What need-supportive and need-thwarting teaching behaviors do university teachers use in their honors classes? An observational study

Teacher classroom behavior is an important factor in student learning and motivation. Past research within higher education has primarily concentrated on identifying teaching behaviors that teachers and students deem important in honors classrooms. Yet, what specific teaching behaviors either support or thwart the needs of students in real-world honors classrooms is currently not clear. This study, which utilizes video observation, sheds light on teaching behaviors that either support or thwart students ’ needs, as viewed through the lens of self-determination theory, within the context of Dutch honors education. We developed an observation tool to analyze video recordings of 12 lessons from four different teachers, and identified the types of behaviors making up the various dimensions of need-supportive and need-thwarting behaviors. We found nine types of behaviors that had not previously been identified in observational studies. Structure-providing behaviors were the most common need-supportive behavior, while need-thwarting behaviors always occurred alongside need-supportive behaviors. The observation tool introduced here can be used for further study of teaching behaviors in honors education practice. These results also make an important contribution to teachers ’ further professionalization and instructional practices.


Introduction
Teaching behavior in the classroom is critical for student learning (Schneider & Preckel, 2017).Teaching behavior involves instruction, connecting with students, and responding to their interests and needs (Muijs & Reynolds, 2017;OECD, 2020).Teachers can motivate or demotivate their students during the learning process through their teaching behavior.Although we know a great deal about the effectiveness of particular forms of instruction, such as giving lectures, small-group discussions, brainstorming, or asking questions (Brewer & Burgess, 2005), much less information is available about actual teaching behavior in higher education practice.Previous empirical studies looking at teaching behavior, primarily conducted in secondary schools (Aelterman et al., 2019;Hornstra et al., 2021;Stroet et al., 2013) and to a limited extent in higher education (Gucciardi et al., 2020;Vermote et al., 2020), have usually emphasized self-report methods.Observational studies are a vital addition to these studies because of their high ecological validity (Haerens et al., 2013;Van Doren et al., 2023).They provide the opportunity to capture information about real-life examples of behaviors in different combinations and stages in actual classroom situations, rather than perceived information (Haerens et al., 2013;Van Doren et al., 2023).In addition, these concrete observations of teacher behaviors can help train honors teachers in higher education for their role (Gaudin & Chalies, 2015;Haerens et al., 2013;Lemke, 2007;Van Doren et al., 2023).
Through their teaching behavior, teachers support or hinder the satisfaction of their students' basic psychological needs: autonomy, competence, and relatedness (e.g., Ryan & Deci, 2017).Autonomy is the feeling of control over one's own actions, experiences, thoughts, and behavior (e.g., Deci & Ryan, 2000;Haerens et al., 2013;Ryan & Deci, 2017).Competence is defined as feeling confident about one's abilities to achieve the desired goals (e.g., Haerens et al., 2013;Ryan & Deci, 2017), and relatedness is experiencing a sense of belonging and respect (e.g., Deci & Ryan, 2000;Haerens et al., 2013;Ryan & Deci, 2017).Meeting these basic needs is a necessary condition for intrinsic motivation (Deci & Ryan, 2000), which contributes to deep learning, continuing growth, well-being, and better performance (Deci & Ryan, 2000;Orsini et al., 2015).Teacher behavior plays an important role in the satisfaction of the students' basic needs in education.
Previous research has shown that teachers simultaneously exhibit various need-supportive and need-thwarting teaching behaviors in their lessons (Aelterman et al., 2019;Van den Berghe et al., 2013).The blend of teaching behaviors students encounter influences their motivation and performance (Deci & Ryan, 2000).To date, only a small number of observational studies on need-supportive and need-thwarting teaching behaviors have been conducted in secondary schools and secondary vocational institutions, and not in higher education (Cents-Boonstra et al., 2020;De Meyer et al., 2014;Haerens et al., 2013;Jang et al., 2010;Jiang et al., 2019;Reeve et al., 2004;Van den Berghe et al., 2013, 2016).
This study was conducted in a 4-year bachelor's degree program at a Dutch institute of higher professional education (University of Applied Sciences).The program in question is an honors program that selects students based on a motivation letter and an intake interview.The program aims to create an educational environment where teachers nurture students' intrinsic motivation through need-supportive teaching behaviors.This study aims to shed light on the need-supportive and need-thwarting teaching behaviors of teachers in honors education.Honors education is designed to educate students who are intrinsically motivated, inquisitive, have a profound desire for knowledge (Wolfensberger & Offringa, 2012), and are eager to tackle challenges (Scager et al., 2014).Students appreciate teachers more who are demanding, challenging, and inspiring (Wolfensberger & Offringa, 2012).Empirical studies on teaching behaviors in honors education are scarce (Scager et al., 2014;Wolfensberger, 2012).We are not aware of any observational studies in the classroom of honors education.Therefore, we developed an observation tool for analyzing videos of actual lessons in honors education.Our research question was as follows: What need-supportive and need-thwarting teaching behaviors do university teachers in honors education demonstrate during their classes?

