Investigating the difficulty level of aspects of teaching during field placement: from coursework to practice

ABSTRACT

The purpose of this study was to explore whether particular aspects of quality of teaching belong to a single scale and identify those aspects which are more difficult to master for student teachers (STs) in the context of early field placement. It also examines whether these aspects could be grouped in stages according to their difficulty. The participants of the study were 154 STs enrolled in an early field placement program at the University of Cyprus. Real-time classroom observations were conducted in different subject areas, corresponding to the STs’ teaching efforts. Analysis of data using the Rasch and the Saltus models revealed that the eleven aspects of teaching are structured in four stages with a diverse transition, in terms of difficulty, from one stage to the other. Such findings stress the importance of a research-based coursework design during practice, based on the challenges STs face during early field placement.

KEYWORDS:

• Early field placement
• coursework
• quality of teaching
• student-teachers’ practice
• initial teacher training

Introduction

A gamut of teacher education programs during initial teacher training are expected to offer a range of courses to student teachers (STs), enabling them to apply lessons learnt in university-based coursework to teaching practices in the field (Zeichner 2010). Acknowledging the challenges emerging during STs’ first teaching efforts through (early) field placement (see Tang et al. 2019), we should rethink: which are the most prominent aspects of teaching (AOT) needed, context-wise, during STs’ transition from university coursework to (early) field placement? Which AOT are considered more or less challenging during (early) field placement? Are these AOT grouped according to their difficulty level during STs’ first teaching efforts and if so, how? Answering these questions may provide evidence leading towards a research-based design of coursework addressing STs’ needs during field placement. In this study, we use particular AOT of existing conceptual frameworks which are examined in an effort to classify the level of challenge STs encounter during their first implementation in practice. Such frameworks focus on a conceptualisation of particular dimensions and their AOT as being generic in nature and applicable across subject – domains and grade levels (e. g. classroom management and questioning, see Brophy 2006; Katharina et al. 2018; Kyriakides and Creemers 2008). Frameworks focusing on teacher-pupil interactions were also considered (see Borko 2004; Pianta and Hamre 2009). Hence, particular AOT strongly related with initial teacher training were conceptualised in the aforementioned frameworks of instructional quality (see Katharina et al. 2018; Kyriakides and Creemers 2008), namely: aspects of generic teaching, classroom organization/management and classroom learning environment.

Drawing upon studies investigating teachers’ instructional quality and professional competence (see Blömeke et al. 2022; Kyriakides et al. 2020) such AOT are expected to belong to an overarching factor (i. e. the generic factor of quality of teaching) which implies that they should be presented to the STs as interrelated and not in an isolated way (see Kyriakides, Christoforou, and Charalambous 2013). In such efforts important assumptions of ‘counteractive’ models suggesting that teachers may have distinct competence profiles with particular strengths and witnesses should not be neglected (Kunter et al. 2013). It is also important to find out whether such AOT can be grouped into different stages in terms of their difficulty levels. In this study, stages are defined as grouping categories of AOT based on the level of challenge STs encounter during enactment (Kyriakides, Creemers, and Antoniou 2009). Blömeke, Gustafsson and Shavelson (2015) argue that teachers’ competence maybe viewed along a continuum of aspects underlying perception, interpretation and decision-making skills which in turn reveals a behaviour in real school learning environments. This argument is empirically tested. Specifically, this study investigates the extent to which particular AOT vary in terms of their difficulty level and whether a single scale defining the challenges that STs encounter during early field placement can be developed. Investigating whether these AOT establish or not a single scale is related with the fact that educational effectiveness studies show that teaching factors are interrelated. However, these studies were only concerned with in-service teachers’ practices and no data on STs’ use of these AOT is available. In addition, this study investigates whether particular AOT might be more challenging for STs to enact during their first teaching efforts due to their lack of experience and expertise, namely differentiation and/or adaptive teaching (see Dack 2018) and/or formative assessment (see Windschitl et al. 2012), while others (e. g. lesson structure, see Muijs et al. 2014) might be less challenging to comprehend and enact during practice.

In this study, STs were placed in schools, as visitor teachers, for their first teaching experiences during which they followed an early field placement coursework mirroring particular AOT. This mode of placement enables STs to follow a carefully designed coursework introducing particular AOT, while they teach their first lessons in schools, before enacting full-time field placement. Early field placement is defined as STs’ first (teaching) experience on developing and implementing particular pedagogical components and their AOT during practice (Anderson, Barksdale, and Hite 2005). Consequently, this study does not only aim to identify the most challenging AOT during practice but also to investigate whether these AOT can be classified into different groups that may follow a different level of challenge during STs’ transition.

