice supervisors have limited opportunities to familiarize themselves with the current curriculum literature and the theoretical principles emphasized in campus teaching, it is unreasonable to expect them to relate their guidance to the curriculum literature. In an ideal world, however, practice teachers would know the curriculum literature, which would enhance the perceived coherence between campus teaching and practice for student teachers, a goal in all Nordic countries, as the Swedish School Commission acknowledges: ‘It is important to enable the teachers and other teachers of the schools to collaborate with universities and colleges in their employment’ (SOU, 2017, p. 35).

In all Nordic countries, student teachers are more satisfied when they receive personalized support (for example, ‘The tutorial talks at the practical schools help me understand what I need to do to improve myself as a teacher’) from a practice supervisor than when the practice supervisor is able to link the content of guidance to the literature in the syllabus. This aligns with an expert evaluation of Danish primary school teacher education: ‘The students are overall satisfied with the internship’ (The Board of Research and Education, 2018, p. 12). Even though practice supervisors may have few opportunities to become acquainted with the theoretical principles of campus teaching and the curriculum literature, they are still perceived as supportive of student teachers’ practice.

The academic study requirements received scores somewhat higher than the neutral midpoint (4) in all Nordic countries except Sweden, where study requirements are generally lower than what is desirable (University Chancellor’s Office, 2015). The indicators we used are not ideal, but our results are consistent with those in other studies (e.g., Lärarnas Riksförbund, 2016; The Swedish Confederation of Professional Employees (TCO), 2013; Universitetskanslersämbetet, 2017) and indicate the need to increase the academic requirements at the Swedish teacher education institutions that provided our data material. The Danish Reform Agreement (2012) expressed an ambition to strengthen the intensity with which student teachers study. Additionally, the expert evaluation of Danish primary school teacher education points out that

with regard to the academic requirements of the students, the expert group considers the achievement of the target to be average. On the one hand, the introduction of competency goals places high demands on students, but the extent of goals has led to disadvantages related to evaluation and in-depth learning for the students. (Danish Agency for Research and Education, 2018, p. 32)

Our analysis of the Danish data set generally corresponds to this evaluation, but the academic requirements are lower than what we measured in the Norwegian and Icelandic data sets.

The reported noise and disorder while teachers observe or teach was highest in the Swedish data set. The differences in student teachers’ perceptions of noise and disorder in Norwegian and Swedish classrooms roughly correspond to the data obtained from the most recently published PISA survey (Kjærnsli & Jensen, 2016). In the other Nordic countries, teachers’ perceptions of noise and disorder differ somewhat from students’ perceptions of the same. This can, of course, depend upon the specific classrooms in which teachers practice. Too much noise and disorder during student teachers’ practice teaching is empirically related to weaker expectations of mastery of classroom management (Christophersen et al., 2016; Juuti et al., 2018). Thus, it is not beneficial for student teachers to experience the most demanding classroom situations in their practice teaching.

Based on the measurements discussed so far, the word ‘crisis’ is hardly a general characteristic of quality dimensions in the Nordic countries’ teacher education programmes. The negativity with which teacher education in Scandinavian countries is discussed indicates that teacher education environments may face public relations challenges. The term ‘crisis’ should be reserved for the case of providing teachers with adequate competence. It is the responsibility of the Nordic states to ensure that sufficient education is provided and that teachers develop adequate competence.

We believe that the samples capture significant trends in student teacher populations, although we cannot claim that the samples are representative. The possibility of sample bias must be investigated in future studies, especially those based on representative samples, so that conclusions can be drawn with more confidence about the samples’ reference populations and quality differences between Nordic countries. Nordic comparisons can also be useful for setting benchmarks for education policy with the aim of improvement.

This comparison of the central tendencies and spread of different quality aspects has revealed nuances in the Nordic countries that should be carefully considered. We note above that there is more than one standard deviation between Finnish and Danish student teachers’ evaluations of the relevance of educational theory teaching to practice. This significant difference in Denmark’s favour will likely surprise many readers, as Finnish teacher education has been characterized as ‘outstanding’ and ‘excellent’ by international organizations (e.g., OECD, 2012, 2016) and other evaluators (e.g., Darling-Hammond, 2017; McKinsky, 2007). In addition, several books by Pasi Sahlberg that deal with Finnish teacher education and Finnish schools have received significant international attention, giving a positive image to Finnish teacher education. Renowned educational scholars who publish in high-ranking scientific journals have also emphasized the high quality of Finnish teacher education (e.g., Darling-Hammond, 2017; Tatto, 2015).