Dimensions of need-supportive teaching behavior
Teachers who exhibit need-supportive teaching behaviors support their students' intrinsic motivation (Skinner & Belmont, 1993).They appreciate the students' perspective (Reeve, 2009;Ryan & Deci, 2020).They engage in autonomy support (supporting the need for autonomy), providing structure (supporting the need for competence), and teacher involvement (supporting the need for relatedness).Table 1 shows an overview of the basic needs and need-supportive teaching behaviors.In order to support autonomy, teachers try to understand, acknowledge and respond to their student's individual interests, perspectives, preferences and initiatives (Niemiec & Ryan, 2009;Reeve, 2009;Ryan et al., 2023).For example, they do so by soliciting students' input regarding their interests and lesson contents (Cents-Boonstra et al., 2020), by providing students with options in assignments, or providing tasks that align with their goals or interests (Assor et al., 2002;Patall et al., 2010), and by using inviting, non-controlling language (Jiang et al., 2019;Reeve, 2009) such as "You might consider ...".However, when teachers give choices in irrelevant and uninteresting tasks (Assor et al., 2002) or by using verbs like "you must" or "have to" choose (Sarrazin et al., 2006), choices can also feel like pressure (Ryan et al., 2023), and become need-thwarting.
By providing structure before and during class, teachers support students in their need for competence (Haerens et al., 2013).For example, they do so by being clear about what they expect from their students (Sierens et l., 2009;Vansteenkiste et al., 2012).They also offer challenging tasks (Niemiec & Ryan, 2009;Reeve, 2009;Reeve et al., 2004) such as an open or complex assignment, give feedback (Mouratidis et al., 2008), and explain why particular knowledge is of use or relevant (Assor et al., 2002;Ingram, 2012;Reeve & Jang, 2006;Skinner & Belmont, 1993).When teachers help students in recognizing the long-term relevance of their learning students are likely to be more engaged in learning activities and, as a result, perform better in demonstrating their competence (Vansteenkiste et al., 2006).However, students become frustrated when they feel overchallenged by complex tasks or expectations that are too high (Scager et al, 2014).Thus, structure, not to be confused with control, provides helpful informational support through clarity of goals and guidelines (Aelterman et al., 2019).With the help, guidance, and feedback students need, they can make progress and feel competent.Good structure scaffolds learning when students encounter challenges, combined with positive guidance and feedback to grow and help when obstacles are encountered (Ryan et al., 2023).It is a challenge for teachers to find the right balance between supporting autonomy and providing structure (Vansteenkiste et al., 2012) to meet the diverse needs, interests and abilities of students Scager et al. (2017).
Teachers who show that they are involved with their students meet their students' need for relatedness and a sense of belonging.Students can experience a sense of belonging by, for example, feeling socially comfortable in classroom interactions, academically respected and heard, and feeling interpersonally connected to their peers and the teacher (Brekelmans et al., 2000;Ingram, 2012).Involved teachers display an open attitude and a personal interest in and concern for their students (Ryan & Deci, 2017).The student's needs for relatedness calls for teachers to know their students' concerns (e.g., Niemiec & Ryan, 2009).It also necessitates fostering positive academic connections between students, for instance during collaboration on a project (Ingram, 2012).It is important for teachers to be aware of the influence that their facilitation of positive peer interactions has on their students' sense of social belonging (Ingram, 2012).Fulfilling the need for relatedness contributes to student engagement and well-being (Ryan & Deci, 2000).

