FACULTY OF SOCIAL- AND BEHAVIORAL SCIENCES Graduate School of Childhood Development and Education
The Development and First Test of a Reflection Coding Instrument Designed to Measure Teacher Reflection During Teacher-Child Interaction Coaching
Master Thesis Studies of Special Education in the Family and at School Pedagogical Science University of Amsterdam
Name: S. Brinkman Student number: 10005897 Thesis supervisor: Dr. H.M.Y. Koomen
Second evaluator: Dr. M. Zee Amsterdam, April 2017
The Development and First Test of a Reflection Coding Instrument Designed to Measure Teacher Reflection During Teacher-Child Interaction Coaching
Master Thesis Studies of Special Education in the Family and at School Pedagogical Science University of Amsterdam
Name: S. Brinkman Student number: 10005897 Thesis supervisor: Dr. H.M.Y. Koomen
Second evaluator: Dr. M. Zee Amsterdam, April 2017
Contents
Abstract………...4
Samenvatting………...5
The development and first test of a reflection coding instrument designed to measure teacher reflection during teacher-child interaction coaching………...6
Attachment theory and the developmental systems theory……….6
Interventions for the teacher-child relationship………..8
The relationship-focused reflection program………10
A history of reflection and reflection research………..12
The present study………..13
Method………..15
Participants and procedure………15
Development of the Reflection Coding Instrument………..16
Intervention script: the second session of the RFRP……….…17
Measures and missing data………18
Nature and degree of teacher reflection……….18
Teacher’s perspective taking……….18
Attributions of student behaviour………..19
Teacher’s perception of relationship change……….19
Data analysis……….19
Results………...20
Descriptive information on reflection scales……….20
Relation between teacher reflection and teachers’ perspective taking………..22
Relation between teacher reflection and teachers’ attribution of student behaviour…23 Relation between teacher reflection and change in the teacher-child relationship…...23
Examination of the treatment integrity………..24
Explanation of constructs………..24
Explanation of received ratings………...25
Stimulation for reflection………..25
Use of interview techniques………..28
Discussion……….………28
References……….34
Appendix 1 - First version of the Reflection Coding Instrument (Dutch version)…………...38
Appendix 3 - Second version of the Reflection Coding Instrument (Dutch version)………...42 Appendix 4 - Second version of the Reflection Coding Instrument (English version)………44 Appendix 5 - Third version of the Reflection Coding Instrument (Dutch version)…………..46 Appendix 6 - Third version of the Reflection Coding Instrument (English version)………...49
Abstract
The Relationship-Focused Reflection Program (RFRP) aims to improve dyadic relationships between teachers and disruptive students. It is assumed that to achieve this goal teachers must reflect on their relationship with individual students. The current research aimed to design a Reflection Coding Instrument to quantitatively measure the nature and degree of teacher reflection. The overall usefulness and the psychometric properties, among which aspects of validity, were tested in this explorative study. The Reflection Coding Instrument was developed by using a grounded theory approach and included the dimensions focus, change and elaboration. Audiotapes of the RFRP from twenty-four dyadic relationships were listened to, transcribed and scored for teacher reflection. To examine aspects of validity, the reflection scores on the dimensions were associated with consultants’ ratings of
perspective taking, teachers’ attribution of student behaviour and teachers’ experience of positive/negative change in the teacher-child relationship. Positive associations were found between the focus dimension and the consultants’ ratings of perspective taking and between the change dimension and the teachers’ attribution of student behaviour to the teachers’ own way of teaching. Negative associations were found between the change dimension and the teachers’ attribution of student behaviour to the character of the student. In particular, the results supported the validity of the dimensions focus and change. Less evidence was found for the validity of the elaboration dimension.
Keywords: Teacher reflection; Teacher-child relationship; Reflection Coding Instrument; Relationship-Focused Reflection Program
Samenvatting
Het “Relationship-Focused Reflection Program” (RFRP) richt zich op het verbeteren van de dyadische relatie tussen leerkrachten en probleemleerlingen. Er wordt verondersteld dat leerkrachten moeten reflecteren op hun relatie met individuele leerlingen om dit te bereiken. De huidige studie richtte zich op het ontwerpen van een reflectiecoderingsinstrument
waarmee de aard en mate van leerkrachtreflectie kwantitatief kon worden gemeten. De
algehele bruikbaarheid en psychometrische eigenschappen, waaronder aspecten van validiteit, werden in dit exploratieve onderzoek getest. Het reflectiecoderingsinstrument is ontwikkeld door gebruik te maken van de “grounded theory” benadering en bevatte de dimensies
oriëntatie (focus), verandering (change) en uitwerking (elaboration). Audio-opnames van het RFRP van vierentwintig dyadische relaties werden beluisterd, getranscribeerd en gescoord op leerkrachtreflectie. Om aspecten van validiteit te onderzoeken, werden reflectiescores op de dimensies in verband gebracht met perspectief nemen beoordeeld door de gespreksleider, hetgeen waaraan de leerkracht leerlinggedrag toeschrijft en de door de leerkracht ervaren positieve/negatieve verandering in de leerkracht-kind relatie. Er werden positieve verbanden gevonden tussen de oriëntatiedimensie en perspectief nemen beoordeeld door de
gespreksleider en tussen de verandering dimensie en het toeschrijven van leerlinggedrag aan het eigen handelen als leerkracht. Negatieve verbanden werden gevonden tussen de
verandering dimensie en het toeschrijven van leerlinggedrag aan het karakter van de leerling. De resultaten ondersteunden met name de validiteit van de dimensies oriëntatie en
verandering. Voor de validiteit van de dimensie uitwerking werd minder evidentie gevonden. Keywords: Leerkrachtreflectie; Leerkracht-kind relatie; Reflectiecoderingsinstrument;
The Development and First Test of a Reflection Coding Instrument Designed to Measure Teacher Reflection During Teacher-Child Interaction Coaching.
Relationships between teachers and disruptive children are often problematic and characterized by conflict. They elicit feelings of helplessness and anger in teachers (Spilt & Koomen, 2009). Moreover, teacher-child relationships with high levels of conflict hamper the development of the child whereas warm and open relationships foster children’s development (Pianta, Steinberg, & Rollins, 1995; Pianta & Stuhlman, 2004; Spilt & Koomen, 2009). The Relationship Focused Reflection Program (RFRP) is an intervention aimed at improving dyadic, one-on-one, relationships between teachers and disruptive students. It is assumed that to achieve this goal teachers must reflect on their relationship with individual students (Spilt, Koomen, Thijs & van der Leij, 2012). However, to what extent do teachers show reflection during the RFRP and does reflection foster improvement of the teacher-child relationship? These questions, especially the first one, are the focus of the current study. More specifically, a Reflection Coding Instrument was developed to quantitatively measure the nature and degree of teacher reflection during the RFRP sessions. The Reflection Coding Instrument was subsequently tested for overall usefulness and psychometric properties. Lastly, the treatment integrity of the RFRP was explored, since a low treatment integrity could have influenced the scores on the Reflection Coding Instrument.