Our study does have some limitations. First, it should be emphasized that our questionnaire instrument may not capture significant quality aspects (see, e.g., Afdal & Nerland, 2014; Andersen, 2015; Lillejord & Børte, 2017). For example, Afdal and Nerland (2014) found some marked differences in the way teachers talk about their professional knowledge and the way they approach challenges in practice, but this was not captured in our study design. It is possible that the instruments we used should have been fine-tuned to tap into more significant qualities that were lost in the first round of our research. On the other hand, Afdal and Nerland’s (2014) findings are based on interviews with 12 newly qualified teachers in Norway and Finland, while our study is based on a much larger sample. When there is discrepancy between studies (for example, this study and Andersen et al., 2017), it would be interesting for future research to explain these differences to achieve a more secure empirical basis for qualitative evaluations. Elstad et al. (2021a) explore differences between the Finnish and Norwegian teacher education arrangements through structural equation modelling and ask whether a common model approach is better than an approach with separate models? The results of testing whether separate models for Norway and Finland fit better than a common model show that the common model approach fits just as well as the separate model approach. One interpretation of this inference is that the statistical associations in the data sets from Norway and Finland follow the same main patterns. These results emerge through an analysis of quality aspects of teacher education in the two countries, as the student teachers perceive quality differently.

We have argued that it is scientifically worthwhile to compare quality aspects of teacher education programmes in the Nordic countries. However, it remains unclear whether this type of analysis and subsequent benchmarking can lead to constructive improvements in the design of teacher education programmes. This topic should be addressed in future research, particularly longitudinal studies.

4 Nordic Student Teachers’ Time on Task: Comparisions

Time on task in higher education has emerged as an increasingly important aspect of the debate over good education. In order to develop both teaching competence and a robust identity as a teacher, student teachers must be present and actively participate in organized teaching (SOU, 2017, p. 35). The Swedish Higher Education Authority (2015) assumes that student teachers should normally study for 37.5–45 h per week, while the Danish Ministry of Education and Research states that 43 h per week constitutes full-time study. The Norwegian authorities also believe that the time on task in higher education as reported in some studies is too low, but they have not established a numerical threshold. Storting representatives from the Labour Party believe that ‘the time spent by student teachers on studies (should) rise to the average of students otherwise, preferably more’ (Stortinget, 2019a).

The results of different time-on-task surveys in the same country may vary, partly because such surveys are based on informants’ memory and thus involve estimates. Indeed, few informants may accurately report their time on task. One could argue that mistakes will be revealed by obtaining information from many informants. However, there may be systematic sources of error in time-on-task studies as different studies are organized differently. This may lead to error if, for example, highly structured studies with a high incidence of laboratory use (e.g., those conducted by pharmacy students) are compared with studies with a high incidence of self-study. Politicians have acknowledged this: ‘The majority of the committee [...] does not deny that there may be a correlation between students’ time on task and student satisfaction and accomplishment, but questions whether a comparison of time on task between highly different educations is relevant as an indicator of the quality of teacher study and students’ efforts’ (Stortinget, 2019a, b).

The time when a time-on-task measurement is performed may be another source of error. For example, there is evidence to suggest that student teachers’ time on task is higher in practice periods than in campus teaching (Martinussen & Smestad, 2011). However, this source of error can be avoided by clarifying that questions refer only to the campus situation. The source of error in time-on-task studies may be less significant if studies with the same characteristics are compared (the approach taken in this chapter). We reduced the problem of discretionary estimates by splitting time on task into predefined categories: lectures, seminaries, colloquia and individual study work.

Student teachers’ time on task has become important in political debates. The following contributions from Storting representatives provide an example of the debate in Norway:

Teacher education also stands out negatively in terms of study intensity, i.e., how much time the students spend on teacher studies on average. The 2018 study barometer shows that while most students in Norway study around 35 hours a week on average, student teachers spend about 5 hours less than this: 30 hours in elementary teacher education for grades 1–7, 28.5 hours in elementary teacher education for the 5th through 10th steps and 30.8 hours in associate’s education. There is a clear connection between organized activity and self-study, and teacher education is among those with the least self-study. (Stortinget, 2019a, b)

Many consider time on task to be an indicator of student quality, and some evidence supports this notion (Karweit, 1984). Furthermore, it is believed that learning progress depends on the quality of campus teaching, the student teachers’ own motivation, prerequisites for performing tasks associated with the teaching profession, the quality of the guidance students receive in practice teaching at schools and their own learning work while they are students. These factors are interconnected, which makes it difficult to isolate the significance of a single factor in a way that clearly indicates the quality of a study.