Early field placement components and their aspects of teaching during coursework

Research results on relating teaching components (and their AOT) with learning reveal that ‘black box’ models seeking to directly relate such components to pupils’ learning are not always accurate (Blömeke et al. 2022). Educational effectiveness research acknowledges potentially confounding variables that would create bias in the results (Koedel et al. 2015), like pupils’ prior achievement, pupils’ background and school context, whereas competency-based approaches oversimplify teaching as a group of isolated components which can be learned in a number of training sessions (see Kokkinou and Kyriakides 2022). An implicit assumption in these models suggests a fixed set of (generic) teaching skills which must be performed optimally from STs to expert teachers, ignoring the dynamism of skills and changes in skills (Raduan and Na 2020). While there is empirical evidence to suggest that particular teaching skills related with the corresponding AOT might be crucial for high-quality instruction (see Blömeke et al. 2022; Kyriakides and Creemers 2008), some studies suggest distinct competence profiles contradicting the idea of how global components (and their AOT) might be interrelated (Kunter et al. 2013). Such evidence concern mainly in-service teacher professional development whereas a lack of research in STs’ initial training indicates that such interrelations need further examination. For instance, Kyriakides, Creemers and Antoniou (2009) suggested that teaching skills could be grouped into five types of teacher behaviour which are distinct and move gradually from skills associated with more advanced skills concerned with differentiation of teaching, while this may not be the case for STs (see Dack 2018). Research conceptualisation on teaching expertise seems to be moving from a successive, linear, hierarchical and progressive development towards a dynamic approach of expertise, leading to the need for a multidimensional model of development integrating different components of expertise (Raduan and Na 2020).

A considerable amount of research advocates in favour of approaches considering the complexity between teacher learning and practice (e.g. Boylan et al. 2018; Strom and Mitchell Viesca 2021) criticising the linear and ‘process-product’ driven way of thinking (Strom and Adrian 2017). However, there are no conclusive findings regarding how introducing particular field placement components and their AOT can be more or less conducive towards STs’ quality of teaching (Azkiyah and Mukminin 2017; Ronfeldt 2012). This study aims to address the need of generating reliable and valid measures on how STs implement these AOT during early field placement. We concentrate on global components and their AOT (Borko 2004) emerging from STs’ needs, namely: particular aspects of generic teaching, classroom organization/management and classroom learning environment.

Aspects of generic teaching

Pupils’ active engagement in achieving learning outcomes

Learning as a self-regulated pupil-centred and constructive process (Bransford, Brown, and Cocking 2000) implies that pupils’ active engagement should be encouraged (Hampton and Reiser 2004). Provoking pupils’ interests is considered as a fundamental aspect of STs’ teaching performance in authentic learning environments during field placement (Houtveen et al. 1999). Acknowledging pupils’ interests, their different needs and characteristics is likely to promote pupils’ learning outcomes. STs are also expected to adapt their teaching to the needs of different groups of pupils (Maulana, Helms-Lorenz, and Van de Grift 2017). As a consequence, this AOT addresses the extent to which STs differentiate their instruction (see Dack 2018).

Lesson structure

The selection and structure, of a series of tasks in goal-oriented activities, which are progressively enrolled in addressing prospective learning outcomes, have been considered crucial in teaching effectiveness research (Kyriakides et al. 2020; Seidel and Shavelson 2007). Activity sequence and gradation in the classroom remains an AOT in which STs inevitably integrate a range of further decisions during enactment (König et al. 2020). Making the activity sequence meaningful for pupils is also related with knowing what is expected of them during the lesson, hence an explicit lesson structure adjusted on pupils’ needs and skills is an important indicator of instructional clarity (Muijs et al. 2014).

Strategies and practices

Using various strategies and teaching practices seems to be the corner stone of classroom learning environments, namely: group work, autonomous work, whole class discussion, peer interactions and teacher-pupil interactions (Maulana, Helms-Lorenz, and van de Grift 2015). Drawing on the findings of teacher effectiveness research, the integrated approach to teaching, in which both direct instruction (cf. Joyce 2000) and constructivist approaches blend together, overcomes dilemmas related with the use of one instructional practice over the other (see Creemers, Kyriakides, and Antoniou 2013). In this context, question posing remains a fundamental feature in engaging and challenging learners (Harrop and Swinson 2003). A skilful use of purposeful questions guides classroom discussion through a series of aligned and goal-oriented activities, remaining a priority at both fronts: STs’ teaching practice and field placement coursework (Azkiyah and Mukminin 2017).

Assessment and feedback

Providing scaffolding and feedback or teaching pupils strategies for self-regulation (Boekaerts 1999) is strongly connected with experiences STs need to endure during their early field experience. Research stresses the importance of scaffolding and feedback towards pupils’ performance and the achievement of learning goals (C. Smith 2008). Any goal-directed learning calls STs and their pupils to reflect upon those learning goals and assess their achievement. Consequently, it is expected that assessment will be conducted for formative purposes (Cowan 2009; Moore 2003). Formative assessment brings together pupils’ ideas with STs’ knowledge of their teaching in an evidence-informed system of learning activities (Windschitl et al. 2012). In this case, the term formative assessment is tailored to the practices of eliciting and providing constructive feedback to pupils’ ideas (including errors) and organising classroom activities around those (alternative) ideas (Bennett 2011; Windschitl et al. 2012).