Attachment Theory and The Developmental Systems Theory
In recent years, relational processes between teachers and students have proliferated as a focus to support children’s development in the classroom (Sabol & Pianta, 2012). Robert Pianta can be seen as the pioneer, since he explored the role of adults, beyond parents, in children’s lives (Pianta, 1992; Verschueren & Koomen, 2012). Multiple researchers have now empirically demonstrated that the quality of the teacher-child relationship influences the academic performance, the motivation and engagement in school, and the psychosocial functioning of the child (Howes, Hamilton, & Matheson, 1994; Pianta, Steinberg, & Rollins, 1995; Pianta & Stuhlman, 2004). Most theoretical and empirical work about relational processes between teachers and students has been strongly influenced by the attachment theory and the developmental systems theory (Sabol & Pianta, 2012). Both theories will be briefly described in this paragraph.
The idea that the relationship between a student and a teacher can play an important role in the development of the student, is largely derived from the notion that the attachment between a child and his/her parent(s) strongly influences the development of the child (Ainsworth, Blehar, Waters, & Wall, 2015; Sabol & Pianta, 2012). According to the
attachment theory, an attachment figure can serve as a secure base for a child from which the child can explore the world or as a safe haven to retreat when necessary (Waters, Crowell, Elliott, Corcoran & Treboux, 2002). Verschueren and Koomen (2012) state that empirical research has revealed several similarities between the parent-child relationship and the teacher-child relationship. They argue that a teacher can serve as a temporary attachment figure, since teachers can function as a secure base and safe haven (Verschueren & Koomen, 2012; Zajac & Kobak, 2006). However, this does not mean that (all) children have a full attachment bond with their teachers (Verschueren & Koomen, 2012). The relationship between a teacher and a child is in most cases not an exclusive and durable attachment bond. Children have different teachers over time and must share their teacher with other children in the classroom. Also, the primary role of teachers is, even though they engage in caregiving behaviours, instruction (Hamilton & Howes, 1992; Verschueren & Koomen, 2012).
Verschueren and Koomen (2012) conclude that the role of a teacher as an attachment figure is expected to be of greater importance for younger children and/or more vulnerable children.
Research on teacher-child relationships is also strongly being influenced by the developmental systems theory (DST). DST states that people are embedded in dynamic systems that include proximal and distal levels of influence (Sabol & Pianta, 2012). The interactions between a child and his/her teacher is an example of a proximal process. Other examples of systems are families, child-parent relationships, classrooms and peer groups (Pianta, Hamre & Stuhlman, 2003). According to DST, systems and their components are embedded within other systems. Interactions take place across systems and within systems over time. These interactions are reciprocal and bidirectional. DST can help to integrate the various factors that influence the development of young children (Pianta et al., 2003).
The attachment theory and the DST provide the basis for understanding that children and teachers form mental representations, based on early experiences, of attachment and of their relationship (Pianta et al., 2003; Sabol & Pianta, 2012; Waters & Waters, 2006). A mental representation, also called a representational model or working model, is defined as a set of beliefs and feelings that has been internalised about a relationship (Pianta et al., 2003; Waters & Waters, 2006). Mental representations include feelings, thoughts and expectations about the self, the other and the relationship between the self and the other (Piante et al., 2003). Empirical evidence shows, for instance, that mental representations of negative affect are especially notable in teachers’ relationships with disruptive children (Spilt & Koomen, 2009; Spilt et al., 2012). DST shows us that a teachers’ mental representation of the
specific student, but is also influenced by other systems such as the teachers’ experiences with other students, his/her own past experiences as being a student him-/herself, his/her own experiences with other teachers, etcetera (Pianta et al., 2003). Mental representations are believed to guide behaviour largely outside conscious awareness (Spilt et al., 2012). They influence a person’s memory and response availability in social interactions (Schank & Abelson, 1977; Waters & Waters, 2006). The mental representation of a teacher that consists of negative affect, for instance, influences that teachers’ behaviour towards individual children disadvantageously (Spilt et al., 2012; Stuhlman & Pianta, 2002). Mental representations can reinforce themselves over time since people are inclined to focus on information that is consistent with existing beliefs. Therefore, mental representations can become highly stable and constraining on the teacher-child relationship (Pianta, 1999; Spilt et al., 2012). Although fairly stable, mental representations are considered to be open systems, since they can be changed based on new experiences (Pianta et al., 2003).
In sum, the attachment theory and DST underscore the importance of teacher–child relationships for the development of the child. Both theories also point to the value of developing interventions that are focused on improving the teacher-child relationship to prevent children from developing problem behaviour (Pianta & Hamre, 2001; Vancraeyveldt, Verschueren, Van Craeyevelt, Wouters, & Colpin, 2015).
Interventions for the Teacher-Child Relationship
Several interventions that are based on the attachment theory and/or DST are intended to improve the dyadic, one-on-one, relationship between the teacher and the child in order to positively influence the development of the child (Sabol & Pianta, 2012). Banking Time, Playing-2-gether, and the Teacher-Child Interaction Therapy will be discussed in this paragraph (Driscoll & Pianta, 2010; McIntosh, Rizza, & Bliss, 2000; Vancraeyveldt et al., 2015).
Banking Time contains a set of techniques to build a positive and supportive
relationship between the teacher and the child. During a couple of sessions, the teacher and the child engage in one-on-one activities chosen by the child. The teacher is instructed to listen and to carry out acceptance and understanding during each session. The teacher also has to observe the activities of the child, verbally describe the activities of the student and label the feelings and emotions (Driscoll & Pianta, 2010). Results of a study with an experimental design, showed that teachers reported an increase in experienced closeness with pre-school children in the Banking Time condition in comparison to a no treatment control condition. Additionally, children who participated in Banking Time demonstrated gains in
teacher-reported frustration tolerance, competence and task orientation and decreases in conduct problems relative to their peers in the within-class control condition (Driscoll & Pianta, 2010; Sabol & Pianta, 2012). However, according to observers’ ratings, there were no differences between children in the Banking Time condition and the within-class control condition concerning the displayed positive affect, since children from both conditions showed significant increases from pre-test to post-test (Driscoll & Pianta, 2010).
Playing-2-gether was designed to decrease the externalizing problem behaviour of pre-schooler boys by focussing on the teacher’s interaction with the child. During the intervention of two six-weeks parts, the teacher organises play sessions with the target child on a regular basis. The first part of playing-2-gether is almost completely similar to the procedure of Banking Time and is called Relationship-Game. The second six-week part of the intervention is called Rule-Game. In this part of the intervention, the sessions are led by the teacher whom is taught to use skills to reduce the externalizing problem behaviour and to increase
appropriate child behaviour. Examples of those skills are giving clear commands, introducing rules and praising the child following good behaviour. Results from a longitudinal randomised controlled trial show that, as a result of playing-2-gether, teachers reported a larger decrease in externalizing problem behaviour compared with the control group children (Vancraeyveldt et al., 2015).
