Finnish teacher educators’ preferences for their professional development – quantitative exploration

ABSTRACT

Education makes a difference and teacher educators are an important part of that circle. However, there is very little research done in Finland on teacher educators’ professional development. The main purpose of this study was to develop and test the psychometric properties of three scales that measure the components of teacher educators’ professional development, namely (a) developmental needs, (b) preferred ways of fulfiling those needs and (c) hindrances to fulfiling developmental needs. The differences between different occupational groups were also discussed. The survey was distributed in May 2019 to all eight Finnish universities that offer teacher education. The final sample size was 354. Using scale development techniques, we succeeded in generating items to all three inventories. We also tested the psychometric properties and the construct validity of the inventories. Our study revealed that teacher educators are not a homogenous group. Different occupational groups have different professional development interests.

KEYWORDS:

• Teacher education
• teacher educators
• professional development
• learning preferences

Introduction

Teacher educators have an enormous significance. Without them educating future teachers for our schools, we would not have a single nuclear physicist, top surgeon, or Nobel prize winner. Teachers in our schools educate each and every one, they support and encourage children, adolescents, and adults. They push them to reach for their dreams and make the most of their potential. Education makes a difference and teacher educators are an important part of that circle.

Teacher educators’ professional development has recently received increasing interest worldwide (e.g. Cochran et al. 2019; Tack and Vanderlinde 2014; Kelchtermans, Smith, and Vanderlinde 2017; Tack et al. 2018). Our article is a continuation of our previous qualitative studies and presents a quantitative study of Finnish teacher educators (Maaranen et al. 2019, 2018).

In Finland, the Ministry of Culture and Education has funded several research and development projects during 2017–2020 concerning different educational purposes in teacher education, and teacher educators’ development was one of them. Teacher education in Finland differs from many other countries, since it is only placed in research-intensive universities, and it is quite similar in all eight universities. All Finnish teachers graduate with a Master’s degree.

This research is based on a survey study by Czerniawski, MacPhail, and Guberman (2017) in which they studied the professional development needs of over 1,000 higher education-based teacher educators in Belgium, Ireland, Israel, the Netherlands, Norway, and the UK. We were also inspired by the work of Dengerink, Luneberg, and Kools (2015), who surveyed 268 Dutch teacher educators, including school-based and university-based teacher educators.

Teacher educators’ professional development

The definition of professional development has been vague. However, Smith (2010) defined professional development as follows:

Professional development … is an internal process in which professionals engage within a formal or informal framework. The process is rooted in critical self-analysis of professional practice, and steps taken to improve current practice are called professional development activities. … Professional development occurs when beliefs and assumptions about the profession change, and as a result, professional practice changes. The latter is related to professional decisions taken prior to action. (681)

When we think about becoming teachers, teacher education is responsible for the conditions of access to the profession, as well as learning what is involved in the field of education (Pereira, Lopes, and Marta 2015; Murray and Kosnik 2011). The role of teacher educators in training student teachers is at the core of this educational enterprise (Hadar and Brody 2017). But what should teacher educators know and be able to do (Goodwin et al. 2014) and what steps do they prefer to take in their professional development? According to Dengerink, Luneberg, and Kools (2015), the professional development of teacher educators denotes the ‘individual learning of those who are engaged in activities, which require specialised attitudes, knowledge and skills, rooted in a formal or informal specialist body of knowledge or knowledge base’ (80). Studies about teacher educators’ professional development have gained increasing attention and various studies have been conducted around it (see, e.g. Lunenberg, Dengerink, and Korthagen 2014). Dengerink, Luneberg, and Kools (2015), for example, investigated 377 school-based and university-based teacher educators’ preferences concerning their professional learning. According to their results, teacher educators want to develop their coaching and teaching as well as their knowledge of the theoretical foundation of their work, whereas the improvement of pedagogical content knowledge, skills in curriculum development, assessment, and organisational matters were less emphasised as an improvement aim (Dengerink, Luneberg, and Kools 2015). Murray and Male (2005) investigated 28 teacher educators in their first three years of working in initial teacher education in England. They found that teacher educators met challenges in two main areas: becoming research active and developing a pedagogy for higher education. Based on the results, new teacher educators need to develop pedagogical knowledge and understanding how research and teaching might be interrelated. Furthermore, Boei et al. (2015) explored 13 teacher educators’ professional development in a one-year supportive programme in the Netherlands. They found that professional development was mostly connected to collaboration with peers, a broadened theoretical knowledge base, the capability to connect theory with practice, and a deepening awareness of professional identity and their role as a teacher educator (Boei et al. 2015). Furthermore, studies concerning teacher educators’ professional development have shown a struggle with the competing demands of the university and school contexts (Mayer et al. 2011) and various concerns by experienced teacher educators during their career (Van der Klink et al. 2017) as well as by mentor teachers (Andreasen, Bjørndal, and Kovač 2019).

Based on studies concerning teacher educators’ professional development, Kleinsasser (2017) states that professional development and the work of teacher educators are integrally related. However, the nature of teacher educators’ work is complex because teacher education is multifaceted in nature. As Taylor (1983, 41, in Murray and Male 2005) has put it: ‘In the one direction it faces classroom and school, with their demands of relevance, practicality, competence, techniques. In the other it faces the university and the world of research, with their stress on scholarship, theoretical fruitfulness and disciplinary rigour.’ Thus, the professional development of teacher educators denotes various activities and tasks in different contexts and settings.

This study deals with a similar context, as described above. Finnish teacher education is research-oriented and carrying out research is a fundamental part of a teacher educator’s work. In Finland, teacher education is strongly rooted in research-intensive universities, and contrary to many other countries (see, e.g. White 2019; Vanassche, Kidd, and Murray 2019), there seems to be no need for policy discussions regarding its place. School-based teacher education plays a minor role in Finland.

At the same time, being educationalists, teacher educators’ interests are associated with pedagogical matters. In the Finnish context, university teacher training schools create their own special aspect, where concentrating on questions of supervising student teachers is the core, as we will describe in the following section.