Classroom organization and management

One of STs’ main concerns during practice is classroom management (O’Neill and Stephenson 2012) which is related with developing procedures and routines to maximise the use of teaching time, including ways of dealing with pupils’ misbehaviour and unexpected events (Meister and Melnick 2003). Organizing classroom materials, means and working strategy (e. g. concrete physical materials, technological tools) is also considered crucial (see O’Neill and Stephenson 2012). Various studies conducted in different countries show that time management is one of the greatest concerns for beginning teachers since they are not always aware of how much time pupils need to deliver a task (Kyriakides et al. 2017; Melnick and Meister 2008; Moore 2003). However, only a few studies investigated STs’ use of teaching time (Cakmak 2008).

Classroom learning environment

Treating classroom learning environment as a different component from the classroom organisation and management component reveals the importance of establishing a supportive and business-like classroom learning environment. A major aspect of this component is the encouragement of teacher – pupil and pupil – pupil relationships/interactions (Korpershoek et al. 2016; Kyriakides et al. 2020). Teacher-pupil interpersonal relations in the context of positive and authentic teaching approaches lead to a relaxing learning atmosphere (Kyriakides et al. 2020; Opdenakker, Maulana, and den Brok 2012). For advancing a stimulating environment, teachers’ instructions are not provided authoritatively, but concisely and inclusively leading to a motivating environment (Maulana, Helms-Lorenz, and Van de Grift 2017; Moore 2003). Establishing conventions of social interactions, based on positive verbal and/or nonverbal communication practices during group or autonomous work, is considered part of such an environment (Maulana, Helms-Lorenz, and van de Grift 2015).

Context of the study

In this study, we focus on early field placement in which STs have the opportunity of teaching their first lessons. In designing an early field placement coursework pedagogy, we considered basic theoretical and pedagogical implications supporting practice as they are documented in the literature, translating them into specific teaching components and their corresponding AOT (see Table 1). After defining those AOT, we sought to structure a university coursework based on STs’ needs and the challenges they face during early field placement (e.g. Creemers, Kyriakides, and Antoniou 2013; Moore 2003). Observing and analysing the components (and their AOT) during teaching practice was considered crucial. Observation data were expected to identify whether particular AOT belong to a single factor and if so, the extent to which those AOT could be classified in stages in terms of their difficulty level. The classification of AOT into different stages does not imply that STs are expected to move in a predetermined sequence of progression but it acknowledges the diversity of prospective stages at an introductory level during early field placement.

Table 1. Components and their aspects of teaching included in the early field placement.

Components	Aspects of Teaching
Aspects of generic teaching	Pupils’ active engagement in achieving learning outcomes Active engagement Provoking pupils’ interests/concern
	Lesson structure Adjustment of pupils’ prior knowledge/experiences/skills and needs and relating pupils’ needs with each activity and especially with the context to be learned Activity sequence and gradation Activities addressing the prospective learning outcomes
	Strategies and Practices Engaging various teaching practices Quality of questions based on the content and goals
	Assessment and Feedback Formative assessment Feedback
Classroom management/organisation	Organising space, materials, means and working strategy (whole class, in groups, atomically)
	Establishing procedures and routines
	Managing unexpected events
	Time management
Classroom learning environment	Positive and authentic approach
	Pupils’ interactions and teacher-pupil interactions
	Verbal and non-verbal communication
	Providing pupils with explicit instructions

Aims

Various teacher effectiveness studies conducted in different countries (see Kyriakides et al. 2020) and relevant meta-analyses (e.g. Kyriakides, Christoforou, and Charalambous 2013; Scheerens 2016) reveal that the AOT described above are likely to be related to each other and may belong to an overarching factor (i. e. the generic factor of quality of teaching). This study attempts to test this hypothesis during early field placement and in particular to find out whether all AOT can help us establish a single scale, showcasing the difficulty STs encounter in respect of their implementation during teaching. Different AOT maybe interrelated, but this does not imply that they belong to the same overarching scale. Thus, analysis of data may reveal that either some AOT do not belong to the common scale or that a multidimensional scale should have been developed. Therefore, this study investigates the extent to which these AOT are not only interrelated but can help us establish a scale in order to find out whether these generic AOT can be grouped according to their difficulty level during enactment. Our hypothesis, purposefully, does not take into account the trajectories of AOT emerging through STs’ continuous transformation during placement, since we seek to establish an introductory early field placement coursework based on STs’ initial needs on their first teaching efforts. Hence, we emphasise on particular generic AOT and whether those are interrelated and/or grouped according to their difficulty level during enactment. The construct validity of the observation protocol used to measure these AOT during STs’ practice by means of confirmatory factor analysis is, initially, examined. The criterion-reference and the predictive validity are also examined. Finally, this study does not only examine the scaling of the AOT implemented during coursework and enacted during practice, using the Rasch model (Bond and Fox 2001), but it also examines whether STs can be grouped into stages using the Saltus model (Mislevy and Wilson 1996). Thus, this study aims to provide answers to the following questions:

• Do the AOT implemented during early field placement coursework belong to a single overarching factor and if so, to what extent STs’ scores on this single scale:

  1. are associated with their grades on a theoretical course on the methodology of teaching and

  2. can predict STs’ overall performance during the final phase of field placement?