The Teacher-Child Interaction Therapy (TCIT) strives to improve the quality of the teacher-child relationship by teaching the teachers behavioural therapy skills. The second goal is to provide teachers with training in problem-solving. This way, teachers can develop more adequate strategies for dealing with behaviour problems of their students. Like Banking Time and Playing-2-gether, one-on-one play sessions are an important element of the TCIT. As a result of a case study, TCIT was found to increase the number of positive interactions between the child and the teacher. TCIT was also found to decrease the child’s disruptive behaviours (McIntosh et al., 2000).
The above-mentioned interventions are all designed to change teachers’ behaviour in order to positively influence the relationship between a teacher and a disruptive child. The interventions indeed have shown positive results. However, there are some side notes.
Banking Time only showed an increase of the experienced closeness within the teacher-child relationship in comparison to the no treatment control group, but not in comparison to a within-class control condition. Playing-2-gether showed positive results concerning the development of the child, since the teachers reported a decrease in externalising problem behaviour. The results did not entail positive effects concerning aspects of the teacher-child
relationship. Lastly, TCIT showed positive results concerning the number of positive
interactions as rated by an observer. However, it did not contain teacher and/or child reports. Therefore it can be concluded that even though the above mentioned interventions have shown positive results, there is room for improvement. An intervention that does not focus directly on the teachers’ behaviour but, instead, has another focus, will possibly have more success in improving the teacher-child relationship.
The Relationship-Focused Reflection Program
Since mental representations of negative affect are especially notable in teachers’ relationships with disruptive children (Spilt & Koomen, 2009; Spilt et al., 2012) and these mental representations affect teachers’ behaviour towards individual children
disadvantageously, Pianta stated (1999) that interventions should aim at constructing more flexible and differentiated mental representations of the relationship with a student through a reflective process. Therefore, the Relationship-Focused Reflection Program (RFRP) was designed to indirectly change the behaviour of the teacher by positively influencing teachers’ mental representations of the relationship with the child which can have a positive influence on the teacher-child relationship (Spilt et al., 2012).
The RFRP uses reflection as a tool to stimulate teachers to adjust their mental
representation of the relationship with the child, since a recent finding is that the caregiver’s capacity for reflection fosters a secure attachment (Slade, Grienenberger, Bernbach, Levy, & Locker, 2005). The authors of the RFRP define Reflective functioning (RF) as “the capacity to think about one’s own and the other’s behavior in terms of underlying mental states such as feelings and intentions” (Fonagy, Steele, Steele, Moran, & Higgitt, 1991, 203; Spilt et al., 2012). With the RFRP, the authors aim to foster an open mind-set so that teachers become aware of the positive feelings that they experience in relation to the child instead of only being conscious of the negative feelings. As a consequence, teachers can adjust their mental representation of the child and the relationship with that child. An open mind-set also means that teachers not only look at the child when exploring causes of problems, but also look at their own functioning as a teacher. Lastly, the RFRP aims to challenge the teacher to explore the way in which his/her emotions influence his/her attitudes and behaviours towards the child. This can lay a foundation for actual change in behaviour (Koomen & Spilt, 2015).
The RFRP consists of four one-on-one sessions with a teacher and a consultant. The first two sessions concern one disruptive child from the classroom of the teacher. The last two sessions concern a second disruptive child. Every first session per child is devoted to the teacher’s narratives about the relationship, elicited by the semi-structured Teacher
Relationship Interview (TRI; Pianta, 1999). Every second session per child, the consultant presents a relationship profile with strengths and weaknesses depicted in a bar graph. The relationship profile is used as a starting point for in depth reflection (Spilt et al., 2012). The relationship profile is based on the TRI coding manual and consists of the following
constructs concerning pedagogical practices and feelings: “Sensitivity of discipline”, “Secure base”, “Perspective taking”, “Beliefs about efficacy”, “Feelings of helplessness”, “Negative affect”, “Positive affect”, and “Neutralising negative affect”. It is assumed that these
constructs give insight into the mental representation of the teacher-child relationship (See Appendix Spilt & Koomen, 2009; Koomen & Spilt, 2015; Spilt et al., 2012).
Spilt, Koomen, Thijs and van der Leij (2012) have examined the efficacy of the RFRP. For their research, they included thirty-two teachers and per teacher two kindergartners with above-median levels of externalizing behaviour. The research contained a comparison group who received an alternative intervention called the Interpersonal Skills Training (IST). It can be concluded from teacher reports that there are differences, between the different teacher-child dyads, in the way in which the degree of experienced closeness changes. Fifteen dyads were already characterized by a high degree of experienced closeness. The degree of
experienced closeness did not change for those dyads. Seventeen dyads were characterized by a low degree of experienced closeness. For six of those dyads the degree of experienced closeness decreased as a result of the RFRP. For eleven dyads the degree of experienced closeness increased. The authors also found a moderation effect with teacher efficacy.
Teacher reports showed that teachers with a high feeling of teacher efficacy reported a bigger decrease in conflict. Lastly, observations showed that the degree of teachers’ sensitivity towards the child increased. All these changes were relative to the comparison group who received the IST (Spilt et al., 2012).
It can be concluded that, as a result of the RFRP, teachers’ mental representations of the teacher-child relationship may have been changed, because some teachers reported a decrease and others reported an increase in the experienced closeness with the student. As already mentioned, it is assumed that a teacher has to reflect on his/her relationship with the student during the sessions of the RFRP in order to change their mental representation of the relationship. Consequently, this could positively influence the teacher-child relationship (Spilt et al., 2012). Spilt, Koomen, Thijs and van der Leij did not examine, however, if the results of the RFRP could be related to the nature and degree of teacher reflection. Before this issue is going to be further explored, we first need to know more about the concept reflection. A History of Reflection and Reflection Research
Dewey is considered to be the creator of the concept of reflection. In his eyes,
reflection is a form of problem solving. When you are reflecting, you are thinking in order to resolve an issue which involves the ordering of ideas and linking each idea with its
predecessors. (Dewey, 1993; Hatton & Smith, 1995). In this light, reflection can be seen as a deliberate and active cognitive process in which doubt and perplexity are generally
experienced before possible solutions are reached (Hatton & Smith, 1995). Schön is also an important person in the history of the concept of reflection (1983, 1987). He made the important distinction between “reflection-on-action” and “reflection-in-action”. Reflection-on-action refers to reflection upon action that has happened in the past. The RFRP asks for this kind of reflection, since the teacher-child relationship is being discussed during the sessions with the consultant. Reflection-in-action refers to reflection that occurs
simultaneously with the action itself. The latter implies that people are able to consciously think about what is happening and are able to modify their actions accordingly (Hatton & Smith, 1995). Schön stated that reflection is bound up with practice. With this statement, he meant that professionals, such as teachers, understand and act on their situations in ways that cannot be reduced to rules. This statement has led to the understanding that reflection is an important process for teachers (Ward & McCotter, 2004).