The amount that student teachers learn depends on, among other things, how much time is actually spent on learning work and the effort that students put into working with the learning material. The relationship between time on task while studying and achievement is a controversial topic in both educational research and other types of performance research, but it is widely recognized that this relationship is not linear (Berliner, 1990). If the student teacher does not work effectively, there is no automatic clear connection between time on task and the development of knowledge. On the other hand, it is difficult to envision learning progress without spending sufficient time on learning work. Here, time on task during study is regarded as a condition for learning in campus teaching. We argue that reasonably high time on task during study and in the campus-based part of teacher education are important prerequisites for campus education to function as intended, to create a foundation of competence so that student teachers can teach academic content and direct students’ learning work during the practical periods and later as professional teachers and to contribute to change competence; that is to say, the competency provider is sufficiently robust to enable the student to later relate to changes that occur as a professional teacher. In this chapter, we refine the problem area, focusing on student teachers’ time on task while studying on campus.

The time-on-task measurements used in research are controversial (Fisher & Berliner, 1985; Rosenshine & Berliner, 1978; Walberg, 1988). It is frequently used in mission research (e.g., Das Deutsche Zentrum für Hochschule- und Wissenschaftsforschung (DZHW), 2018). Assignment research has found that the time-on-task profiles vary not only in different types of studies but also systematically between countries. For example, one study showed that the average time on task for organized education (including higher education) is lowest in Sweden. at 10 h per week, with the other Nordic countries ranging from 13 to 19 h per week. In contrast, the number of hours of self-study is highest among Swedish students. However, the measured time on task among Swedish students varies significantly by type of study. This justifies a new look on how time on task is divided into different types of activities – lectures, student-led colloquia, teacher-led seminaries and individual studies – and a comparison of the total time on task in studying (calculated as the sum of lectures, seminaries, colloquia and self-study).

Student teachers’ motivation to study has been highlighted as an important factor. Motivation requires goals for one’s own actions, and motivation researchers have identified different types of motivation (external, intrinsic and altruistic) for choosing teacher education and thus the teaching profession (Balyer & Özcan, 2014; Kyriacou & Coulthard, 2000; Roness & Smith, 2010; Thomson et al., 2012). External motivation can be divided into several subcategories (Ryan & Deci, 2000), but generally, recognition will result in motivation (often referred to as achievement goals; see Elliot & McGregor, 2001). Intrinsic motivation is an inner drive to perform a task. For example, a student teacher may want to become a teacher because he or she wants to help students learn and is excited by the profession. An altruistically motivated teacher views teaching as socially worthy and important work. This is an important category of motivation for student teachers (Roness, 2011). However, in many cases, student teachers must find a balance between several types of motivation (Struyven et al., 2013), making it necessary to investigate altruism, recognition and study absorption more closely.

A number of studies have found that self-discipline (also called self-control) is of great importance for the completion of studies, study performance and other factors (e.g., Tangney et al., 2004). Self-discipline can be understood as a more or less permanent personality trait (Blickle, 1996), but no one is inextricably locked into a personality. Except in very special situations, everyone has choices and thus the opportunity to change their behavioural patterns. In the context of this study, self-discipline is treated as a driver of individuals’ ability to carry out their studies (Zimmerman, 2002) and may be triggered by academic requirements. Therefore, it is important to study self-discipline and academic requirements in context (Christophersen et al., 2015). In the survey, we also included questions about the degree to which student teachers had teacher role models. Furthermore, we included two different goals related to mastering expectations about teaching during practical periods (engaging students in academic work and the ability to keep order in teaching situations, which we call self-concepts 1 and 2, respectively, in Table 12.2).

Table 12.2 Results of various aspects of Nordic student teachers’ motivation and time on task
Full size table

So far, we have mentioned factors that can be considered characteristics of individual students and that we believe are related to time on task while studying. However, student teachers’ time on task can also be understood as a response to demands and tasks that require student teachers to perform in the study programme. For example, some parts of the teacher education programme may require students to submit assignments that are corrected and evaluated by teacher educators, who can demand greater effort in their feedback on student presentations and assignments.

Table 12.2 shows the results of various aspects of Nordic student teachers’ motivation and time on task. A seven-part scale was used in the questionnaire (1 = strongly disagree, 4 = neither agree nor disagree, 7 = strongly agree). With the exception of the term ‘turmoil in practice’, a high average value indicates high quality. The individual questions included in the aggregate variables are presented in Appendix. Table 12.3 shows the values of the time on task we included. As the Swedish survey used a different length of time than a week, we converted the Swedish values to hours per week.

Table 12.3 Accepted values when estimating time on task
Full size table

Time on task among Swedish student teachers shows that a relatively small proportion of the time on task is used for teacher-led teaching. This aligns with previous Swedish studies that problematize this topic (see The National Union of Teachers in Sweden, 2016; The Swedish Confederation of Professional Employees (TCO), 2013; The Swedish Higher Education Authority, 2017). The average amount of time that Swedish student teachers engage in teacher-led teaching is 9 h per week (DZHW, 2018), while self-study and group work combined constitute 22 h per week (The Swedish Higher Education Authority, 2017). In our survey, Norwegian, Swedish and Finnish student teachers spend about the same time on campus studies, but the distribution of study activities is different. The time on task in Danish teacher education is far higher than in the other Nordic countries.