Finnish teacher education and educators

Teacher education is a popular field of study in Finland. Institutes of higher education are therefore in a good position to select the best-suited and the most highly motivated applicants, as well as the most highly qualified teacher educators. On the university level, most teacher educators carry out research that could be considered part of their personal professional development. Finnish senior lecturers are generally required to hold a doctoral or other postgraduate degree and they must complete pedagogical studies. Most of them also have a teaching qualification (for early-childhood education, classroom teaching, or subject teaching), as well as some working experience at kindergarten or school level.

Finnish teachers are highly trained academic professionals, teacher education being research-based (cf. Maaranen et al. 2019), and the link between teaching and research is emphasised. The objective is to produce teachers with a research orientation in their work who are capable of independent problem solving and have the capacity to apply the most recent research to the fields of education and the subjects taught. Using these skills, they are able independently to develop themselves as teachers, and their work community, in cooperation with other staff in the educational institution (Ministry of Education and Culture 2016).

There is no specific requirement for teacher educators to be competent in the use of ICT hence skills and knowledge in this field vary a lot. According to the Ministry of Education and Culture (2016), focal points in developing the content of teacher education include the teaching and guidance of students requiring a higher level of support, the development of learning environments, the utilisation of ICT, issues related to a multicultural society, cooperation between school and home, and cooperation with the world of work. The University of Helsinkioffers optional courses in technology use in which the faculty can participate, and many teacher educators have taken advantage of this. There are also a number of development and research projects concerning the educational use of technology being carried out in the teacher education unit. The research we report here belongs to one of these projects. Professional development is the responsibility of each individual teacher educator.

To qualify as a teacher in Finland, student teachers practise in a university teacher training school during their teacher education. University teacher training schools are administratively part of the Faculties of education and have a high degree of autonomy. In addition to offering practicum placements for student teachers, training schools participate in the development of teacher education and co-operate with faculty departments, other university departments as well as national and foreign educational institutions. In other words, the teachers at the university’s teacher training schools are part of university personnel who actively participate and carry out a wide range of experiments, research and in-service training (see eNorssi.fi/ftts). In this sense, the Finnish university teacher training school system differs from what is generally understood as school-based teacher education. Teachers at training schools have at least MA degrees, but also many of them have a doctoral degree (Niemi and Jakku-Sihvonen 2011).

Aims of the study

There is very little research done in Finland on teacher educators in general or their professional development. The field seems to be quite novel, though it is increasingly growing. Our first studies on Finnish teacher educators (Maaranen et al. 2018, 2019) showed that further research is needed, especially based on larger scale data. Thus, our study has four main questions:

1. Is it possible to develop a questionnaire that measures three closely related multidimensional components of teacher educators’ professional development, namely (a) developmental needs, (b) preferred ways of fulfiling those needs and (c) hindrances to fulfiling developmental needs?

2. What preferences do Finnish teacher educators have concerning their professional development?

3. How are they able to realise these professional development opportunities and what kind of hindrances do they identify?

4. What kind of differences, if any, can be found between teachers at the university teacher training school and teacher educators from the TE departments?

Method

Subjects

In May 2019 the survey was distributed to all eight Finnish universities that offer teacher education. The estimated total population of teacher educators was about 1,400. We received 354 responses to the survey, which is a response rate of about 25%. The responses from the University of Helsinki were emphasised in the data, as the response rate in University of Helsinki was 39.7%. We obtained 126 responses from teacher training schools that operate under the universities and 216 responses from the universities’ departments of teacher education. The majority of participants were women, only 83 men taking part in the research.

The core duties of Finnish universities are research, teaching, and societal impact. However, depending on the job requirements at the university, the share of the three duties varies. For instance, the job description of professors emphasises research whereas the job description of university lectures and supervising teachers in teacher training schools emphasises teaching. Thus, it does not make sense to examine occupational groups as one group. Based on the job duties teachers had, we composed five occupational groups, which are presented in Table 1. We placed PhD students and researchers in the same sub-group because they had very similar response profiles.

Table 1. Definition summary of the occupational groups of the data

Occupational group	Definition
Supervising teachers, n = 126	Supervising teachers organise the teaching practice included in the University’s teacher education in a teacher training schools. They supervise and develop this teaching practice.
PhD students and researchers, n = 33	Both PhD students’ and researchers’ main job is to do research. For PhD students this means the preparation of a doctoral thesis.
Lectures (MA), n = 64	The position is teaching-focused. This teaching position entails teaching undergraduate and graduate courses. Essential duties include the development of course contents and teaching methods to meet the needs of the various fields studied at Finnish universities.
Senior lecturers (PhD), n = 86	The duties of the senior lecturer include providing teaching in accordance with the degree requirements, supervising and examining theses and dissertations, and conducting research in the field. According to the regulations of Finnish universities, holders of university lectureships are required to possess an appropriate doctoral degree, to have the ability to provide high quality research-based teaching and to supervise theses and dissertations.
Professors and Assistant Professors, n = 30	A tenure track appointee will be employed as an assistant professor for a duration of three to five years, depending on his or her background and experience. If the appointee fulfils the predetermined criteria and successfully performs the duties of the position, he or she may be appointed professor or the employment may be continued with a second contract for three to five years, after which the appointee may receive a permanent position as a professor. Professors should take advantage of national and international research funding opportunities, participate in Department’s teaching activities and doctoral education, and actively participate in the interaction between science and society at large.

Measures

Where applicable, the scale development was based on the suggestions of Boateng et al. (2018) as well as Netemeyer, Bearden, and Sharma (2003). The questionnaire contained five parts: professional development and satisfaction with the possibility of developing professionally over the past two years (2 items); degree of interest in future professional development (1 item); preferences for professional development (PPD-scale, 24 items); what professional learning opportunities would best meet the professional learning preferences (PLO-scale, 17 items); things that prevent professional development (PD-prevent scale, 11 items). Items were assigned on a six-point rating scale ranging from ‘not at all’ to ‘very much’ or ‘very well’, depending upon the item context. PPD, PLO- and PD-prevent scales also had an open-ended question, ‘something else’. This question was also estimated on a scale of 1–6.