Van Manen (1977) was one of the first authors who aspired to find a hierarchy in reflection. He derived three levels of reflection based on theories of Habermas: “Technical reflection”, “practical reflection” and “critical reflection”. Technical reflection, the first level, refers to the efficiency and the effectiveness of means to achieve certain ends. Practical reflection, the second level, refers to the open examination of means, goals, the assumptions on which these are based and the actual outcomes. There is acknowledgment that meanings are embedded in and negotiated through language and that meanings are not absolute. Critical reflection, the third level, refers to the considerations involving moral and ethical criteria. It locates analysis of personal actions within wider socio-historical and politico-cultural contexts (Hatton & Smith, 1995).
Hatton and Smith combined the work of Dewey, Schön and van Manen to be able to identify types and patterns of reflection shown by teacher education. They define reflection as “deliberate thinking about action with a view to its improvement” (Hatton & Smith, 1995, page 40). Based on their research, Hatton and Smith (1995) identified four types of
(reflective) writing: “Descriptive writing”, “Descriptive reflection”, “Dialogic reflection” and “Critical reflection”. These types are, except for the first type, characterized as different kinds of reflection-on-action. Descriptive writing is sometimes referred to as technical or routine
reflection and reports events or literature (Hatton & Smith, 1995; Ward & McCotter, 2004). The second type, descriptive reflection, tries to provide reasons, which is often based on a person’s own judgement or on the reading of literature. The third one, dialogic reflection refers to a discourse with one’s self during which possible reasons are explored. Lastly, critical reflection is defined similar to Schön’s critical reflection. It involves reason giving for decisions or events together with considering the broader historical, social, and/or political contexts (Hatton & Smith, 1995).
More recently, Ward and McCotter (2004) have enlarged and adjusted Hatton and Smiths’ types of reflection. Their most important point of critique on most hierarchical
frameworks of reflection, including that of Hatton and Smith, states that those frameworks are solely designed to describe the process of reflection and not to identify the qualities of
reflection that are related to improvement of practice (Ward & McCotter, 2004). Therefore, Ward and McCotter decided to develop a rubric for evaluating teacher reflection. They used a grounded theory approach to develop this rubric. After a cyclical process of examination of theory and data, Ward and McCotter ended up with three dimensions of reflection; “Focus”, “Inquiry” and “Change”. They combined these with the hierarchical types of reflection defined by Hatton and Smith. In this way, the three dimensions of reflection can be seen as aspects of the different types of reflection. The first dimension, Focus, asks the question “What is the focus of concerns about practice?”. This dimension ranges from focus on self, focus on students to focus on the impact of broader concerns on student learning. The Inquiry dimension addresses the question “What is the process of inquiry?”. This dimension concerns how questions are asked rather than whether they are present. The Change dimension
concerns the question “how does inquiry change practice and perspective?”. This dimension includes new insights that arise from a “synthesis of situated experiences” (Ward & McCotter, 2004).
After a careful examination of the existing literature on levels of reflection, it has become clear that the existing studies about reflection mainly focus on reflection in journals written by teachers (Hatton & Smith, 1994; Surbeck, Han & Moyer, 1991; Ward & McCotter, 2004; Wong, Chung & CertEd, 1995). However, the current research focuses on reflection in spoken conversation during the RFRP. Therefore, a Reflection Coding Instrument needed to be developed that was usable to measure the nature and degree of reflection that teachers show during the RFRP.
It is assumed that the RFRP positively influences the dyadic relationship between teachers and disruptive students through reflection of teachers on their relationship with the students during the sessions of the RFRP (Spilt et al., 2012). The question arises, however, if teachers indeed show reflection during the RFRP and, if so, if reflection fosters improvement of the teacher-child relationship. In order to be able to explore the first question, an instrument needed to be developed to quantitatively measure the nature and degree of teacher reflection during the RFRP sessions. Therefore, the current research concerns an explorative study in which a Reflection Coding Instrument was designed and tested in order to judge the overall usefulness and psychometric properties.
A grounded theory approach was used to design the Reflection Coding Instrument. The reflection rubric of Ward and McCotter (2004) was used as a starting point. Audiotapes from a pilot sample were listened to and transcribed to inductively construct the Reflection Coding Instrument (Strauss & Corbin, 1990). In order to judge the overall usefulness and psychometric properties of the Reflection Coding Instrument, audiotapes of every second session of the RFRP from a different sample were listened to, transcribed and scored by using the Reflection Coding Instrument. The results on the Reflection coding Instrument gave information about the dispersion of the scores within the different dimensions of reflection, the relevance of the different dimensions of reflection for the measurement of teacher reflection that is shown during sessions of the RFRP and the internal consistency of the different dimensions of reflection. Subsequently, the results of the Reflection Coding Instrument were associated with different variables to gain insight into aspects of validity.
First, reflection scores were related to the degree of perspective taking by the teacher. Reflecting, according to Dewey, is thinking in order to resolve an issue which involves the ordering of ideas and linking each idea with its predecessors (Dewey, 1993; Hatton & Smith, 1995). In this light, reflecting upon another’s thoughts, feelings, and internal mental states is necessary when considering another’s perspective (Epley & Caruso, 2008). More specifically, perspective taking theoretically has a large overlap with the Focus dimension of reflection defined by Ward and McCotter (2004), since the Focus dimension mainly concerns the degree in which a teacher is capable of taking the perspective, the focus, of the student. It is expected, then, that the more a teacher is able to take the perspective of his/her student, the more he/she is also able to reflect. In particular, it is expected that there is a positive relation between perspective taking and the Focus dimension of reflection.
Second, reflection scores were related to teacher attributions for student behaviour. When a teacher is considering the possible causes of student behaviour, he/she is, again in
light of Dewey’s definition, reflecting (Cothran, Kulinna & Garrahy, 2009). Two attributions are examined, namely the attribution of student behaviour to (1) the character of the student and (2) the teacher his/her own way of teaching. It is expected that the more a teacher
attributes the behaviour of the student to his/her own way of teaching, the more he/she is able to reflect. More specifically, it is expected that there is a positive relation with the Focus dimension of reflection, since this dimension also concerns the impact of the teacher him/herself on student learning. In contrast, it is expected that there is a negative relation between the attribution of student behaviour to the character of the student and teacher
reflection. When a teacher is attributing student behaviour to the character of the student, then he/she is not considering the impact of him/herself on the behaviour of the student and on student learning. Therefore, it is especially expected that there is a negative relation with the Focus dimension of teacher reflection.
Lastly, reflection scores were related to the degree in which teachers experience change in the relationships with the student after participating in the RFRP. As a result of reflection, the mental representation of a relationship can change (Pianta et al., 2003) which can have a positive influence on the teacher-child relationship (Spilt et al., 2012). Therefore, it is expected that the more a teacher reflects about his/her relationship with a student, the more he/she is able to positively change his/her mental representation of that relationship and experience actual positive change in the teacher-child relationship. Additionally, it is expected that the more a teacher reflects about his/her relationship with a student, the more he/she is able to positively change his/her mental representation of that relationship and the less he/she experiences actual negative change in the teacher-child relationship. These relations are especially expected for the change dimension of reflection, since this dimension concerns the degree in which a teacher shows inclination to change.