We find somewhat higher time on task in the Danish data set than Rambøll (2018), in which student teachers were asked the following:

How many hours did you spend on the following activities during an average teaching week (seven days) in the last semester?

  • Educational activities in which a teacher participates (lectures, team teaching, tutoring, practice lessons, etc.)

  • Teacher-initiated preparation (reading syllabus, doing homework, exercises or assignments individually or in reading groups, analysing data, other academic activities)

  • Self-initiated preparation (e.g., reading supplementary literature and independent studies)

  • Activities that are related to education but are not directly related to a subject (e.g., participation in professional, social or political student organizations)

  • Study-relevant work

Rambøll (2018) found that a group of ‘ordinary students’ spent an average of 30.3 h per week on these activities. If we deduct activities related to education that are not directly related to a subject, the average number is 28.8 h per week. This questionnaire survey was conducted by Rambøll (2018) among all fourth-year students teaching in a primary school who were enrolled in a teacher education programme in Denmark at approximately the same time as we conducted our survey (December 2017 and January 2018). Of 2222 students asked, 792 answered Rambøll’s (2018) questionnaire, corresponding to a response rate of 36%. We do not know whether the dropout rate was random.

Our survey was conducted at two large teacher education institutions in Denmark, and our response rate was close to 100%. No student who participated in compulsory education chose not to participate in the study, but those who were not present in the teaching situation were obviously not included. The choice of only two professional colleges may result in a sample bias even though the response rate was so high. We also acknowledge that the results do not matter much in the two time-on-task surveys. Our measurements were carried out with the same instrument for measuring teacher education in all Nordic countries. Our surveys indicate that the average in the Danish data set is clearly the highest among the Nordic countries (although Iceland was not included in the time-on-task measurements because a significant number of students engaged in part-time study and distance education).

The Danish press reported that students in Danish primary school teacher education spend an average of 24.6 h on study activities, and Education and Research Minister Tommy Ahlers argued that this was not enough (Politiken, 2018):

I am concerned about the academic level and especially the study intensity, which is too low. Teaching is an incredibly important job, and we need dedicated and academically strong teachers who can give our children a good start in life, so that they have the skills and desire to take a youth education.

Our survey shows that the Danish student teachers excel in terms of time on task while studying compared to student teachers in other Nordic countries. Indeed, the average number of teaching and supervision hours increased following the 2012 Danish teacher education reform (Rambøll, 2018).

Student teachers’ time on task while studying has received a great deal attention in the Nordic countries. Our study shows that student teachers’ time on task is similar in Norway, Sweden and Finland but clearly higher in Denmark. We have previously analysed the precursors of time on task among Norwegian student teachers using structural equation modelling and showed that academic requirements, self-discipline and external recognition are statistically related to time on task, while altruistic motivation is not (Christophersen et al., 2015). One possible conclusion to draw from these results is that if time on task is too low, it can be increased by introducing higher professional requirements. Time on task can thus be seen as the response of rational decision makers to situational conditions (Elstad, 2003). However, should the time on task be increased through academic requirements (such as multiple assignments, more exams or more time-intensive work), changes in resource use or additional resources to fund such initiatives are required. The Norwegian government has contributed to the financing of time-intensive forms of work and expressed its expectation that teacher education institutions will follow up with students to ensure higher time on task (Nybø, 2019):

How we can make student teachers spend more time on their studies is a recurring theme in our communication with the sector. The Ministry meets universities and colleges with clear expectations and has provided funding for the sector to identify ways of working that can promote higher study intensity and student activity.

The Minister of Higher Education is correct in noting that the ways in which students study may impact their time on task, but other factors must also be highlighted, such as the types of exams they are given and the degree to which the seriousness of life is reflected in study requirements and exams (Elstad, 2003). Future research must investigate whether these measures lead to the desired results.

5 Which Factors Are Related to Nordic Student Teachers’ Turnover Intentions?

The dropout rate of student teachers is a problem in all Nordic countries – and in many other places – but to varying degrees (Sutcher et al., 2019). Dropout in a study programme can be measured as the difference between the number of registered students in a programme in the first year of study and the number of registered students who complete the programme. Cumulative dropout can also be measured throughout the period of study. However, measuring dropout is complex, as some student teachers complete their studies many years after the bulk of their cohort has graduated.