The item generation phase of the scale development was mainly deductive (Boateng et al. 2018). Where appropriate, the items of the generated scales were based on previous studies (Czerniawski, MacPhail, and Guberman 2017; Dengerink, Luneberg, and Kools 2015). However, an inductive method was also used in the item generation phase. Qualitative data obtained through previous studies (Czerniawski, MacPhail, and Guberman 2017; Dengerink, Luneberg, and Kools 2015; Maaranen et al. 2018, 2019) was utilised in item development. The research team worked as expert judges to evaluate the content validity of the generated items. In several meetings, the content of the items were compared to the target concepts and modifications were made according to the discussions. Together the research team has over 100 years of experience in teacher education and in positions related to it.

For the PPD-scale, suitable items were translated from English into Finnish and they were also modified to suit the Finnish context. Some items that were context specific to Finland were also added to the questionnaire. The Finnish teacher educators’ job description consists of three parts: teaching, research, and societal impact. New items were generated so that the PPD-scale measures all three components of the teacher educators’ job description. The generation of items for the PLO- and PD-prevent scales was based on the study of Czerniawski, MacPhail, and Guberman (2017). In this study, learning opportunities and things that prevent professional development were asked using open-ended items. Based on these responses and our experience as teacher educators we generated closed items to PLO- and PD-prevent scales. Participants also provided open-ended responses to the survey question ‘What are your two most important professional learning needs?’ The questionnaire was implemented as an e-form that was sent to all eight Finnish universities that offer teacher education in Finland.

Data analyses

Using IBM SPSS Statistics 24 we first tested the skewness and kurtosis values of the variables. All values were between +2/-2, thus indicating acceptable normal univariate distribution (George and Mallery 2010). Means and standard deviations were used to summarise the items of the scales. Exploratory factor analysis (EFA) was used to test the psychometric properties and the construct validity of the inventories (Boateng et al. 2018; Gorsuch 1997; Haig 2012; Netemeyer, Bearden, and Sharma 2003). The convergent and discriminant validity of the scales was investigated by using Person correlation coefficients. As a criterion to retain or delete items from the scale, we used the suggestions of Tabachnick and Fidell (2014) and Worthington and Whittaker (2006). However, to avoid capitalising on chance we did not use these criteria categorically. As a rule of thumb, items with a factor loading of less than .32 or cross-loadings less than .15 difference from an item’s highest factor loading were deleted.

The independent t-test was used to investigate the differences between the two respondent groups mean values. One-way analysis of variance (ANOVA) was used to determine whether there were any significant differences between the variances of the different participant groups. The assumptions of ANOVA (normality and equality of variance) were checked before conducting the analyses. In the situations where the equality of variances assumption was violated Welch’s-test was used instead of normal ANOVA. Canonical correlation analysis (CCA) was used to study how two sets of variables (subscales) relate to each other (Tabachnick and Fidell 2014, 617).

Results

Teacher educators’ satisfaction with past opportunities for professional development

Our first question asked respondents about satisfaction with their professional development over the previous two years. Teachers in university teacher training schools (M = 4.43, SD = 1.16) were more satisfied with their professional development than the respondents in the departments (M = 3.96, SD = 1.31). The difference was statistically significant t(340) = 3.405, p = .001. Cohen’s d was 0.388, which according to Cumming and Calin-Jageman (2017) can be considered in education a medium effect. However, there were also statistically significant differences among the different sub-groups of the departments F(3.212) = 4.867, p = .003, η² = .064. Lecturers (M = 4.41, SD = 1.09) were the sub-group that was most satisfied with their professional development over the previous two years, and senior lecturers (M = 3.61, SD = 1.35) were the least satisfied sub-group. The results were similar when we asked respondents about opportunities for professional development in the previous two years. Teachers in training schools (M = 4.14, SD = 1.27) were more satisfied with their opportunities for professional development than respondents in the university departments (M = 3.50, SD = 1.61), t(310) = 3.825, p < .001, d = .442. Again, senior lecturers (M = 3.01, SD = 1.60) were the least satisfied sub-group in the whole sample.

The mean value of interest in future professional development was high in both main respondent groups, M = 5.38, SD = 0.99 (teachers in training schools) and M = 5.28, SD = 1.10 (departments), t(340) = 0.826, p = .410. In the university departments PhD students and researchers (M = 5.76, SD = 0.61) were the sub-group most interested in future professional development. As expected, experience correlated negatively to interest in future professional development in both main respondent groups, r = −.323, p < .001 (training schools) and r = −.387, p < .001 (departments). Thus, experience decreases the interest in future professional development.

What preferences do Finnish teacher educators have concerning their professional development?

Czerniawski, MacPhail, and Guberman (2017) subjected responses that concerned teacher educators’ professional development interest to principal component analysis. The analysis yielded two main factors. The first factor was named ‘Academic Interest’ and the second factor ‘Educational Interest’. To validate further the structure of the scale we followed the suggestions of Tabachnick and Fidell (2014) and continued the analyses with factor analysis. We used the maximum likelihood method with Promax oblige rotation. The two-factor solution was very similar to the solution presented by Czerniawski, MacPhail, and Guberman (2017). The correlation between marker variables on the corresponding factor was 0.91. However, in our data, the two factor solution showed signs of underfactoring, namely the eigenvalues of the third and fourth factors were over 1 and the goodness of fit test was statistically highly significant (χ²(207) = 602, p < .001). The residual correlation matrix of the two-factor solution also contained several meaningful substantial correlations. Thus, we also extracted three- and four-factor solutions. Of these two solutions, the three-factor solution was interpretably more meaningful (cf. Gorsuch 1983) and it corresponded to the university’s core duties, which are research, teaching, and societal impact. In the three-factor solution, the Academic interest factor (cf. Czerniawski, MacPhail, and Guberman 2017) split into two components (Factors 1 and 3), which we named ‘Knowledge Sharing Interest’ and ‘Research Skills Interest’. As expected these two factors correlated moderately (r = .635). The item ‘Students scientific mentoring’ was named according to its secondary loading on Factor 3. This item had primary loading (λ = .467) on Factor 1. The item ‘Academic administration competence’ was named with caution on Factor 1 (cf. Czerniawski, MacPhail, and Guberman 2017). Following Czerniawski, MacPhail, and Guberman (2017), Factor 2 was named ‘Educational Interest’. Together the three factors accounted for 49.28% of the variance among 24 items. The factor loadings of each item are given in brackets in Table 2. The table also contains the item means and standard deviations of each sub-sample.