In addition to the question of validity, this research also contains a qualitative part. Every second session of the RFRP was also listened to in order to explore the treatment integrity, since a shortcomings concerning the treatment integrity could have influenced the scores on the Reflection Coding Instrument and with that the outcomes on the above-mentioned hypotheses.
Method Participants and Procedure
For the current study, participants and data from a larger research project on the efficacy of the RFRP intervention were used. This project was previously approved by the
Ethics Review Board of the Faculty of Social and Behavioral Sciences, University of
Amsterdam (project 2015-CDE-4482). From the participants of this research project, twenty-four dyadic relationships - twelve teachers and twenty-twenty-four students - were included in the current study. Five male and seven female teachers from five regular elementary schools (third to sixth grade) in the Netherlands participated. The students in the sample were selected by the teachers based on experienced problems within the student-teacher relationship.
Eighteen of the students were male and six of the students were female.
The RFRP took place in October and November of 2015. The teachers were asked to fill in a questionnaire at three occasions: pre-intervention, post-intervention and follow-up. For the purpose of the current study, questions from the pre-intervention questionnaire concerning teachers’ attributions of student behaviour and questions from the
post-intervention concerning teachers’ experience of positive and negative change of the student-teacher relationship were used. Apart from the questionnaire data, the current study also included data from the larger research project concerning consultants’ ratings of teachers’ perspective taking. These ratings were elicited from the relationship profiles that the consultants presented during every second session of the RFRP.
Development of the Reflection Coding Instrument
A cyclical process of examination of theory and data gave rise to the Reflection Coding Instrument. The first version of the Reflection Coding Instrument (Appendix 1 and 2) consisted of the dimensions Focus, Inquiry and Change from the Reflection Rubric of Ward and McCotter (2004) and, additionally, consisted of hierarchical levels of reflection per dimension. These hierarchical levels were only slightly adjusted to be used for the RFRP. A pilot set of audiotapes of the six dyadic relationships, three teachers and six students, were used to test the first version of the Reflection Coding Instrument. Two types of statements made by teachers were rated: (1) Every answer of the teacher concerning the selected students and/or the relationship with the selected students, following a question of the consultant (2) Every statement of the teacher concerning the selected students and/or the relationship with the selected students, which he/she gave as a reaction to a statement made by the consultant. The test of the first version of the Reflection Coding Instrument revealed that the Inquiry dimension was not relevant for the RFRP. It turned out that, since the consultant asks most of the questions during the RFRP, the teacher him-/herself in most cases did not ask questions to him-/herself. Therefore, the Inquiry dimension was deleted from the Reflection Coding Instrument. Instead, it was judged valuable to include another aspect that is Elaboration (Surbeck, Han, & Moyer 1991). Elaboration concerns the way in which the first reaction is
expanded by an explanation, an example and/or reference to another situation. This, together with some other small changes, resulted in the second version of the Reflection Coding Instrument (Appendix 3 and 4).
The second version of the Reflection Coding Instrument was again tested on the pilot set of audiotapes. This time, also a second evaluator tested the Reflection Coding Instrument. As a result, the descriptions of some of the levels changed. Also, some levels were left out and others were added. For instance, one of the levels of the Focus dimension of the second version of the Reflection Coding Instrument concerned both the teacher’s focus on the student’s perspective and the teacher’s own influence on the student. The test showed that teachers also make statements in which the teacher takes the perspective of the student, but does not talk about his/her own influence on the student. Therefore, an extra level for this kind of statements was added. The changes resulted in the third, and final, version of the Reflection Coding Instrument (Appendix 5 and 6).
Intervention Script: the Second Session of the RFRP
In every second session of the RFRP, the consultant and the teacher look back at the TRI that was administered in the first session in order for the teacher to reflect upon his/her relationship with the student in question. With the use of six steps, the consultant is supported to give the teacher information and asks him/her questions with the aim to stimulate the teacher to explore the relationship with the student from different perspectives (Koomen & Spilt, 2015). These steps will be briefly described below to provide an image of the sessions that were rated for the nature and degree of teacher reflection.
In the first step, the consultant and the teacher discuss what has happened with the teacher and the student since the first session. During the second step, the consultant explains that he/she has interpreted the TRI and that this will be discussed during the second session. He/she emphasises that it is merely the interpretation of the consultant about the student-teacher relationship obtained by a qualitative scoring system and that it is solely based on the TRI. The teacher is encouraged to speak up when he/she has a different view on the matter. During the third step, the consultant globally explains the relationship profile. The consultant indicates, at the beginning of step four, that he/she is going to give a lot of information and that the teacher, therefore, is encouraged to make notes. Then the consultant starts to explain every construct. For every construct, the consultant first gives a short description. Then he/she informs the teacher if he/she has received a low, an average or a high score on the construct. The consultant does not communicate the exact scores in terms of numbers. To illustrate the received score, the consultant gives one or two examples from the TRI. Then, the consultant
explains why he/she has not given a higher or a lower score. During step four, the consultant gives the teacher regularly space to react. In other words, the consultant holds up a mirror and gives the teacher the opportunity to ask questions and to respond. During step five, the teacher is asked to what extend he/she recognizes him/herself in the relationship profile. The final step, step six, is to provide the teacher with a summary of what has been discussed during the session. The consultant also asks if the teacher sees linkages between the different constructs of the relationship profile and if he/she has gained anything from the profile. Lastly, when it is the last session for the teacher (the second session concerning the second student), the two relationship profiles concerning the relationship with the two students are placed next to each other. The teacher is then asked if he/she recognizes similarities and differences and if he/she can think of an explanation for those similarities and differences (Koomen & Spilt, 2015). Measures and Missing Data
Nature and degree of teacher reflection. The Reflection Coding Instrument consists of three scales for the dimensions Focus, Change and Elaborations. The Focus scale concerns who or what the teacher holds responsible for a situation or a problem. It also takes into account if a teacher takes the perception of the student and considers his/her own influence on the student. The scale has 5 levels that range hierarchically from “The teacher barely gives any reaction. Therefore, it seems as if the teacher has no focus at all” (1) to “The teacher focuses on the student and the influence he/she has on the student and, additionally, on all the students in the classroom and/or the people outside the classroom” (5).The Change scale concerns the degree to which the teacher shows inclination to change. The scale has 4 levels that hierarchically range from “There is no personal reaction. It is as if there is a distance between the teacher and the situation” (1) to “The teacher uses the situation to develop new insights about his/her own way of acting as a teacher and about his/her strengths and weaknesses” (4). Lastly, the Elaboration scale concerns the degree in which the teacher elaborates his first reaction (meaning, feeling, thought, etc.) with arguments or considerations. The scale has 4 levels ranging hierarchically from “The teacher does not elaborate on his first reaction” (1) to “The teacher elaborates upon his first reaction by considering different perspectives and, in the end, develops a general principle and/or theory” (4) (See Appendix 1 to 6). Unfortunately, for one teacher a part of the audiotape was missing. Therefore, not the whole second session could be scored for teacher reflection.