The reasons why student teachers quit are also complex. The decision to drop out may be related to many factors, such as personal reasons (e.g., the kindergarten attended by their own child is too far from the teacher education institution) or to pursue other studies (the so-called opt-out). In this chapter, we focus on the process and thoughts that precede the choice to quit teacher education. The path from intention to action involves multiple mental transitions (Ajzen et al., 2009), meaning that student teachers can think about quitting before – perhaps long before – making the final decision to leave.

The purpose of our analysis is to seek factors that may be empirically related to thoughts of quitting, including perceptions of the quality of campus teaching in educational theory and subject didactics, perceptions of the integration of theory and practice into practice guidance, how the practice supervisor supports the student teacher’s development of mastery of teaching assignments, perceptions of the academic requirements of the teacher education programme, various aspects of teaching (including expectations of mastery of engaging students and classroom management), student teachers’ engagement in the programme, their motivations (grounded in altruism and recognition), the importance of experiencing role models in teachers and their total time on task in their studies. First, we describe the situation in the five Nordic countries in terms of recruitments and dropouts; we then present the correlation coefficients between the aforementioned factors and turnover intentions. Finally, we report implications for further research.

5.1 The Situation in Different Nordic Countries

The labour market for teachers differs widely around the world. In the Nordic countries, varying proportions of principals are reported to face difficulty hiring people they consider well-qualified teachers (OECD, 2016), indicating that there may be geographical differences within individual Nordic countries. Teacher shortages are usually defined as the inability to fill vacant teacher positions at current salaries with people qualified to teach the subjects and levels in question.

Finland

In Finland, teacher education programmes are popular. However, the total number of applicants for these programmes has decreased over time, from 24,466 in 2015 to 19,601 in 2018. About 72% of applicants are asked to take an entrance exam, and about 34% of those who took the entrance exam enrolled in a teacher education programme. In 2018, most Finnish universities had to admit applicants with slightly weaker results on the entrance exam than in previous years (National selection cooperation network in the field of education (VAKAVA, 2019)). There are no official statistics regarding dropout from teacher education programmes, but several sources (e.g., Bjerril, 2019; Darling-Hammond, 2017) indicate that the dropout rate is low. However, others state that some students delay completing the programme for up to 10 years (Juuti, 2019). Despite this, a lack of teachers has not been identified as a problem in Finland. In fact, there have been times when there were too many teachers.

Sweden

As of 2011, in order for a teacher to gain permanent employment and assign grades independently in Sweden, they must have a teacher certification. One-fifth of certified primary and secondary school teachers are 60 years of age or older, and in many municipalities, the proportion is even greater. Thus, many teachers are expected to retire in the next few years. In the 2018–2019 school year, 70.5% of primary school teachers were qualified and had credentials (Skolverket, 2019a, b), indicating a notable lack of fully qualified teachers. The number of learners is expected to increase by 350,000 over the next decade, an increase of 15% over 2017. In total, schools must recruit 187,000 full-time teachers by 2031 (Statistics Sweden (SCB), 2018; Swedish Association of Local Authorities and Regions SKL, 2018). To meet Sweden’s future need for competent teachers, the volume of admissions to teacher education programmes must increase by a significant proportion(Swedish Association of Local Authorities and Regions (SKL), 2018).

This is major challenge for Swedish society, and there are several indications that Swedish authorities have given up on teacher credentials as the main model for education. In June 2019, a measure called ‘Introduction for unauthorized teachers in the school’ was implemented (Skolverket, 2019a, b). The initiative was based on texts, films and links to in-depth information about, among other things, school tasks and management documents, teachers’ assignments and how to provide effective and varied teaching. Thus, a situation in which a significant proportion of teachers lack adequate education is being normalized.

The number of applicants to Sweden’s teacher education programmes has not been sufficient to meet the future need for teachers. This has allowed many students with very low grade point averages to be admitted to these programmes. To overcome this, in 2021, grade requirements for admission to teacher education programmes, similar to those in Denmark and Norway, will be implemented.

Whether this measure will lead applicants with stronger higher education records finding teacher education programmes more attractive remains unknown, as this is an empirical question that is impossible to answer today. A report from The Swedish Higher Education Authority (2015) shows that higher education grades are statistically related to both dropout and student achievement in teacher education programmes; teacher education students with good grades from higher education tend to complete the programmes and achieve higher grades. The Swedish Higher Education Authority (2017) notes that continuing to expand the volume of teacher education applicants is likely to be difficult if higher grade requirements are implemented, but that the dropout rate may be reduced if the intake requirements are increased.

In order to meet the future need for teachers with an increased supply of newly qualified teachers, 25% of all upper secondary school graduates would have to enrol in a teacher education programme (Bergling, 2017). This is scarcely possible, let alone likely. However, there are 40,000 educated teachers who are no longer in schools, comprising 14% of all educated teachers. About 60% of this population may want to work as teachers again if working conditions were suitable (SCB, 2017).