• Can the AOT, implemented during early field placement coursework, be classified in terms of their difficulty and if so, which particular combinations of AOT are more or less challenging for STs during early field placement?

Methods

Sample and setting

The participants were 154 STs (139 females, 15 males) enrolled at an early field placement program at the University of Cyprus. These STs were randomly selected for real classroom observation during their first teaching efforts from a list of 310 STs enrolled in the early field placement. STs participated in a full semester program, comprising the coursework and the field placement teaching practice. During coursework, STs had three-hours weekly meetings for thirteen weeks, in which they followed a structured program involving particular AOT selected through the literature review (see Blömeke, Gustafsson, and Shavelson 2015; Kyriakides and Creemers 2008; Seidel and Shavelson 2007) corresponding to three teaching components: aspects of generic teaching, classroom organization/management and classroom learning environment. These AOT were theoretically introduced during coursework, surrounding the context at hand with practical ideas and examples from different subject areas. STs had the chance to critically analyse research papers arguing about the importance of each AOT and discuss under which conditions and how each AOT may promote learning. All AOT were introduced and discussed during the first nine weeks of the coursework, whereas in the four remaining weeks STs reflected on implementing those AOT in their teaching practice.

The field placement program

The field placement program comprises three phases, namely: i) the preparatory field placement, ii) the early field placement and iii) the final field placement. The main priority of early field placement is to gain teaching experiences and at the same time bring together STs’ underlying theoretical principles with their actual teaching practice. To this end, a coursework, aligned with their teaching experiences in schools is conducted, which emphasises on the theoretical and practical implications on planning and enactment of the teaching components and their AOT presented in Table 1.

Data sources

Real-time classroom observations were conducted, corresponding to at least two lessons of each ST during the second phase of early field placement. Data were also collected on STs’ overall performance on a previous course regarding the methodology of teaching taught by another instructor, and their overall performance during their final field placement. Data collection approval was obtained by the ethics committee of the Department of Education of the University of Cyprus; STs’ consents were also collected. All data collected were coded and analysed anonymously. The same observer (i. e. the first author of this paper) conducted all the observations on a weekly basis according to STs’ timeline without having access to their overall performance during the theoretical course on methodology of teaching. The observation protocol used included all the AOT implemented during early field placement coursework. The observation protocol was developed by faculty members of the field placement committee and tutors who were enrolled in an ongoing, development and refinement, procedure which lasted approximately three years. The procedure involved discussing/analysing every component’s item while observing STs’ lessons during field placement. A number of video-recorded lessons during field placement were used to test the content validity of the observation instrument. Six cycles of video observations and real time discussions based on classroom observations were conducted (from 18 tutors and 7 teacher educators). During this procedure, constant analysis and reflection based on two fronts were conducted: i) selecting components/AOT and ii) the use of each component/AOT by STs during lessons. As a result, the observation protocol received the form used in the current study. Each AOT was coded on a four-point Likert scale representing the quality of enactment, ranging from 1 (predominantly weak) to 4 (predominantly strong). For each aspect, second level explanatory codes were used in order to reach the final coding on the four-point scale (see the codes of each AOT in Appendix A). Finally, support to the content validity of the final version of the observation protocol was provided by each member (n = 25) of the expert group mentioned above.

For each lesson observed, an observation rubric protocol was completed (based on STs’ efforts on each AOT) and a reflection diary was produced to justify the codes given for each lesson. The observation protocol developed was in line with the AOT included in the three main components used in the analysis (i. e. generic teaching skills, classroom management/organisation, classroom learning environment). The results of several meta-analyses were considered while selecting components and their AOT. These meta-analyses reveal the generic nature of specific aspects of effective teaching (e. g. Kyriakides, Christoforou, and Charalambous 2013; Seidel and Shavelson 2007). Specifically, these AOT were found to promote pupils’ learning outcomes in different domains and subjects (Katharina et al. 2018; Scheerens 2016). Components of domain specific AOT were excluded since this was an early phase of field placement, focusing only on generic AOT.