Teacher’s perspective taking. The consultants rated the TRI narratives on eight relationship constructs by using a qualitative scoring system from the TRI coding manual in order to form a relationship profile (See Appendix Spilt & Koomen, 2009; Koomen & Spilt,
2015; Spilt et al., 2012). Data on the Perspective taking construct for each individual student of the relationship profile was used for the current research. Perspective taking refers to the extent to which the teacher is aware of the internal states of the student in question and that he/she provides possible reasons for that state (See Appendix Spilt & Koomen, 2009). The consultants could range the Perspective taking construct on a 7-point scale ranging from low (1-2), middle (3-4-5) to high (6-7). A high score means that the construct is clearly present within the student-teacher relationship (Koomen & Spilt, 2015)
Attributions of student behaviour. For the current research, the results on two questions about teacher attributions were used. The questions are as follows: “In which degree would you attribute the behaviour of the student to the nature/character of the
student?” and “In which degree would you attribute the behaviour of the student to your own acting/ways of teaching as a teacher?”. The teachers could rate each question on a five point Likert scale ranging from “definitely not the cause” (1) to “definitely the cause” (5). For the purpose of this research, the answers that the teachers had given pre-intervention were used. In this way, their answers couldn’t be influenced by the RFRP. For four dyads, the answers on these questions about attributions were missing.
Teacher’s perception of relationship change. Two questions about perception of relationship change were used, originating from the study of Pianta and Nimetz (1991): “My relationship with this student has become more positive over the last school period” and “My relationship with this student has become more negative over the last school period”. Again, a five point Likert scale was used for both questions, with answers ranging from “totally disagree” (1) to “totally agree” (5). For these questions, the answers were used that the teachers gave post-intervention. For five dyads, the answers on these questions were missing. Data Analysis
Since this study concerns an explorative research in which a Reflection Coding Instrument was designed and tested in order to judge the overall usefulness and psychometric properties, descriptive information about the reflection scales and the corresponding
hierarchical levels will be provided in the paragraph below. This information gives insight into the quality and relevance of the different reflection scales and the corresponding hierarchical levels. The number of scores per level of the reflection scales, the lowest and highest scores on the reflection scales, and the mean and the standard deviations are shown to examine the dispersion and to examine the relevance of every level of the three reflection scales for the RFRP. It was also examined if the three reflection scales form an internally consistent scale.
In order to examine aspects of the validity of the Reflection Coding instrument, in other words whether the scores on the reflection scales related to teachers’ perspective taking, teachers’ attribution of student behaviour, and the experienced change in the teacher-child relationship, Kendall’s tau-b coefficient was used. This statistical measurement has been chosen, since the data set was small (Field, 2009). Considering the sample size, the
significance level was set at .10 to reduce the chance of Type-II error. The Kendall’s tau-b coefficient was tested two-tailed. The results on the Kendall’s tau-b coefficient were also used to examine whether the mean scores on the reflection scales or the highest scores on the reflection scales are the best indicator to measure teacher reflection. Therefore, both the mean scores and the highest scores were used as reflection measures in the analyses.
Lastly, the audiotapes of every second session of the RFRP were listened to and transcribed in order to qualitatively measure the treatment integrity. This information is important, since the usefulness and psychometric properties of the Reflection Coding Instrument, among which aspects of the validity, might be influenced by shortcomings concerning the treatment integrity.
Results Descriptive Information on Reflection Scales
Table 1 presents the number of scores per level for the scales Focus, Change and Elaboration. The first thing that is notable is that level 5 of the reflection scale was not scored at all. This level indicates that a teacher takes the perspective of the student and considers his/her own influence on the student and additionally on all the students in his/her class and/or persons outside his/her class. It is also notable that for all three scales, level 2 was most
scored. This means that most teachers did not take the perspective of the student, did not consider his/her own influence on the student, did not use the situation to change his/her perspective, did not show inclination to change and did not elaborate on his/her first reaction by taking multiple perspectives/angles into account. For both the Change and Elaboration scale, level 3 was the second most scored. This indicates that teachers’ responded personally on the discussed situation, showed inclination to change and elaborated on their first reaction, opinion, feeling, and etcetera, from different angles of incidence. For the Focus scale,
however, level 4 was second most scored. This indicates that teachers took the perspective of the student and considered their own influence on the student.
Table 1
The number of scores per level for the scales Focus, Change and Elaboration
Focus Change Elaboration
Level 1 27 24 26
Level 2 403 462 312
Level 3 58 70 238
Level 4 95 27 7
Level 5 0 - -
Note. Level 5 does not apply for the scales Change and Elaboration.
Tables 2 and 3 present the means, standard deviations, minimum scores and maximum scores for the mean scores on the three reflection scales and the highest scores on the three reflection scales. The standard deviations show that the dispersion within the three reflection scales is quite small. The standard deviations range from 0.26 to 0.34, with one outlier of 0.82 for the mean scores on the Change scale. It is notable that, concerning the mean scores on the three reflection scales (Table 2), the Focus scale has a slightly higher maximum score in comparison to the other two scales. This cannot be due to the fact that the Focus scale
contains five levels in comparison to the other two scales which contain four levels, since the fifth level of the Focus scale has not been scored at all. The slightly higher maximum score can be explained by the fact that the number of level 4 scores for the Focus scale is
considerably higher than the number of level 4 scores for the Change and Elaboration scales.
Table 2
Mean, standard deviation, minimum and maximum score of the mean scores on the scales “Focus”, “Change” and “Elaboration”
Focus mean scores Change mean scores Elaboration mean scores
Mean 2.39 2.17 2.40
Std. Deviation 0.33 0.26 0.26
Minimum 1.88 1.80 1.88
Maximum 3.33 2.83 2.83
Table 3
Mean, standard deviation, minimum and maximum score of the highest scores on the scales Focus, Change and Elaboration
Focus highest scores Change highest scores Elaboration highest scores
Mean 3.88 3.33 3.13
Std. Deviation 0.34 0.82 0.34
Minimum 3.0 2.0 3.0
Maximum 4.0 4.0 4.0
As shown in Table 4, the mean scores on the three scales correlate significantly. The mean scores on the Focus scale correlate significantly with the mean scores on the Change
scale, r = .44 (p = .003) and those on the Elaboration scale, r = .50 (p = .001). The mean scores on the Change scale and the mean scores on the Elaboration scale also correlate significantly, r = .40 (p = .007). The highest scores on the three scales, however, do not correlate significantly (p >.10). The mean scores on the three scales form a reliable scale (Cronbach’s α: .83) which will be called the “Total reflection scale”. However, the highest scores on those scales do not form a reliable scale (Cronbach’s α: .37).
Table 4
Kendall’s tau-b Correlations between the mean and highest scores on the scales Focus, Change and Elaboration
Focus scores Change scores Elaboration scores
Focus scores .15 .14
Change scores .44 ** .32
Elaboration scores .50 ** .40 **
Note. The correlation coefficients between the highest scores are above the diagonal line and the correlations between the mean scores are below the diagonal line.