Dimensions of need-thwarting teaching behavior
Teachers who demonstrate need-thwarting teaching behavior may hinder their students' intrinsic motivation (Bartholomew et al., 2011;Van den Berghe et al., 2013).They are less responsive to students' perspectives (Soenens et al., 2012) and more distant (Aelterman et al., 2019;Brekelmans et al., 2000).Teachers can thwart students' basic needs by being controlling and pressuring students to behave, by creating a chaotic environment without information and guidance, and by having cold interactions with their students that lead to feelings of loneliness and alienation (Deci & Ryan, 2000;Haerens et al., 2013;Ryan et al., 2023;Van den Berghe et al., 2016).Table 1 provides an overview of need-thwarting teaching behaviors.
Through both direct and indirect control, teachers may thwart their students' need for autonomy (Assor et al., 2005;De Meyer et al., 2014;Haerens et al., 2015;Reeve, 2009;Van den Berghe et al., 2013;Vansteenkiste et al., 2005).Controlling teachers take the lead and steer their students' thoughts, feelings, and actions in a particular direction (Assor et al., 2002).For example, they may use commanding language (Assor et al., 2005) such as "You have to meet all the criteria."They may push students towards a single correct answer (Sarrazin et al., 2006) or may not allow any criticism (Assor et al., 2002).Controlling teacher behavior may indirectly induce feelings of guilt, shame, or fear (Soenens et al., 2012) for example through a statement such as: "If you are not able to pass this simple test, you are not fit for this profession." Teachers who create a chaotic learning environment may thwart satisfaction of their students' need for competence (Ryan & Deci, 2017).They do so through unclear lesson objectives and by giving too few rules or too little information about what they expect from their students (Reeve, 2009;Van den Berghe et al., 2013).They also adopt a laissez-faire approach (Aelterman et al., 2019).
Finally, by being unfriendly, impatient, and distant towards their students, teachers may hinder satisfaction of their students' need for relatedness (Ryan & Deci, 2017;Van den Berghe et al., 2013).They do so, for example, by ignoring or excluding students (Skinner & Belmont, 1993).

Observable teaching behaviors
Studies of teaching behaviors from the perspective of selfdetermination theory have previously been carried out in primary and secondary schools, with a focus on student perceptions of teaching behavior (Stroet et al., 2013) and self-reports from both students and teachers (Aelterman et al., 2019;Hornstra et al., 2021).Self-reports have also been obtained from teachers in higher education (Vermote et al., 2020).Using a vignette-based self-report questionnaire, this study focused on how teachers perceived their motivating and demotivating teaching styles (i.e., autonomy support, structure, control, and chaos) and the possible antecedents of their style (i.e., motivation and mindset).They aimed to gain insight into how these teaching practices are interrelated and form a cyclical structure within higher education.
In earlier research, need-supportive and need-thwarting teaching behaviors were viewed as opposites (Jang et al., 2010).Later studies have shown that teachers combine both need-supportive and need-thwarting teaching behaviors in their lessons (Aelterman et al., 2019;Vermote et al., 2020).However, these studies of teaching behavior have provided little insight into concrete teaching behavior in class.Observational studies of teaching behavior have mainly been conducted in a laboratory setting (e.g., Deci et al., 1994;Reeve & Jang, 2006).Only a limited number of observational studies have been conducted in authentic teaching situations, often in secondary schools and secondary vocational institutions (Cents-Boonstra et al., 2020;De Meyer et al., 2014;Haerens et al., 2013;Jang et al., 2010;Jiang et al., 2019;Van den Berghe et al., 2013, 2016).In this study, we complement the existing experimental and self-report studies by conducting video observations in authentic honors education teaching settings.

Context, sample, and procedure
The 25 teachers and 250 students of this honors program form a close-knit community.Each class has 3 h of lessons in the morning or afternoon, and students work independently or in groups for the rest of the day on lesson-related assignments.This structure ensures that focus on the day's lesson topic is maintained.
Teachers were informed about the study's purpose in a team meeting, after which the specific information was sent by email.Four teachers expressed their willingness to participate.In individual agreements, it was then decided which three lessons and which classes would be filmed.We opted for several lessons per teacher to obtain as varied a picture as possible of different teachers' need-supportive and needthwarting teaching behaviors.We also decided to record lessons in different classes and weeks to minimize random factors (e.g., the time of day or unexpected situations; Cents-Boonstra et al., 2020).These 12 lessons formed a convenience sample (Marshall, 1996).The first author informed students about the study during class, 1 week before the first recording.All participating teachers and students took part voluntarily and signed an informed consent form.Confidentiality was guaranteed for teachers and students.
The observed teachers had 6 to 25 years of higher education teaching experience and 2 to 10 years of honors teaching experience.They taught first or second-year students in the course units on Organizational Behavior, Project Management, Global Challenges, and Statistics, focusing on knowledge, understanding, and application.The sizes of the observed classes ranged from 16 to 20 students.