The observation protocol was used as a high inference observational instrument (see Borich and Klinzing 1984), hence coding STs’ performance on each code by the end of the lesson based on the observer’s general impression (as it is usually the case in using such observational instruments, see Borich and Klinzing 1984; O’Leary 2020). It is acknowledged that there are some instruments which expect observers to rate separately different segments/periods of the lesson, rather than the whole lesson. However, studies investigating the quality of these instruments reveal that data emerged from these instruments are generalisable to the level of the lesson (Creemers, Kyriakides, and Sammons 2010; Gitomer et al. 2014). For this reason, the observer was expected to evaluate the whole lesson in regard to each AOT. It was also explained that, for example, if a ST provided adequate feedback to pupils for a significant amount of time on the task(s) at hand during the lesson, the code predominantly adequate would be used concerning the corresponding AOT (i. e. feedback).

Only one observer was used during the lessons’ observation, and it is, therefore, not possible to test the inter-rater reliability. One could argue that the descriptors used in the observational instrument were generic in nature, hence vulnerable to multiple interpretations. However, an intensive training session on the use of the observation protocol¹ was conducted preceding the data collection phase. During the training session, several observers (including the first author) were enrolled, by applying each descriptor on real-time video observations of STs’ teaching efforts in field placement. The second level explanatory codes and the terms used (i. e. absence, inadequate, adequate, excellent) were explicitly defined during the training session in terms of coding (see an example of a classroom management AOT in Appendix A). The inter-rater reliability of each observer was found to be higher than 0.80. The internal reliability of the data emerged from this study was satisfactory (see the results section). These results corroborate with tutors’ extensive training during video observation, since a high level of analysis of each descriptor was enacted during coding and interpretation of each video segment used. Since more than one observation per ST were conducted, one-way ANOVA was employed and revealed that the data were generalisable at the level of ST {F(152, 285) = 18.6, p < .001}.

Data analysis

Using structural equation modelling to test the construct validity of the instrument

Structural equation modelling (SEM) was employed to test the construct validity of the observation instrument. Specifically, Confirmatory Factor Analysis (CFA) was conducted to examine if the internal factor structure of the observation instrument was in line with the theoretical framework used to measure the AOT shown in Τable 2. Before running the CFA, a descriptive statistical analysis was conducted to identify any items suffering from either ceiling or flooring effect. AOT concerned with differentiation of teaching was found to suffer from the flooring effect, since STs’ success rate was extremely low {i. e. almost all of them (94%) had a score lower than the adequate level (i. e. 1 or 2)}. Flooring effect was also observed for the motives aspect under the classroom learning environment (i. e. almost 90% of students received a score lower than the adequate level). No other item was found to suffer either from the ceiling or the flooring effect. Thus, data from all the other items were considered in testing the factorial structure of the instrument. In line with the theoretical framework of the instrument, a three-factor model was tested, where each item representing performance on one of the three components was prescribed to load on a single factor, and the three factors were allowed to correlate with each other. The fit of this model was good (i. e. X² (df = 41) = 47.2, p < .001; CFI = 0.961; RMSEA = 0.038) and all parameter estimates were statistically significant (p < .001). Moreover, all factor loadings were higher than .67 whereas the between-factor correlations were higher than .50. The magnitude of the factor intercorrelations justifies our attempt to identify a higher-order model which may explain the correlations among the three first-order factors. In this model, all three first-order factors were regressed on a second-order factor. The standardised path coefficients relating the first-order factors to the second-order factors were higher than .55 and significant (p < .001). The fit statistics (i. e. X² (df = 38) = 37.8, p < .001; CFI = 0.972; RMSEA = 0.035) were acceptable. By comparing the second-order factor model with the theoretical first-order factor model, we could identify a minor decrease of the RMSEA (i. e. from 0.038 to 0.035) and a minor increase of the CFI (i. e. from 0.961 to 0.972). Thus, the single second-order factor model was considered as preferable for both statistical reasons and reasons of parsimony.

Using the Rasch model to specify the hierarchy of AOT difficulties

The identification of a strong second-order factor reveals that performance on the whole set of items included in the observation protocol could be reducible to a common scale. The Rasch model is appropriate for the specification of this scale because it enables the researcher to test the extent to which the data meet the requirement that both STs’ performances on the items of the protocol and the difficulties of the items form a stable sequence (within probabilistic constraints) along a single continuum (Bond and Fox 2001). Rasch analysis provides indicators of how well each item of an instrument (in this study each AOT) fits within the underlying construct and thereby the construct validity of the instrument can be examined (E. Smith 2001). In regard to the criterion-reference and the predictive validity, statistically significant correlations with two other scores [i. e. grades on i) a course on the methodology of teaching and ii) overall performance during the final phase of field placement] were identified, thus providing empirical support to the validity of the observation instrument.

Using cluster analysis to specify stages of effective teaching

Having established the reliability of the scale, one might ask if the various AOT are systematically grouped by level of difficulty, which move from relatively easy to more difficult. As such, the procedure for detecting pattern clustering in measurement designs developed by Marcoulides and Drezner (1999) was used. This procedure enables us to segment the observed measurements into constituent groups (or clusters) so that the members of any group are similar to each other, according to selected criterion that stands for difficulty.