** p < .01
Relation Between Teacher Reflection and Teachers’ Perspective Taking
Reflection scores were related to the degree of perspective taking by the teacher. The results in general show, as expected, positive correlations between the scores on the reflection scales and the scores on the Perspective taking scale of the TRI (Table 5). The mean scores on the scales Focus and Elaboration and the Total reflection scale correlated significantly with the scores on Perspective taking , respectively r = .38 (p = .029), r = .37 (p = .034) and r = .43 (p = .013). The highest scores on the Focus scale also correlate significantly with the
Perspective taking scale, r = .50 (p = .016).
Table 5
Kendall’s tau-b Correlations between the scores on the Reflection Coding Instrument and the Perspective taking construct of the TRI
Perspective taking
Focus mean scores .38 *
Change mean scores .29
Elaboration mean scores .37 *
Total reflection scale .43 *
Focus highest scores .50 *
Change highest scores .06
Elaboration highest scores .14
Relation Between Teacher Reflection and Teachers’ Attribution of Student Behaviour The reflection scores were also related to the teacher’s attribution of student behaviour to the character of the student and to the teacher’s way of teaching (Table 6). The results in general show, again as expected, positive correlations between the scores on the Reflection Coding Instrument and the scores on the attribution of student behaviour to the teacher’s way of teaching. The mean and highest scores on the Change scale correlated significantly with the scores on the attribution of student behaviour to the teacher his/her own way of teaching, respectively r = .36 (p = .048) and r = .36 (p = .081). In general, there were negative correlations between the scores on the Reflection Coding Instrument and the attribution of student behaviour to the character of the student. This corresponds with the assumed expectation. The mean scores on the Change scale correlated significantly with the answers on the question, r = -.42 (p = .029). The Total reflection scale also correlated significantly with the attribution of student behaviour to the character of the student, r = -.33 (p = .088). In contrast to what was expected, there were no significant correlations between the teacher attribution of student behaviour and the mean and highest scores of the focus scale.
Table 6
Kendall’s tau-b Correlations between the scores on the Reflection Coding Instrument and the teacher’s attribution of student behaviour to the student and to the teacher him/herself
Attribute behaviour to self Attribute behaviour to student
Focus mean scores .29 -.15
Change mean scores .36 * -.42 *
Elaboration mean scores .08 -.24
Total reflection scale .26 -.33 †
Focus highest scores .26 -.17
Change highest scores .36 † -.22
Elaboration highest scores .07 -.28
† p < .10 * p < .05
Relation Between Teacher Reflection and Change in the Teacher-Child Relationship
Lastly, reflection scores were related to the degree to which teachers experienced change in the relationships with the student after participating in the RFRP (Table 7). In general, against expectations, the results showed negative correlations between the scores on the Reflection Coding Instrument and the degree to which the teacher experienced a positive change in the relationship. There was only a positive correlation with the highest scores on the Change scale. However, none of the correlations were significant. Additionally, there were, in
general, positive correlations between the scores on the Reflection Coding Instrument and the degree in which the teacher experienced a negative change in the relationship. Again, this did not correspond with the assumed expectation. The mean scores on the Elaboration dimension and the Total reflection scale correlated significantly with the degree to which the teacher experiences a negative change in the relationship, respectively r = .41 (p = .032) and r = .39 (p = .039).
Table 7
Kendall’s tau-b Correlations between the scores on the Reflection Coding Instrument and the degree in which teachers experience a positive and a negative change in the relationships with the student after participating in the RFRP
Relationship with the student has become more positive
Relationship with the student has become more negative
Focus mean scores -.10 .30
Change mean scores -.03 .19
Elaboration mean scores -.22 .41 *
Total reflection scale -.15 .39 *
Focus highest scores -.23 .18
Change highest scores .18 .00
Elaboration highest scores -.20 .27
* p < .05
Examination of the Treatment Integrity
Since there were a lot of differences in the way the consultants acted during the RFRP which could have affected the treatment integrity, this paragraph contains an overview of these differences. The differences are clustered as follows: “Explanation of constructs”, “Explanation of received ratings”, “Stimulations for reflection” and “Use of interview techniques”.
Explanation of constructs. The consultants differed in the way they explained the meaning of the key constructs, or scales, of the relationship profile. This is an important aspect, since the consultants are instructed to explain the meaning of each construct in such a manner that the teachers get a clear image of the constructs (Koomen, Verschueren & Pianta, 2015). Sometimes a consultant did not explain the meaning of a construct. When they did explain the meaning of a construct, the consultants differed in the extensiveness of their explanation. The construct “sensitivity of discipline”, which is called “behaviour regulation” during the RFRP, is a good example. Some consultants were succinct in their explanation: Consultant: “Behaviour regulation is a theme that’s about clarifying and discussing rules and expectations in the classroom in a sensitive and proactive manner.”
Others explained the constructs more comprehensively:
Consultant: “The first aspect of the relationship is behaviour regulation. Behaviour regulation is the degree to which you act sensitive, so adapted to the student, but also
proactive concerning the rules and that you take measures so that the student can fully benefit from the learning opportunities.”
And some consultants expanded their explanations with an example:
Consultant: “I want to start with behaviour regulation. This dimension of the relationship is about the degree in which you clarify and discuss rules and expectations with the student in a sensitive and proactive manner so that the student can fully benefit from the learning
opportunities. It is really about whether you can adjust your behaviour to the student. Can you see in advance that the student is probably going to misbehave and can you then
preventively give little signs to the student so that you retain the student from misbehaving?”
Explanation of received ratings. After the meaning of a construct has been explained, the consultant is supposed to clarify the rating that the teacher received on the construct. The consultant is instructed do this by giving examples from the TRI that endorse the rating and by describing what teachers do who receive a higher and a lower rating
(Koomen & Spilt, 2015). However, not all the consultants always clarified the ratings that the teacher received on the constructs the way they were supposed to do. Sometimes consultants only informed the teacher about the rating he/she received on a construct without any
explanation. There were consultants who did not always discuss what teachers do who receive a higher and a lower rating on the construct. There were also consultants who discussed what they were supposed to discuss, but who every now and then did this in a different order. For example, they discussed what teachers do who receive a higher and a lower rating on the construct, but did this before they clarified which rating the teacher received on the construct. Additionally, when the consultant has clarified the rating the teacher received on the
construct, the consultant has to ask the teacher if, and why or why not, he/she agreed with the consultant (Koomen & Spilt, 2015). This question is important to ask, because it makes the teacher reflect on the way he/she sees him-/herself concerning the construct. Unfortunately, sometimes a consultant skipped this question.
Stimulations for reflection. Consultants differed in the questions that they asked to stimulate the teacher to reflect on their relationship with the student. For example:
Consultant: “What do you need specifically with X to realise the goals. Do you have any idea what you need so that he is going to listen to you?”
Consultant: “Do you have any idea why X behaves like that in these kinds of situations?”