Data collection
Data were collected between October 2019 and January 2020.A video camera with robotic support (SWIVL) was set up at the back of the class and followed the teacher's movements.This setup meant the teacher was always in the frame, while students were filmed mainly from behind.The teachers wore a transmitting microphone around their neck, including during the coaching discussions with smaller groups of students.The use of the SWIVL camera meant that no adjustments to the classroom setup were required.The study was approved by the ethics committee of the University of Amsterdam under the number 2019-CDE-11252.

Unit of analysis
Previous observational studies of teaching behaviors frequently opted for the time-sampling method, in which observations were made at 5-minute time intervals (Cents-Boonstra et al., 2020;Haerens et al., 2013;Van den Berghe et al., 2013) or during a single specific lesson phase, for example, the beginning of the lesson (Van den Berghe et al., 2016).The disadvantage of this approach is that the interactions between the teacher and students become fragmented.
To avoid fragmentation, we opted for instructional patterns as the meaningful unit of analysis in this study, as this gave us a better T. Kingma et al. understanding of the relationship between teaching behaviors and the method of instruction.We used Brekelmans et al. (2000), Holt et al. (2006), andKember (1997) as a starting point for describing some general information about the three main groups of instructional patterns in order to delineate the roles of students and teachers in the classroom context.These groups of instructional patterns are teacher-centered patterns, teacher-student interactive patterns, and student-centered patterns.
Through carefully watching the videos, we found the features of the five instructional patterns we used for the analysis (see Table 2).We distinguished the different instructional patterns by when a teacher switched to a distinctively different form of instruction.For example, the teacher might begin by giving an explanation, and then switch to a group assignment.In order to distinguish the instructional patterns that were operating, the first two authors examined one recorded lesson for each teacher, and identified types of instructional patterns.Table 2 below gives an overview of the features of the instructional patterns we found in the video recordings.
Each instructional pattern was established by the beginning of a different pattern, a break, or the beginning or end of a lesson.We found that each instructional pattern lasted 10 to 15 min on average before a transition occurred.Teachers alternated between the different instructional patterns several times in their lessons and made sure to use all three groups of instructional patterns.The video data were analyzed using ATLAS.tisoftware (Friese, 2019).In the next phase of the coding, each teaching behavior code could be assigned a maximum of one time within a single instructional pattern.

Developing the observation schedule
To develop the observation schedule, we built on existing SDT observation schedules from educational sectors outside higher education (Haerens et al., 2013;Jiang et al., 2019;Morgan, 2006;Van den Berghe et al., 2013).We supplemented them with descriptions of teaching behavior from experimental, self-report, and lab studies from various educational contexts (Aelterman et al., 2019;Gucciardi et al., 2020;Reeve, 2009;Scager et al., 2014;Stroet et al., 2013).Working with existing (validated) SDT observation schedules made it possible to compare teaching behaviors found in other contexts (Klette & Blikstad-Balas, 2018).This initially resulted in a long list of descriptions of need-supportive and need-thwarting teaching behaviors, some of which overlapped.Table 3 shows the final observation schedule for the types of need-supportive teaching behaviors, and Table 4 shows the types of need-thwarting teaching behaviors.
The final observation schedule was developed as follows.The first author coded the behaviors shown in one lesson recording.The extensive list of teaching behaviors from previous observation, experimental, self-report, and lab studies was a source of inspiration, and inductive codes were added where necessary.A joint coding session with the second author followed, to merge overlapping codes for teaching behaviors and reformulate ambiguous codes.Once the two authors agreed, the codes and the descriptions were adapted, where necessary, to the honors education context.Codes were also merged because of the considerable overlap in the extensive list of teaching behavior codes.
After recoding the first video per the modified observation schedule, the first author coded the behaviors from a recorded lesson by a different teacher.The second author then independently coded 20 % of the instructional patterns already coded by the first author.The authors then discussed the teaching behaviors codes where they did not agree.Together, they decided whether a code for a teaching behavior and the corresponding description were correct, should be modified or removed, or whether a new, inductive code was required.After agreement was reached, the first author adjusted the observation schedule, reviewed the previously coded lesson recordings, and recoded the first and second lesson recordings using the modified schedule.This process was repeated for the first lesson recording for the remaining teachers.After four coding sessions across four different recordings, the observation schedule was finalized, and the first author coded the behaviors from the remaining eight recordings using the definitive observation schedule.
Table 3 shows the definitive observation schedule for the types of need-supportive teaching behavior, with a description of the behavior and an example.The observation schedule consists of six existing and six new, inductively generated codes for providing structure (S2, S3, S5, S7, S8, S11), five existing and one new code for teacher involvement (I4), and eight existing codes for autonomy support.We had to adapt existing codes and descriptions to the honors education context.Based on the video recordings, addressing students by their first name (I1) was identified as a teacher-involvement behavior.Other studies have identified it as a type of structure-providing behavior (e.g., Gucciardi et al., 2020;Haerens et al., 2013).The note to the table indicates whether a code is deductive or inductive and what the deductive sources are.
Table 4 shows the definitive observation schedule for the types of need-thwarting teaching behavior, with a description of the behavior and an example.The observation schedule consists of five existing, one new inductively generated control code (C5), and two existing and one inductively generated cold teaching behavior code (col3).No teaching behaviors that created chaos were found in the video recordings.The note to the table indicates whether a code is deductive or inductive and what the deductive sources are.