Table 2. The factor loadings, means (neutral point = 3) and standard deviations for statements concerning teacher educators’ interest in professional learning activities

		Supervising teachers n = 126		PhD students and researchers n = 33		Lectures n = 64		Senior lecturers n = 86		Professors and Assistant Professors n = 30
	Variable	M	SD	M	SD	M	SD	M	SD	M	SD
F1: Knowledge sharing	Presentations and participation at conferences (.895)	2.61	1.88	4.50	1.31	3.19	1.77	3.49	1.70	4.07	1.68
	Conference presentations (.822)	2.23	1.66	3.97	1.49	3.19	1.65	3.33	1.66	3.67	1.91
	Reviewing journal articles, etc. (.734)	1.84	1.71	4.29	1.49	2.73	1.78	3.30	1.68	4.07	1.33
	Media impact (.582)	2.59	1.60	4.06	1.07	3.23	1.31	3.23	1.63	3.80	1.47
	Internationalisation (.530)	3.75	1.91	4.56	1.40	3.68	1.82	3.88	1.86	4.03	1.95
	Academic administration competence (.506)	2.48	1.93	2.79	1.32	2.50	1.45	2.51	1.55	2.83	1.86
	Popularised writing (.393)	3.23	1.83	4.21	1.32	3.97	1.32	3.33	1.51	3.03	1.50
F2: Educational	Mentoring practical teacher training (.672)	5.38	0.90	3.71	1.96	4.36	1.66	3.77	1.62	1.93	1.78
	Teaching skills (.627)	5.00	1.26	5.00	1.07	5.17	1.09	4.24	1.32	3.17	1.63
	Current developments in teacher education (.548)	4.83	1.12	4.88	1.34	4.95	1.15	4.28	1.48	4.77	1.46
	Assessment (.504)	4.38	1.31	4.06	1.35	4.67	1.40	4.47	1.40	3.63	1.52
	Researching one’s own practice (.500)	4.04	1.57	4.21	1.24	4.54	1.20	3.75	1.62	3.57	1.52
	Subject knowledge enhancement (.483)	4.44	1.46	4.44	1.46	4.66	1.31	4.13	1.45	3.70	1.58
	Curriculum development (.467)	3.91	1.60	4.36	1.25	4.75	1.07	4.03	1.39	3.67	1.71
	Mentoring student teachers (.434)	4.10	1.47	3.94	1.48	3.55	1.47	3.17	1.69	3.69	1.58
	Learning material generation (.410)	3.84	1.68	3.74	1.38	4.63	1.10	3.78	1.58	3.67	1.42
	Participating in union activities (.400)	2.21	1.80	2.27	1.66	2.27	1.50	2.13	1.45	1.71	1.63
	Integrating technology into teaching (.383)	4.40	1.39	3.79	1.67	4.30	1.43	4.10	1.41	3.87	1.78
F3: Research skills	Qualitative research skills (.813)	3.12	1.75	5.15	0.89	4.47	1.50	4.34	1.55	4.10	1.58
	Quantitative research skills (.595)	2.84	1.70	4.86	1.35	3.97	1.70	3.91	1.75	4.23	1.57
	Action research (.564)	3.50	1.71	3.97	1.70	4.45	1.41	3.55	1.77	3.67	1.95
	Empirical research (.539)	3.06	1.68	4.91	0.93	4.17	1.54	4.29	1.54	5.10	0.88
	Academic writing (.488)	3.01	1.67	5.18	0.94	4.38	1.32	4.34	1.48	4.87	1.25
	Students’ scientific mentoring (.315)	3.13	1.73	5.29	0.91	3.84	1.71	4.65	1.38	4.83	1.29

As seen from Table 2, the teachers at university teacher training schools were mostly interested in developing themselves professionally in activities which loaded on the factor ‘Educational Interest’, namely Instruction skills; Supervision of teaching practice; Developing teacher education; Enhancing content knowledge; Enhancing assessment skills; and Integrating technology in teaching. In the departments, the sub-sample group lecturers’ response profile was very similar to teachers at Teacher Training Schools. They also emphasised activities that loaded on the ‘Educational Interest’ factor.

PhD students’ and researchers’ response profile was very similar to that of senior lecturers. Both groups were mostly interested in developing themselves professionally in Developing teacher education; Supervising students’ scientific theses; Skills related to research methods (qualitative and quantitative); Instruction skills; Empirical research skills; and Scientific writing skills. Professors’ and assistant professors’ interest was directed towards developing teacher education and research skills.

Regarding professional development, teachers in university teacher training schools were least interested in Preparing conference presentations; Reviewing scientific articles; Participation and presentation in conferences; Supervising students’ scientific theses; Media presence; and Scientific writing.

Regarding professional development, the respondents of teacher education departments were least interested in Administrative skills; Participation in professional organisations; Writing popular publications; Media presence; and Reviewing scientific articles. In addition to the items of the questionnaire, the respondents were able to add things or express themselves more precisely in an open-ended question. Some concrete ideas and interests showed up. These included teamwork and personal relationships with colleagues as well as collegial support. A part of the interest concerned well-being and controlling stress.

Teacher educators’ preferred ways of fulfiling developmental needs

The 17 multiple choice items on the PLO-scale were based on our experience as teacher educators and on the previous studies (Czerniawski, MacPhail, and Guberman 2017; Maaranen et al. 2018, 2019), where only qualitative data were gathered with open-ended questions. Thus, our first aim was to test how the invented 17 Likert-type items work to measure preferred ways of fulfiling developmental needs. On the other hand, we also wanted to know what construct would explain the intercorrelations among the 17 variables. Both the KMO-index and Bartlett’s test of sphericity supported the use of factor analysis. The value of KMO-index was 0.842 and the Bartlett’s test of sphericity was highly significant (χ²(136) = 2863.70, p < .001), thus supporting the use of factor analysis. Four factors with eigenvalues greater than one were found. A series of factor analyses were conducted which indicated that three factors gave the most interpretable solution. These three factors explained 58% of the sample variance. The first factor was named ‘Learning by Doing Research’ (eigenvalue 5.059). In the interpretation and naming process of Factors 2 and 3 we utilised some previous studies. The second factor was named according to Lunenberg, Dengerink, and Korthagen (2014) ‘Learning from and with Colleagues’ (eigenvalue 3.619). Only two items loaded on the third factor. However, this factor had a clear interpretation (c.f., Langelotz 2017; Smith 2010). We named it ‘Collegial Interaction’ (eigenvalue 1.186), which refers to collegial conversations and reflective practices (cf. Selkrig and Keamy 2015), and also to collegial partnerships (cf., Smith 2016). Factors 2 and 3 correlated moderately (r = .511). The factor loadings of each items are given in brackets in Table 3. The table also contains the item means and standard deviations of each sub-sample.