Using the saltus model to specify the structure of AOT

The Rasch model cannot provide answers on how deep the divide is separating the stages emerging from cluster analysis and which can be ordered into different stages according to their difficulty. Wilson developed a variant of the Rasch model, the so-called Saltus model (Mislevy and Wilson 1996; Wilson 1989), as a method that can differentiate between different levels. Specifically, we use the Saltus model to differentiate between major and less pervasive changes in moving from one stage to the other without sacrificing the idea of one common underlying continuum.

Findings

The Rasch model was applied on the whole sample (N = 154) and there was no item or participant who did not fit to the model. Specifically, all items were found to have item infit with the range 0.73 up to 1.20, and item outfit with the range of 0.71 up to 1.42. Moreover, the separation indices of items were found to be higher than 0.90. This implies that the reliability of the scale is high (Bond and Fox 2001). Finally, the infit mean squares and the outfit mean squares are 1 and the values of the infit t scores and the outfit t scores are approximately zero.

In order to further investigate the validity of the Rasch scale measuring STs overall performance on the whole set of items included in the observation protocol, we searched for the relations between the STs’ Rasch score and the overall grade of STs in a course on the methodology of teaching offered to them during the second year of their study. Pearson correlation coefficient was found to be statistically significant and relatively high (r = 0.46, n = 152, p < .001). We also examined the predictive validity of the instrument by searching for the relation between the Rasch score of STs and their overall grade/performance during the final phase of field placement which took place one year after their early field placement coursework. A statistically significant relation with a high correlation coefficient was also identified (r = 0.51, n = 151, p < .001). Tutors offering the final phase of the field placement had no access to the results of the Rasch analysis of STs’ performance during the early field placement.

Having established the reliability and the validity of the Rasch scale, cluster analysis was employed to classify the 11 items into specific stages by considering their difficulty level. It was found that they can be optimally clustered into four types of ST behaviour (i. e. stages of teaching). Specifically, the cumulative D for the four-cluster solution was 72% whereas the fifth gap adds only 3%. These four clusters are further explored and specified by using the Saltus model.

To apply the Saltus model, we assumed that the 11 AOT are structured in the four stages of effective teaching identified through the cluster analysis. The Saltus solution was found to represent a better fit to the actual data rather than the Rasch model. This solution offers a statistically significant improvement over the Rasch model which is equal to 703 chi-square units at the cost of 20 additional parameters (i. e. 9 t values, 4 means, 4 standard deviations, and 3 independent proportions). Table 2 presents the difficulty parameters of the 11 AOT for STs in the easiest stage of effective teaching (i. e. stage 1 shown in column 3) and the implied within stage difficulty (i. e. columns 4, 5, and 6). A comprehensive account of the Saltus model is presented in Appendix B of the supplementary material.

Table 2. Rasch and Saltus parameter estimates for factor scores measuring aspects of quality of teaching on STs’ teaching efforts during early field placement.

			Rasch	Implied within-stage difficulty (Saltus)
Aspects of quality of teaching			All	Level 1	Level 2	Level 3	Level 4
Positive and authentic approach			−1,79	−2,21	−2,21	−2,21	−2,21
Strategies and practices			−0,93	−1,08	−1,92	−1,89	−1,87
Organizing space, materials, means and working strategy (whole class, in groups, atomically)			−0,84	−0,94	−1,88	−1,8	−1,78
Lesson structure			−0,82	−0,9	−1,76	−1,71	−1,7
Verbal and non-verbal communication			−0,54	−0,78	−1,7	−1,66	−1,62
Pupils’ active engagement in achieving learning outcomes			−0,35	−0,71	−1,65	−1,62	−1,65
Pupils’ interactions and teacher-pupil interactions			0,32	0,35	0,13	−1,53	−1,57
Providing pupils with explicit instructions			0,44	0,48	0,24	−1,41	−1,48
Managing unexpected events			0,51	0,58	0,27	−1,35	−1,39
Time management			1,05	1,72	1,58	−0,13	−1,21
Feedback and assessment			1,08	1,89	1,64	0,23	−1,12
The saltus parameter estimates (i. e. τ values)
Item class
	Examinee stage
	1	2	3	4
1	0*	0*	0*	0*
2	0*	0,90	0,86	0,85
3	0*	0,26	1,90	1,95
4	0*	0,20	1,76	2,97

Empty lines are used to separate the four levels of STs’ performance emerged by cluster analysis.

*Fixed at zero for model identification.