Therefore, the solution to the increased need for teachers involves a combination of several factors, including adding qualified teachers with a background in other professions (by making teaching a more attractive profession), increasing the volume of teacher education students and reducing the dropout rate in these programmes. Additionally, as part of a new strategy, Swedish authorities have decided that first-time teachers should receive a salary supplement of 10,000 Swedish crowns (almost 1000 euros) per month, and the school sector has been provided with more funds to create more equity in the schools. In many of today’s Swedish schools, teachers with credentials are in the minority.

Denmark

In 2012, the Danish Parliament passed a new teacher education reform that introduced the first admissions requirement for teacher training programmes. Since 2013, prospective students must have an average grade of 7 (on a scale of 1–12) to be admitted to such a programme; applicants with a lower average must undergo an admissions interview. One of the aims of the 2012 teacher education reform was to make primary school teacher education more attractive and avoid dropout. But this has not happened, according to figures from Denmark’s Ministry of Education and Research, as 15% of students in one cohort had quit by the end of the first year, 25% by the second, 29% by the third and 31.5% after the fourth. This is roughly the same pattern in average dropouts from 2006 to 2012, after which the government changed the admission requirements for education programmes. The number of primary school teachers fell by 15% from 2009 to 2017, with the number of primary school learners dropping by only 5% over the same period (Ministry of Education, 2019). Thus, a significant number of those teaching primary school lack adequate education, and this number has more than doubled since 1997. Furthermore, analysis shows that the selection of student teachers is worrying, as fewer and fewer of the country’s most talented young people become teachers, while a significant proportion of educated primary school teachers do not work in schools. Thus, primary schools face difficulties in retaining the best teachers. The proportion of teachers without education has grown from 8% in 1997 to 18% in 2017. Finally Danish student teachers have an average skill level below that of student teachers in the other Nordic countries (Søndergaard et al., 2019).

Forecasts indicate a drop in primary school teachers until the year 2022. This represents a challenge for some Danish municipalities. However, a surplus of primary school teachers is expected between 2023 and 2030 (Lange Group, 2017). These figures are based on certain assumptions about the future need for teachers, geographical patterns of access to vocational colleges’ teacher education programmes, current completion patterns and current geographical patterns of transition from educational programmes to employment in municipal primary schools.

Norway

As of 2009, prospective students aiming to enrol in a teacher education programme in Norway must have a grade average of 3.5 (with 6 the best grade) in upper secondary school. In 2016, the requirement for a grade of 4 in mathematics was added. Norwegian education authorities have argued that grade requirements are important for improving the status of the teaching profession. However, there has been no discussion of legislating the right to be taught by qualified teachers, and there is no political majority that supports that principle. Even with stricter grade requirements and a longer period of study in Norway’s new 5-year master’s degree programme, primary school teacher education still attracts many students, with the number of students increasing from 2017 to 2018. More women than men choose to pursue teacher education, but the gender distribution has become somewhat more even. Statistics Norway has made forecasts regarding the relationship between teacher demand and access to different teacher groups based on realistic expectations of the number of students, completion percentage, retirement propensity (for those over 62 years) and employment distribution. However, regional biases were not taken into account. Furthermore, supply and demand are projected separately, meaning that adjustment mechanisms between supply and demand (such as changes in wages) were not included in the calculations. Of the various teacher groups, it is only among primary school teachers that Statistics Norway estimates a deficit (i.e., a higher demand than supply). Starting in 2017, the number of elementary school teachers was expected to fall. This deficit was expected to increase to around 3000 full-time equivalents in 2020, 4000 in 2030 and 6700 in 2040. It is difficult to maintain a sufficient number of primary school teachers given the projected growth in demand (based on the number of future children of primary school age). It is worth noting that these calculations are based on assumptions about the current degree of completion, and if more people complete their education, future conditions may change. Factors that may influence the degree of completion include current educational policy, changes in salary conditions or other incentives. If a higher completion rate is achieved by addressing these factors, the deficit in primary school teachers will be reduced.

A dropout survey conducted on behalf of an expert group showed that many students quit for reasons that have nothing to do with the quality of their studies or other factors that the educational institution can address (Follow-up Group, 2015). However, that analysis also showed that the grade point average required for admission and learning pressure (as measured by the number of required assignments) can each play a role.

Just under 50% of primary school teacher education students who started in 2010 completed their education in the standard period of time; a smaller proportion of students whose progress was delayed completed the programme at a later date.

From 2008 to 2018, 8100 teachers switched from teaching in one school to another (Statistics Norway, 2019). Of the almost 20% of teacher educators who quit their jobs at schools, 5% went to jobs in other areas of the educational sector and 12% went to other industries (e.g., public administration). In the same period, 3700 former teachers, many of whom were younger women, returned to the teaching profession after working elsewhere, but the current size of this reserve of teachers is uncertain (TSN Gallup, n.d.).