Table 3. The factor loadings, means (neutral point = 3) and standard deviations for statements concerning how professional development needs can best be realised

		Supervising teachers n = 126		PhD students and researchers n = 33		Lectures n = 64		Senior lecturers n = 86		Professors and Assistant Professors n = 30
	Variable	M	SD	M	SD	M	SD	M	SD	M	SD
F1	Research, together with other researchers (.921)	3.04	1.80	5.73	0.45	4.61	1.58	4.88	1.51	5.63	0.49
	Work in a research team (.868)	2.84	1.68	5.59	0.70	4.21	1.71	4.33	1.88	5.50	0.78
	Writing publications (.850)	2.57	1.75	4.88	1.12	4.00	1.56	4.26	1.70	5.17	0.91
	Doing own research (.833)	3.06	1.82	4.88	0.98	4.20	1.62	4.36	1.55	4.90	1.03
	Participating seminars and congresses (.508)	3.82	1.59	4.59	1.10	3.92	1.69	4.20	1.43	4.93	0.92
	Research, together with supervising teachers (.444)	3.59	1.72	3.41	1.74	3.30	1.91	2.98	1.85	3.76	1.83
	Reading literature of the field (.347)	4.55	1.33	5.03	0.97	5.03	1.02	4.79	1.43	5.13	1.01
F2	Visiting other depts. of teacher education (.758)	4.21	1.64	3.91	1.38	4.08	1.48	3.20	1.70	2.48	1.74
	Observing the teaching of colleagues (.598)	4.13	1.50	4.38	1.26	4.05	1.55	3.20	1.73	2.90	1.56
	Staff training (.498)	3.92	1.53	3.24	1.35	3.64	1.54	2.94	1.60	2.60	1.79
	Studying together with colleagues (.482)	4.16	1.38	4.68	1.32	4.45	1.42	3.69	1.68	4.03	1.86
	Participating in innovative educational projects (.475)	3.74	1.71	4.82	1.57	4.31	1.52	3.79	1.82	4.37	1.67
	Volunteer education and further training (.473)	4.52	1.45	3.97	1.34	4.34	1.62	3.77	1.64	2.83	1.87
	Participating in exchange programmes (.465)	3.11	1.73	3.65	1.52	3.19	1.77	3.41	1.64	3.53	1.80
	Getting better acquainted with school reality (.318)	2.34	1.89	4.15	1.79	4.02	1.67	3.52	1.85	3.07	1.66
F3	Informal discussions with colleagues (.875)	4.84	1.23	4.73	1.24	5.11	1.06	4.34	1.46	4.00	1.41
	Consulting colleagues (.740)	4.64	1.19	4.82	1.09	4.72	1.24	4.07	1.56	3.77	1.57

F1 = Learning by Doing Research, F2 = Learning from and with Colleagues and F3 = Collegial Interaction.

Teachers at the university teacher training schools considered the ways they could realise their professional development needs in the best possible ways. They were Unofficial discussions with colleagues; Consulting colleagues; Voluntary training and qualification; Reading relevant professional literature; Observing colleagues’ teaching; and Visiting other teacher education departments.

The faculty of teacher education departments considered the ways they could realise their professional development needs in the best possible way. They were Reading relevant professional literature; Doing research with other researchers; Working in a research group; Unofficial discussions with colleagues; Participating in innovative educational projects; and Conducting one’s own research.

The least mentioned ways of realising professional development for teachers at the Teacher Training Schools were Research and Writing scientific publications, whereas for the faculty of the departments of teacher education it was Personnel courses; By getting to know school reality better; and Conducting research together with teachers at Teacher Training Schools.

Hindrances to fulfiling developmental needs

The value of KMO-index and Bartlett’s test of sphericity supported the use of factor analysis also in the analysis of the PD-prevent scale. The value of the KMO-index was 0.736 and Bartlett’s test of sphericity was statistically highly significant, χ²(55) = 892.65, p < .001. Three factors with eigenvalues greater than one were found. A series of factor analyses were conducted, which indicated that two factors gave the most interpretable solution. The first factor was labelled ‘Lack of Time’ (eigenvalue 3.153) and the second factor ‘Lack of Competence’ (eigenvalue 1.898). Factors 1 and 2 correlated moderately (r = .511). The factor loadings of each items are given in brackets in Table 4. The table also contains the item means and standard deviations of each sub-sample.

Table 4. The factor loadings, means (neutral point = 3) and standard deviations for statements concerning hindrances to professional development

		Supervising teachers n = 126		PhD students and researchers n = 33		Lectures n = 64		Senior lecturers n = 86		Professors and Assistant Professors n = 30
	Variable	M	SD	M	SD	M	SD	M	SD	M	SD
F1	Too much work (.911)	4.67	1.66	3.91	1.86	4.72	1.49	4.85	1.31	4.67	1.32
	Lack of time (.795)	5.04	1.41	4.23	1.72	4.91	1.40	4.91	1.40	4.71	1.72
	Job description too fragmented (.451	3.75	1.90	3.85	1.81	3.94	1.95	4.56	1.49	4.03	1.92
	Pressures to produce results (.410)	2.16	1.81	3.12	2.07	3.09	1.97	3.33	1.70	3.43	2.12
	Tiresomeness of administrative jobs (.348)	2.03	1.86	1.24	1.16	2.30	1.74	2.77	1.58	3.47	2.05
F2	Lack of academic writing skills (.701)	1.49	1.55	1.47	1.52	2.20	1.74	1.16	1.15	0.97	0.93
	Lack of research skills (.687)	2.47	1.68	2.47	1.56	2.89	1.68	2.01	1.46	1.37	1.38
	Lack of networks (.488)	1.81	1.51	2.53	1.89	2.09	1.57	1.88	1.60	1.33	1.42
	Professional uncertainty (.487)	1.04	1.34	2.68	1.92	1.91	1.81	1.61	1.54	1.23	1.65
	Lack of language proficiency (.442)	1.36	1.52	1.47	1.80	1.95	1.85	1.99	1.83	0.87	1.33
	Lack of technological skills (.394)	1.58	1.51	1.58	1.70	1.66	1.61	2.02	1.51	1.80	1.47

F1 = Lack of Time and F2 = Lack of Competence.