The following observations arise from this table. First, difficulty parameters of ΑΟΤ for STs in Stage 1 (i. e. the values, shown in the third column of Table 2) are more spread out than those of the Rasch Model (shown in the second column). This finding reveals on one hand that for STs in Stage 1 a large gap between the AOT of Stage 1 and the AOT of Stages 2, 3 and 4 can be observed. On the other hand, for STs who belong to Stage 2, AOT of Stage 2 are as easy as AOT of Stage 1. Stage 3 AOT are relatively difficult for Stage 2 but are as easy as Stage 2 AOT for Stage 3 STs. Similar observations can be made in relation to Stage 4 AOT since STs at Stage 4 found these AOT as difficult as Stage 3 AOT whereas, all the other STs found these AOT as the most difficult. Based on Table 2, we then calculated the asymmetry and segmentation indices (see Appendix B) and found that the gap between Stages 1 and 2 is much smaller than the gap between Stages 2 and 3 and the gap between Stages 3 and 4. This implies that the transitions between consecutive stages are not equally difficult. Specifically, the difficulty of transition from Stage 2 to Stage 3 in comparison with the transitions from Stage 1 to Stage 2, delineates STs difficulty on moving from organising working strategy for pupils, enacting practices in activating them during the lesson using positive verbal and non-verbal communication (Stage 2) towards managing unexpected events, providing pupils with explicit instructions and causing meaningful interactions within the lesson (Stage 3). Results show that the transition from STs’ positive and authentic approach (Stage 1) to enacting practices and organising working strategy in relation with all the other AOT of Stage 2 was not as difficult as the transition discussed above between Stage 2 and Stage 3. The transition from Stage 3 to Stage 4 was the most difficult one.

Discussion

This study generates some support to the early field placement coursework structure since the AOT introduced during coursework were found to be interrelated and to fit the Rasch model. These findings also reveal that AOT should not be treated as isolated skills. In addition, descriptive analysis revealed that two AOT concerned with differentiation and pupils’ motivation suffered from the flooring effect providing empirical support to the initial assumption that STs lack clear understandings of how differentiation can be introduced in their teaching (see Dack 2018). Providing adaptive teaching was found to be very challenging even for expert teachers (Kyriakides et al. 2020). However, differentiation affects the motivational aspect of teaching, since pupils’ characteristics and different needs should not be neglected (see Maulana, Helms-Lorenz, and van de Grift 2015). In this context, we argue for further research exploring ways to support STs differentiate their instruction, thus contributing to the development of pupils’ positive attitudes towards learning.

Second, the results of CFA revealed that all AOT belong to a single overarching factor and support to the construct validity of the observation instrument is provided. In addition, the Rasch analysis shows that all AOT in this early field placement coursework belong to a scale that meets the requirements of unidimensionality. These findings imply that all aspects of generic teaching, including classroom management/organisation and learning environment are interrelated. This result is in line with current literature on teacher improvement (e.g. Creemers, Kyriakides, and Antoniou 2013; Scheerens 2016) which reveals the weakness of the competency-based approach. This implies that the coursework should follow a more comprehensive approach demonstrating the interrelations among the AOT and revealing the complexity and dynamic nature of teaching. Such results seem to provide empirical support to those arguing for the complexity between teacher learning and practice (e.g. Boylan et al. 2018) contrasting the homogenous, linear perspectives of rationalistic approaches (Strom and Viesca 2021).

Finally, this study was in a position to classify AOT in four specific stages. These combinations of AOT imply that specific AOT are not only interrelated, but their difficulty levels are similar. This implies that they can be presented together during coursework and relevant support to STs should be provided to help them use each combination of AOT during teaching (Creemers, Kyriakides, and Antoniou 2013). The Saltus model has also shown that STs follow specific challenges in moving from one stage to the other and reveals the weaknesses of a linear progression perspective during STs’ teaching practice.

The four types of teacher behaviour clustered in the Saltus model suggest a particular research direction towards the design of the coursework. The results also showed that the most challenging transition was from Stage 3 to Stage 4, meaning that STs possessing the generic AOT as well as the components of classroom management/organisation and classroom learning environment presented in Table 1, confronted a great challenge in managing time and in providing feedback and conducting formative assessment during their teaching. The grouping of time management, feedback and assessment at the fourth stage of the Saltus model indicates the need of further research on how to promote (if possible) these AOT during early field placement coursework. STs’ issues with time management, feedback and assessment are discussed in the literature and were considered during coursework, albeit this implementation was not enough (Beeth and Adadan 2006; Moore 2003). For instance, our findings on time management corroborate with the literature indicating that this particular AOT is considered to be of grave importance for beginning teachers (Melnick and Meister 2008). Although time management might not be the most prominent task STs have to deal with in the first place, it seems from our results that this AOT is one of the most difficult for STs to achieve. A similar challenge emerged regarding the difficulty level of the AOT concerned with formative assessment and specifically on assessing the learning outcomes when exploring lesson planning and teaching during early field placement (see Windschitl et al. 2012).