With the first question, the consultant asked the teacher directly to reflect on possibilities to change the relationship. With the second question the consultant asked the teacher to take the perspective of the child. The answers that the teacher gave on these questions could be rated on the focus and change dimensions of the Reflection Coding Instrument. Because consultants asked these kinds of questions, teachers were provided with opportunities to reflect on their relationship with the student. However, there were also consultants who barely asked these kinds of questions. Consequently, teachers received fewer opportunities to reflect on their relationship with the student.
Apart from differences in questions, the consultants also differed in the degree of suggestive comments. Consultants were supposed to present the relationship profile and ask the teacher questions to stimulate reflection. Offering ideas or telling a teacher how to act is not a part of the RFRP (Koomen & Spilt, 2015). However, consultants were often
suggestively in their comments. For example, they told the teacher their opinion about the cause of a problem or they informed the teacher about what they should do. Here are two examples:
Consultant: “… at the start of the interview you had to give three words that you thought were typical for your relationship with X. You mentioned clearness and you told me that she
benefits from clearness and structure and that you want to make her understand that see is worth it and that you appreciate what she does. So, that are very good examples of you trying to provide safety for her. And also, mentioning what we (stops)
Teacher: “Yes”
Consultant: ”Just what you said that it is a challenge and it is fun to try to open her a little since she is pretty timid”
Teacher: “Yes”
Consultant: “Yes so the interview contains a lot of examples of that” Teacher: “Good”
Consultant: “Yes” Teacher: “Thank you”
Teacher: “Yes I think that way you try to manage” (stops) Consultant: “Hm hmm”
Teacher: “To continue”
Consultant: “Yes yes yes yes so that the child feels save enough to learn” Teach: “Yes yes”
Consultant: “Yes oké”
In this example, the consultant suggestively says that the way the teacher acts makes the child feel save enough to learn. It would have been better if the consultant would have asked the teacher how he/she thinks the way he/she acts influences the child. This way the teacher would have been stimulated to reflect on this matter instead of just listening to the opinion of the consultant.
Consultant: “…You say that he is reachable, but it is very thin and momentary” Teacher: “Yes”
Consultant: “He does not come to me that often and he does not use me as a confidential person so that is” (stops)
Teacher: “I know.”
Consultant: “He wants to solve it himself” (stops) Teacher: “Yes”
Consultant: “You said”
Teacher: “I do not know if he has those needs”
Consultant: “Yes yes and look if, for example, if you would support X with his feelings of impotence and shame”
Teacher: “Hm hmm”
Consultant: “Then you would get a higher score so that you teach him how he can deal with that”
Teacher: “Yes”
Consultant: “With his doubt. He is going towards puberty and his high intelligence that he feels differently sometimes and that you develop a greater understanding for that and he will get more thrusted with you and will reach out for you more”
In this example the teacher mentions that she doubts if the student has a need for safety. The consultant kind of ignores this statement and tells the teacher what she must do to provide safety and how this would influence the student. This does not suit the purpose of the RFRP,
which is not to tell teachers what to do, but to hold up a mirror, figuratively speaking, and to stimulate teachers to reflect on their relationship with the student (Koomen & Spilt, 2015). Because the consultant from this example tells the teacher what she should do, the teacher is not stimulated to reflect on the relationship with the student.
Use of interview techniques. The consultants also differed in their use of interview techniques. Firstly, most of the consultants asked questions in an uninterrupted sequence. For example:
Consultant: “Can you give some concrete examples? Because if you look at the profile in total, which dimensions stand out for you? What fits you the best? “
Because the consultant in this example asked so many questions, the teacher possibly did not know what the consultant wanted to know and, consequently, was not stimulated to reflect on the relationship with the student. Furthermore, there were consultants who barely asked why a teacher thinks or acts a certain way. That is unfortunate, since the “why” question encourages the teacher to reflect. Lastly, one consultant repeatedly asked “so, you agree with that?” instead of “do you agree with that?”. The way this consultant formulated the questions, presumably stimulated the teacher to answer the questions with a yes.
Discussion
The present study concerned an explorative study in which a Reflection Coding Instrument was designed and tested. With this Reflection Coding Instrument, the nature and degree of reflection that teachers show during every second session of the RFRP could be quantitatively measured. The study was not intended as a comprehensive examination of the question if teacher reflection fosters improvement of the teacher-child relationship. Rather, it was meant to design the Reflection Coding Instrument and to examine its overall usefulness and psychometric properties, among which aspects of the validity. Additionally, the treatment integrity of the RFRP was examined, since shortcomings concerning the treatment integrity could have influenced the scores on the Reflection Coding Instrument.
The descriptive information on the reflection scales showed, for one thing, that the highest level (5) of the Focus scale was not scored at all. This level indicates that a teacher takes the perspective of the student and considers his/her own influence on the student and additionally on all the students in his/her class and/or persons outside his/her class. An explanation could be that the RFRP does not lend itself for this kind of reflection, since the intervention encourages teachers to reflect on their relationship with specific students
(Koomen & Spilt, 2015). If this is the case, then level 5 should be removed from the
Reflection Coding Instrument. Additionally, it could be that the shortcomings concerning the treatment integrity, such as the suggestive comments that were made or the fact that some consultants barely asked why a teacher thinks or acts a certain way, could have prohibited teachers to reach level 5 of the focus dimension of reflection.
There appeared to be a notable difference in the distribution of the scores between the Focus scale and the Change and Elaboration scale. Level 4 of the Focus scale was relatively often (second most) scored in comparison to the Change and Elaboration scale in which level 3 was relatively often (second most) scored. It seems as though when a teacher takes the perspective of the student it more often coheres with taking his/her own influence on the student into account then it does not. In other words it might be that, for teachers, taking the perspective of the student and taking your own influence into account are highly connected. However, the differences in number of scores are not substantial enough to fully endorse this assumption. If this assumption is true, then it is necessary to take another look at the
description of the different levels of the Focus scale and to examine if they should be altered. The descriptive information also showed that the dispersion within the three reflection scales is quite small. This is due to the fact that for all the three scales level 2 was by far most often scored. This could mean that, for all three scales, the description of level 2 is too wide and should be narrowed down. The small dispersion, however, could also be related to
shortcomings concerning the treatment integrity. For instance, the degree to which the teacher has a clear image of a construct and the received score on that construct, may be considered to influence the degree to which the teacher can reflect on the relationship with the student concerning that construct. Therefore, it is important that the consultant explains the meaning of the construct and the received score in such a manner that the teacher can get that clear image. As the findings regarding the treatment integrity show, there were differences in the way the consultants explained the constructs and the received scores which could have influenced the degree of reflection that the teachers have shown during the RFRP. This also applies to the degree to which consultants used suggestive comments, the degree in which consultants asked questions that stimulate reflection, the degree in which consultants asked questions in an uninterrupted sequence, the degree in which consultants asked why a teacher thinks or acts a certain way, etcetera. Therefore, taken together, the shortcomings concerning the treatment integrity could have narrowed and influenced the results on the Reflection Coding Instrument.