Iceland

Eyjólfsson and Jónsson (2017) estimate that Iceland will see a 14% increase in the number of young people aged 6–16 from 2017 to 2034, followed by a decrease from 2034 to 2050. However, the average age of teachers in Iceland has been rising in recent years, and Sigurdardottir et al. (2018) documented a significant reduction in the number of educated teachers following the 2011 Icelandic teacher education reform. Icelandic teacher education is in a difficult situation. Many leave the profession after a short time, and there are a substantial number of people who have completed a teacher education programme but work in other professions (Eyjólfsson & Jónsson, 2017).

The problem is that admission to primary school teacher education programmes was sharply reduced after the recent teacher education reform and has not rebounded sufficiently to meet the need for this skilled labour. Simultaneously, the largest cohort of teachers in history is retiring, meaning that Iceland’s shortage of teachers will increase significantly over the next few years. In recent years, more and more student teachers have attended two-year master’s programmes after obtaining bachelor’s degrees. These programmes involve practice in schools, in which student teachers are given the full responsibility and workload of regular teachers and receive salaries (albeit at slightly lower rates than those earned by fully certified teachers).

5.2 Results and Discussion

Table 12.4 shows the correlations between Nordic student teachers’ intention to quit and a number of factors related to their perceptions of aspects of the study programme, their motivation and their time on task. Correlation measures the underlying dependence between two variables. The coefficient will always be between −1 and 1, and a correlation coefficient close to zero means that there is no linear relationship between the two variables. A positive coefficient indicates a positive empirical relationship, while a negative coefficient indicates the opposite. Correlations greater than 0.30 are in bold in the table. The individual questions underlying the aggregate variables in Table 12.4 are presented in Appendix.

Table 12.4 Relationship with Nordic student teachers’ turnover intentions: correlations
Full size table

Turnover intentions are measured by indicators of the degree to which a student teacher thinks of career opportunities other than teaching, the degree to which other professions appear more attractive than teaching and so on. The factor that is most strongly associated with ​​completion of a teacher education programme is professional identity, as measured through questions about the degree to which a student teacher looks forward to working as a teacher, how connected the student feels to the teaching profession, and whether the student feels good when thinking about 1 day becoming a teacher. Our findings are similar to those of Bergmark et al. (2018), who found a close relationship between students’ motives for choosing teacher education (the opposite of considering quitting) and what they call ‘attachment to the teacher profession’. For all Nordic countries, the size of this correlation factor was high and negative in our results. This means that students with a weak professional identity score high on the question about thoughts on quitting the teacher education programme. The measurements do not show the direction of causal processes, but it is reasonable to assume that identity is connected – positively and negatively – to thoughts of quitting the programme. One possible interpretation is that if a teacher education institution admits more student teachers with a strong potential professional identity, the likelihood of dropout is lower. If this assumption corresponds to causal processes, a system in which student teachers are admitted to a programme involves an admission test of, among other things, the strength of the applicant’s desire to become a teacher, would be well justified. Admission interviews are conducted in Finnish – and sometimes Danish – teacher education programmes, and there are plans to establish a similar process in Sweden.

The relationship between professional identity and dropout rate has also been identified in studies involving other countries. For example, Hong (2010) found that student teachers’ professional identity is related to aspects of their decision to leave education. Several different groups of student teachers and newly qualified teachers were included in Hong’s survey. Interestingly, student teachers who had completed practice teaching had less idealistic views than those who had not completed it. In other words, practical experience strengthens students’ perceptions of reality.

One implication for practice is that if the statistical relationship between professional identity and thoughts of dropping out is causal, we can say that awareness of applicants’ potential professional identity will be beneficial for preventing dropout. In the Norwegian data set, there are clear correlations between student teachers’ expectations of mastering various aspects of teaching (classroom management and eliciting student engagement) and the strength of their turnover intentions. This correlation is weaker in the data sets from the other Nordic countries. Noise and disorder in classes in which students practice teaching are not strongly associated with the idea of ​​dropping out in any of the Nordic countries. This is somewhat surprising in light of the results of a study of Finnish primary school teachers (Heikonen et al., 2017). Academic requirements and thoughts of dropping out are weakly associated in all Nordic countries, with the highest correlation in the Swedish data set.

When it comes to motivational factors, there are several subtle differences among the Nordic countries. The correlation between thoughts of dropping out and the degree of absorption in studies is negative in all Nordic countries, particularly Finland and Iceland; the stronger the thoughts of dropping out, the weaker students’ absorption in their studies. Similarly, Kim and Corcoran (2018) found a strong empirical relationship between students’ engagement in a campus environment and endurance (the opposite of thinking about dropping out). We assume that study absorption and endurance are related variables.