The hindrances both main groups considered to be the greatest for professional development were the same: lack of time and too much work. Senior lectures also emphasised that fragmented job descriptions were hindrances to fulfiling the developmental needs they had. In none of the examined sub-samples did a lack of competence appear to be a hindrance to fulfiling developmental needs.

Relationships between scales

Canonical correlation analysis (CCA) was used to analyse the relationships between PPD-, PLO- and PD-prevent scales. For this purpose, sum variables were formed. The sum variables were created according to the results of factor analyses. The internal consistency/homogeneity of formed sub-scales were investigated by Cronbach’s Alpha. The resulting Alpha coefficients are presented in Table 5. The internal consistencies of all subscales were in the optimal range (Boyle 1991).

Table 5. Internal consistencies of the subscales

Subscale	No. items	Cronbach’s Alpha
Knowledge Sharing Interest	7	.86
Educational Interest	11	.79
Research Skills Interest	6	.85
Learning by Doing Research	7	.87
Learning from and with Colleagues	8	.78
Collegial Interaction	2	.88
Lack of Time	5	.74
Lack of Competence	6	.70

After the sum variable construction, the assumptions of canonical correlation were tested (see Tabachnick and Fidell 2014). There were no problems concerning the normality, linearity or homoscedasticity of the subscales. However, as Table 6 shows, there was moderate multicollinearity between ‘Knowledge sharing’ and ‘Research skills’ sum variables.

Table 6. Intercorrelation matrix with descriptive statistics for sub-scales

	KS	EI	RSI	LDR	LfwC	CL	LT	LC
KSI	–
EI	.377**	–
RSI	.704**	.360**	–
LDR	.676**	.220**	.780**	–
LfwC	.395**	.538**	.280**	.297**	–
CI	.122*	.423**	.123*	.116*	.572**	–
LT	.082	.031	.135*	.176**	.026	.086	–
LC	−.008	.120*	.087	.030	.122*	.084	.283**	–
M	3.195	4.059	3.910	4.060	3.712	4.544	3.741	1.772
SD	1.251	0.844	1.300	1.265	1.050	1.276	1.210	1.015

*p <.05. **p <.01. All other items were not statistically significant. KSI = Knowledge Sharing Interest, EI = Educational Interest, RSI = Research Skills Interest, LDR = Learning by Doing Research, LfwC = Learning from and with Colleagues, CI = Collegial Interaction, LT = Lack of Time and LC = Lack of Competence.

The first CCA was performed between the set of PPD-subscales and the set of PLO-subscales. This analysis yielded three canonical functions within squared canonical correlations (R_c²) of .641, .285 and .039 for each successive function. The full model was statistically significant and had what may be considered a large effect size, Wilks’s Λ = .246, F(9, 727.84) = 62.93, p < .001, R_c² = 75.4%. All three calculated functions were statistically significant (p < .01). However, following the suggestions of Sherry and Henson (2005), only the first two functions were considered noteworthy. The third function explained only 3.9% of the remaining variance.

Table 7 presents the standardised canonical function coefficients and structure coefficients for Functions 1 and 2. According to the Function 1 coefficients, the relevant PPD-subscales are primarily ‘Knowledge Sharing Interest’ and ‘Research Skills Interest’. This conclusion was supported by squared structure coefficients (r_s²). Among the PLO-subscales the relevant subscale for Function 1 was ‘Learning by Doing Research’. Because all relevant coefficients of Function 1 are negative, this indicates that knowledge sharing interest and research skills interest are associated with doing research.

Table 7. Canonical solution for preferences for professional development

Function 1					Function 2
Subscale	Coef	rs	rs^2 (%)	Coef	rs	rs^2 (%)	h^2(%^)
KSI	−.345	−.857	73.4	.092	.206	4.3	77.6
EI	.073	−.345	11.9	1.055	.928	86.1	98.0
RSI	−.750	−.973	94.7	−.454	−.005	0.0	94.7
LDR	−.966	−.995	99.0	−.357	−.056	0.3	99.3
LfwC	−.125	−.387	15.0	.798	.888	78.8	93.8
CI	.075	−.129	1.7	.357	.760	57.7	59.4

KSI = Knowledge Sharing Interest, EI = Educational Interest, RSI = Research Skills Interest, LDR = Learning by Doing Research, LfwC = Learning from and with Colleagues, CI = Collegial Interaction, Coef = standardised canonical function coefficient.

The coefficients of Function 2 indicate that the only relevant PPD-subscale for this function was ‘Educational Interest’. The coefficient of this subscale was over 1.0, but this result was due to the multicollinearity of the variables. Among the PLO-subscales the relevant subscale for Function 2 were ‘Learning from and with Colleagues’ and ‘Collegial Interaction’. It was also notable that the subscale ‘Doing research’ was inversely related on this function.

We also investigated the relationships between PPD- and PD-prevent subscales. However, neither of the calculated two functions were statistically significant.

Discussion

Our investigation had both methodological and substantial aims. Our first aim was to develop a questionnaire that measures three closely related multidimensional components of teacher educators’ professional development, namely developmental needs (PPD-scale), preferred ways of fulfiling those needs (PLO-scale) and hindrances to fulfiling developmental needs (PD-prevent scale). Using a scale development technique suggested by Boateng et al. (2018) and Netemeyer, Bearden, and Sharma (2003) we succeeded in generating items for these three inventories. Using exploratory factor analysis, we tested the psychometric properties and construct validity of the inventories. As expected, all inventories were multidimensional.