Other than Stage 4, a number of most challenging AOT categorised on Stage 3 were related with classroom organisation and management (managing unexpected events) and classroom learning environment (encouraging teacher-pupil and pupil-pupil interactions and providing pupils with explicit instructions). STs are not really familiar with effective practices on organising and/or provoking interactive behaviour among pupils, albeit their constant efforts in planning such activities sometimes counteract their lack of teaching experience to this end. In addition, establishing positive and authentic learning environments (Hampton and Reiser 2004) (Stage 1) and planning and enacting various engaging practices (Maulana, Helms-Lorenz, and van de Grift 2015) (Stage 2) are not enough for STs to promote classroom interaction, provide pupils with explicit instructions and manage unexpected events (Kyriakides et al. 2020).

Exploring STs’ transitions among the different stages of teaching behaviour during practice is strongly related with STs’ different needs, which begs the question: How do we customise STs’ guidance during early field placement coursework (and practice)? How do we support STs’ transition from one stage to the next? We argue that STs in Stage 1 may need different support during coursework than their counterparts in Stage 2, while this being the case for each of the subsequent stages presented. The stages emerged from the Saltus model provide implications for further research on how we can introduce such AOT during early field placement and whether more explicit guidance is needed during both, coursework and practice. Since STs may follow different learning trajectories during practice than in-service teachers, a customisation of early field coursework towards this end (i. e. differentiation) needs further examination. Such efforts may not only provide implications for (re)designing early field coursework at a university level but may also provide important guidance to mentors in further scaffolding STs who confront particular difficulties (Moore 2003; Perry, Hutchinson, and Thauberger 2008).

The main study limitations are acknowledged. Studies testing the generalisability of these findings are needed. Since this study is focused on particular AOT, future work could also look at other AOT to explore whether and how a comprehensive approach in defining AOT can help us improve the quality of the early field placement. Longitudinal studies should also investigate whether and how STs manage to move from one stage to the next more demanding. Further studies should also investigate the extent to which background factors (e. g. academic performance, year of study) explain improvement in the performance of STs. The type of challenges STs face during early and final field placement can be compared.

Finally, research on effective teaching recognise the importance of searching for the impact of contextual factors on teaching quality (e. g. Campbell et al. 2004; Kokkinou and Kyriakides 2022; Muijs et al. 2014). Specific characteristics of quality of teaching were found to explain variation in student achievement in different types of learning outcomes (not only cognitive). For this reason, these characteristics are treated as generic, meaning that they contribute to the promotion of learning in different subjects and in different phases of education (see Muijs et al. 2014; Scheerens 2016). It is, however, important to note that there is almost no study searching for the extent to which STs’ behaviour depends on the context in which they have to work (e. g. different age groups of pupils, different school subjects, different types of learning outcomes in different classrooms) (see Kyriakides et al. 2020). Research is, therefore, needed on how the context may influence the development of STs’ skills in teaching.

The consistent dissonance between theory and practice in (early) field placement environments, remains a constant challenge to universities (Zeichner 2010). The well-established mechanisms of excessive resources like mentoring and tutoring have long tried to bridge university coursework and teaching practice during field placement without always addressing STs’ traditional challenges (Moore 2003). In bridging coursework and practice there are no readymade solutions. However, the study reported here may be considered as a starting point of (re)designing a research-based early field placement coursework during teaching practice, giving meaning to both, the theoretical underpinnings and the practical ideas surrounding the context at hand.

Acknowledgments

We would like to thank the STs who participated in the study, thus enabling us learn with and from them. We would also like to thank our partners in the Field Placement Committee and the Field Placement Group of the Department of Education at the University of Cyprus, for their contribution in the design of the observation protocol and related material.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/02619768.2024.2370887.

Additional information

Notes on contributors

George Olympiou

George Olympiou is the Coordinator of Field Placement program of the Department of Education at the University of Cyprus. He completed a B.A. in Education and his graduate studies (M.A. and PhD) in Science Education at the University of Cyprus. He worked as a researcher in several research projects concerning initial teacher training and inquiry-based learning that received continuous financial support over the years from the Cyprus Research Promotion Foundation and the European Commission. Dr Olympiou research interests include initial teacher training and learning by observing, analyzing and reflecting upon actual teaching practice. Dr Olympiou has authored or co-authored papers published in peer-reviewed journals including Learning and Instruction, Journal of Teaching and Teacher Education, Journal of Research in Science Teaching, Science Education and Instructional Science.

Leonidas Kyriakides

Leonidas Kyriakides is Professor of Educational Research and Evaluation at the University of Cyprus. His field of research and scholarship is the evaluation of educational effectiveness, whether of teachers, schools or educational systems. Currently his research agenda is concerned with the development of a dynamic model of educational effectiveness, and the application of research to the improvement of educational practice. Leonidas has been involved in several international projects. His work has contributed not only to theory improvement but also to the testing of theoretical models and using them for improving quality and equity in education. Finally, he is the author of more than 130 papers, 8 books, and 100 chapters in books.

Notes

1. The observation protocol was developed by the field placement group coordinated by the field placement committee before this study was conducted.