A relationship between altruistic motivation and thoughts of dropping out can be found in the Norwegian, Finnish and Danish data sets, and a somewhat weaker connection can be found in the Swedish and Icelandic data sets. The stronger the altruistic motivation, the weaker the tendency to think about dropping out. Whereas Bergmark et al. (2018) found that Swedish student teachers had altruistic motives for choosing teacher education, this relationship is relatively weak in our Swedish data set. Roness (2011) found that altruistic motivation is present among Norwegian student teachers, which accords with our analysis. Additionally, we found a strong negative association between altruistic motivation and thoughts of dropping out in the Finnish data set and a moderately strong negative relationship in the Norwegian and Danish data sets. Self-discipline was only very weakly associated with thoughts of dropping out. Flores and Niklasson (2014) found that family and discussions with friends were important factors in the choice to pursue teacher education and thus the teaching profession. In our study, this factor was very weak in all the Nordic countries.

6 Conclusions

An individual who chooses to pursue teacher education and the professional life of a teacher will be able to work in that role for up to 45 years before reaching retirement age. However, many who have undergone teacher training choose to leave the teaching profession in favour of alternatives. Thus, there is a reserve of educated teachers who in principle could choose to return to the profession; whether this is actually possible must be considered, as this is a ready source of teachers who already have adequate education to meet current and perhaps future demands. The challenge associated with qualifying enough teachers with adequate competence is unevenly distributed in the Nordic countries. In principle, there is an internal Nordic labour market for teachers in which current regulations provide opportunities for students to obtain teacher education from another country and achieve an adequate education for the teaching profession. In principle, it is also conceivable that if there are, for example, mathematics teachers with adequate competence working in other fields, market mechanisms can be deployed to set an attractive salary and encourage these teachers to return to the teaching profession. However, the empirical question of whether this type of market mechanism can actually be implemented remains unanswered.

The second source of teachers is teacher education. Dropout rates vary among Nordic countries’ teacher education programmes and are relatively similar within the teaching profession. With the exception of Finland and the nuances of the Danish situation, the inadequate supply of newly qualified teachers is a real challenge. Thus, it is important to explore which factors are related to student teachers’ thoughts of dropping out of teacher education or choosing another profession after they graduate. In this chapter, we have tried to explore statistical relationships between a number of factors and thoughts of dropping out. The clearest connection is the negative relationship between professional identity and the idea of ​​dropping out, which was strong in all Nordic countries. If we believe that this context reflects causal processes (something we cannot prove in this context), we must consider the applicants’ professional identity when they seek admission to teacher education.

Our empirical findings can be built upon by more fine-grained studies of various aspects of professional identity. One promising area to pursue is Hong’s (2010) six subgroups of professional identity: values, mastery expectations, commitment, emotions, knowledge and micropolitics. Those student teachers and novice teachers who had the most naive and idealistic views of vocational tasks showed the greatest emotional fatigue following their practical experience; future studies should follow up on this finding. Emotional fatigue can be a motive for leaving teacher education or the teaching profession (Skaalvik & Skaalvik, 2011). The question, then, is what constitutes a healthy amount of optimism and an unhealthy amount of idealism? Should we recruit student teachers with a realistic view of the stresses in the teacher profession rather than those with a strong, idealistic, and well-considered teacher identity? Conducting interviews as part of the admissions process may be one way to determine whether applicants to teacher education programmes have a sufficiently robust perception of the potential stresses of the profession. In addition, Bruinsma and Jansen (2010) found significant gender differences in the motivation to undergo teacher education, a topic that is worth pursuing in future research.

Weaker empirical relationships can be found for categories other than professional identity and thoughts of dropping out in Table 12.4. Student teachers’ altruistic motivation to become teachers is one factor to consider, but its relationship with thoughts of dropping out is not particularly strong in Sweden and Iceland. Thus, it must be concluded that, beyond these two previously mentioned factors, there is no clear empirical pattern in the correlations between thoughts of dropping out and other factors. This means that we need more research on student teachers’ perceptions and preferences to determine how dropout from teacher education can be prevented. Due to the peculiarity of the phenomenon, longitudinal studies should be funded.

All research has some limitations, and this study is no exception. First, correlation is not necessarily an expression of causality. Furthermore, correlation studies are ranked lower on evidence hierarchies than, for example, efficacy studies based on randomized controlled experiments. On the other hand, Lipsey and Wilson (1993) showed that there are small differences between randomized and non-randomized studies. Regardless, the degree of bias in a sample may play a significant role, as we have noted above. Finally, the samples from Sweden and Finland were each collected at a single institution, which weakens their external validity.