Analyses indicated that three distinct factors underlay respondents’ responses to the 24 items of the PPD-scale. We labelled these factors ‘Educational Interest’ ‘Knowledge Sharing Interest’ and ‘Research Skills Interest’. The ‘Educational Interest’ factor is similar to Czerniawski, MacPhail, and Guberman (2017) corresponding factor. However, in the present study, Czerniawski et al.’ s ‘Academic Interest’ factor split into two moderately correlating factors, namely ‘Knowledge Sharing Interest’ and ‘Research Skills Interest’. This three-factor solution made conceptual (theoretical) sense because the university’s core duties are primarily research, teaching and societal impact. Depending on the job duties at the university, the share of these three duties varies. For instance, the job description of professors emphasises research whereas the job description of supervising teachers and lectures emphasises teaching. It is worth noting that sharing may mean different contents. For faculty members it points, among other things, to societal impact and the sharing of research results. Regarding university teacher training schools, sharing obviously deals with such kinds of things as collegial conversations and co-operation. However, it is good to keep in mind that Finnish university teacher training schools are also research-oriented, as we have described earlier in this article. It is reasonable that the instrument that measures teacher educators’ professional development also takes into account this fact in the university community.

The factor loadings for the individual items showed that the items of the PPD-scale had loadings that were between .393 and .895, thus indicating that they are appropriate indicators of their respective factors (see, e.g. Tabachnick and Fidell 2014). These findings imply that this instrument has adequate construct validity of its factor structure and add strong support to the assumption that teacher educators’ professional development needs to be analysed from a multidimensional perspective. However, the PPD-scale also contained some problematic items. First, item ‘Academic administration competence’ loaded highly (λ = .506) on the Knowledge Sharing factor but has a bad conceptual fit to this context. However, this item also has a bad conceptual fit to all other emerged factors. The item level analyses of the present study also revealed that teacher educators have very little interest in academic administration competence development. The mean values of this item varied from subgroup to subgroup from 2.48 to 2.83 (neutral point = 3). Thus, we suggest that this item can be removed from the scale. It is also notable that the inventory developed by Czerniawski, MacPhail, and Guberman (2017) did not contain this kind of item. Second, the items ‘Popularized writing’, ‘Researching one’s own practice’ and ‘Academic writing’ had cross-loadings less than .15 difference from an item’s highest factor loading. Although we recognise that the cross-loadings of these items may increase the correlation among factors, we did not delete these items from the scale. Worthington and Whittaker (2006, 823) have warned researchers to use cross-loadings as a criterion for item deletion.

The responses to our PPD-scale item ‘Something else’ gave further support to the content validity of our scale (cf. Boateng et al. 2018, 6–7). We analysed all responses to this item and this analysis revealed that we had captured what is relevant to teacher educators’ professional development experience. The only item we suggest might be added to the scale is ‘Applying for research funding’. This nowadays is an important part of the job description of professors and senior lecturers in Finland.

The PLO- and PD-prevent scales also had a clear and conceptually fitting structure in our study. The PLO-scale that measured the best ways of realising professional development needs had a clear three-factor structure. Common to all three factors was learning collaboratively in a team, this feature being especially emphasised in factors two and three. The first factor was named ‘Learning by Doing Research’. The items, which loaded highly on this factor, reflected different research-related activities. That is, doing one’s own research as well as working in a research team (cf. Czerniawski, MacPhail, and Guberman 2017). Items which referred to writing publications and participating in seminars and congresses also had a high loading on this factor. The second and third factors on the PLO-scale correlated moderately. Both these factors include common co-operative learning and working collaboratively. We named the second factor in accordance with Lunenberg, Dengerink, and Korthagen (2014) ‘Learning from and with Colleagues’. Only two items loaded on the third factor. Regardless of this, the factor has both conceptual and empirical (factor loading) grounds. Several researchers (e.g., Langelotz 2017; Smith 2010), have emphasised that collegial learning and a collegial environment enhance teachers’ professional development. Thus, we named our third factor ‘Collegial Interaction’. The PD-prevent scale likewise had a clear and conceptually clear factor structure. The two factors were named ‘Lack of time’ and ‘Lack of competence’. None of the items had salient cross-loadings in the PD-prevent scale. In the PLO-scale items ‘Participating in innovative educational projects’ and ‘Reading literature of the field’ had cross-loadings less than .15 difference from an item’s highest factor loading. Thus, we suggest that the expression ‘innovative educational projects’ should be clarified in the next version of the scale.

The alpha coefficients of sum variables that were formed according the factor analytical results, were at an acceptable level. Thus, using CCA we also investigated the relationships between different scales. The finding that ‘Knowledge Sharing Interest’ and ‘Research Skills Interest’ are linked to ‘Doing Research’ is almost self-evident. However, the other finding from CCA was more interesting. According to our results, ‘Educational Interest’ is positively linked to ‘Learning from and with Colleagues’ and ‘Collegial Interaction’ but inversely related to subscale ‘Learning by Doing Research’. This result means that ‘Educational Interest’ is positively related to ‘Learning from and with Colleagues’ and ‘Collegial Interaction’ but negatively related to ‘Learning by Doing Research. Thus, respondents seem to think that it is not possible to fulfil developmental needs which were related to educational interests through research.

The substantial findings of our study revealed that all respondent groups were extremely interested in professional development (cf. Dengerink, Luneberg, and Kools 2015). However, there were differences among respondent groups when asked how satisfied they were about their professional development and how satisfied they were with the opportunities for professional development over the past two years. In this sense, senior lecturers were the most dissatisfied sub-group. Senior lecturers’ most important interests in professional development were students’ scientific mentoring, assessment, academic writing, qualitative research skills and empirical research. They considered that the best ways to realise these interests were to research with other colleagues, read literature in the field, participate in seminars and congresses, have informal discussions with colleagues, and write publications. Like all the other sub-groups of our study, senior lecturers also believed that the worst hindrances to their professional development were lack of time, too much work and job descriptions that were too fragmented (see Kleinsasser 2017). If compared to all other respondent groups senior lecturers seemed to suffer most from fragmented job descriptions.

Of all the respondent groups, supervising teachers in training schools were the most satisfied sub-group in their professional development. However, in spite of this, they considered that their most important interests in professional development were mentoring practical teacher training, teaching skills, current developments in teacher education, subject knowledge enhancement and integrating technology into teaching. Supervising teachers considered that the best ways to realise these interests were informal discussions with colleagues, consulting colleagues, reading literature in the field, volunteer education and further training and visiting other departments of teacher education. This result is in line with various studies where the importance of communities in supporting professional development is emphasised (Boei et al. 2015; Williams 2014).

Lecturers had very similar interests in professional development to supervising teachers. Our results are a little different from those of Dengerink, Luneberg, and Kools (2015), where school-based teacher educators stress cooperation with teacher education while university-based teacher educators highlight teacher education pedagogy (p.89). However, the best ways to realise these interests were different. Both supervising teachers and lecturers believed that discussions with colleagues, consulting colleagues and reading the literature of the field were good ways to advance their professional development. However, unlike supervising teachers, lecturers considered that studying together with colleagues and doing research with other colleagues were also good ways to advance their professional development.

PhD students and researchers had a very similar response profile to our questionnaires to professors and assistant professors. Their most important interests in professional development were empirical research, academic writing and both qualitative and quantitative research skills. PhD students and researchers were also interested in improving their teaching skills, whereas professors and assistant professors were interested in current developments in teacher education. PhD students, researchers, professors and assistant professors had very similar views about the best ways to realise their professional development interests. They stated that the best ways are research with other colleagues, working in a research team, writing publications, reading literature in the field and participating seminars and congresses. According to Tack and Vanderlinde (2014), this kind of professional development may be viewed as the development of a ‘researcherly disposition’.

Conclusions and practical implications

Goodwin et al. (2014, 299) have noted that where teacher educators’ professional development is concerned, too little attention has been focused on ‘what teacher educators should know … how they should be deliberately prepared to know it, and on how they must be supported, mentored and […] inducted into the profession as scholar-practitioners’. From a practical perspective, our study provides preliminary diagnostic tools for rigorously measuring three closely related concepts: preferences for professional development, what professional learning opportunities would best meet the professional learning preferences, and things that prevent professional development. Thus, our scales may offer preliminary tools to answer the points raised by Goodwin et al. (2014).

Our study also revealed that teacher educators are not a homogenous group. Different occupational groups have different professional development interests. This must be kept in mind when teacher educators are studied. However, we were forced to break this rule in our scale development phase. Factor analysis as a method has limitations that concern the ratio of variables and sample size (see, e.g. Tabachnick and Fidell 2014, 666). Our sample size was only 354 and thus it was not possible to conduct a factor analysis with different sub-groups. This means that further evidence is needed for the construct validity of the invented scales. Whereas the internal structures of our scales were established through exploratory factor analysis, future research should concentrate on confirming the structures through multi-group confirmatory factor analysis (MGCFA) and the procedure suggested by Mulaik and Millsap (2000); see also Byman 2001, Byman (2005). It is also unknown whether the psychometric properties of used instruments are generalisable to other countries or cultures (cf. Netemeyer, Bearden, and Sharma 2003).

As stated above, different occupational groups have different professional development interests. However, the best ways to realise professional development interests are very similar. If we want to contribute to research-related professional development interests we should make sure that all workers are members of some research team and can work together with other researchers (cf. Kleinsasser 2017; Smith 2010). On the other hand, if we want to foster educational interests that are related to teaching and mentoring students we should offer the staff opportunities to engage in collective learning processes that involve joint reflection and learning from and with colleagues (cf. Smith 2010). As Dengerink, Luneberg, and Kools (2015) have stated, bringing occupational groups together while acknowledging their different preferences could help to overcome possible imbalances in different roles. Our study also revealed one worrying fact: all responder groups felt that the worst hindrance to their professional development was too much work and a lack of time. From previous studies (see, e.g. Michie and Williams 2003) we know that these feelings indicate stress, which is disastrous for professional development. Moreover, our results also showed that overly fragmented job descriptions and multiple demands (see, e.g. Smith and Flores 2019) are problematic for teacher educators’ professional development (see Kleinsasser 2017), and this problem especially concerns senior lecturers. Still, we want to emphasise that, for example, having university teacher training schools closely connected to Faculties of Education provides unique research and collaboration possibilities for Finnish teacher educators. This potential offers all teacher educators versatile possibilities for professional development. In sum, the results of our study offer preliminary tools and a deeper understanding of how universities should foster teacher educators’ professional development.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Reijo Byman

Reijo Byman, PhD., is an Adjunct Professor of Education and a university lecturer of didactics in the Faculty of Educational Sciences at the University of Helsinki. He has written several publications concerning curiosity, interest and quantitative research methods. His current research focuses on teacher educators' professional development and well-being.

Riitta Jyrhämä

Riitta Jyrhämä, PhD., is an Adjunct Professor of Education and a senior lecturer in the Faculty of Educational Sciences at the University of Helsinki. She also serves as a Head of the Department of Education. Her main research interests concern teachers' pedagogical thinking, teaching practice supervision and teacher education. 

Katariina Stenberg

Katariina Stenberg, PhD., is a university lecturer of didactics in the Faculty of Educational Sciences at the University of Helsinki. Her research interests include teacher education, teacher identity, student teachers' professional development, reflection and theory-practice relationship.

Katriina Maaranen

Katriina Maaranen, PhD., is an Adjunct Professor of Education and a university lecturer of didactics in the Faculty of Educational Sciences at the University of Helsinki. Her research focuses are teacher education, reflection, personal practical theories, professional development and teachers' professional work. 

Sara Sintonen

Sara Sintonen, PhD., is an Adjunct Professor and works as a senior lecturer in the Faculty of Educational Sciences at the University of Helsinki. Her research interests cover media education, arts education, multiliteracy, as well as early childhood education and pedagogy. 

Heikki Kynäslahti

Heikki Kynäslahti, PhD., is an Adjunct Professor of Education and a university lecturer of media education in the Faculty of Educational Sciences at the University of Helsinki. His latest research concerns theoretical framework of media education, young people's conceptions of media skills, students' role in digitalization of education and professional development of teacher educators.