Results
The results of the study are described in Sections 4.1 and 4.2.Section 4.1 discusses the frequency of occurrence of the different dimensions of need-supportive and need-thwarting behavior.Section 4.2 describes the types of autonomy support, structure providing, teacher involvement, control, and cold teaching behaviors observed in the video recordings.

Table 2
Features of instructional patterns.

Instructional patterns
Features of instructional patterns

Teacher-centered Instruction and explanation of theory
The teacher explains the concepts.
The teacher asks questions to check comprehension.The teacher elaborates on the students' answers.
The students' input is brief and confined to comprehension.
The teacher initiates and manages the discussion in class.

Teacher-student interactive Dialogue with the class
The teacher encourages students to ask questions and share experiences.
The teacher elicits and stimulates different students' input and opinions.
Students provide a lot of their own input.
A mutual, learning-oriented dialogue takes place.
The initiative for interaction sometimes rests with the teacher and sometimes with the students.

Student presentations
Student groups present the results of their group work.
The teacher gives feedback and invites the class to ask questions or add something.
The teacher manages the discussion in class.

Group coaching
Students work together on their project in desk groups.
The teacher coaches the groups on their project's progress in various rounds.
The teacher sits down with each group while coaching.
Students share the progress of their project work with the teacher and can say what they want to discuss.The teacher gives feedback on the points made by the students.
The initiative for what is discussed rests with the students.

Independent work
Students work on a task individually, in pairs, or in groups.
The teacher walks around, offering help and answering questions from students where necessary.
T. Kingma et al.

Dimensions of need-supportive and need-thwarting teaching behaviors
Table 5 presents how often the dimensions of need-supportive and need-thwarting teaching behaviors were observed, given as their percentage of the total number of coded teaching behaviors.
The table shows that the within the need-supportive dimensions of teaching behavior, providing structure occurred twice as often as teacher involvement and autonomy support.In addition, the dimensions of need-thwarting teaching behavior (control and cold teaching) accounted for 11 % of the observed teaching behaviors.
Fig. 1 gives an overview of the occurrence of the different dimensions of need-supportive and need-thwarting behavior during the different instructional patterns.The figure shows the occurrence of the five instructional patterns on the left side and the occurrence of the use of each dimension of need-supportive and need-thwarting behavior across each instructional pattern on the right side.
The comparatively most frequent use of the three dimensions of need-supportive behaviors (autonomy support, providing structure, and teacher involvement) was in teacher-student interactive instructional pattern of dialogue with the class and in the student-centered pattern of group coaching.The comparatively most frequent use of the two dimensions of need-thwarting behaviors (control and cold teaching) was in instruction and explanation of theory (teacher-centered) and independent work (student-centered).Morgan (2006). 4Aelterman et al. (2019). 5Gucciardi et al. (2020). 6Jiang et al. (2019). 7Scager et al. (2014).

Table 4
Observation schedule for types of need-thwarting teaching behaviors.

Code Description Example
Types of control C1 Uses controlling language d, 3 Uses words like "should," "have to," "must," or "got to," and "need to" to direct students' behavior The teacher-centered pattern of instruction and explanation of theory (n = 60) was the most often used instructional pattern, followed by dialogue with the class (teacher-student interactive; n = 40) and group coaching (student-centered; n = 33).Teachers demonstrated all the dimensions of need-supportive and need-thwarting teaching behaviors during all five instructional patterns, albeit in different ratios.
Structure-providing behaviors were the main behaviors exhibited during all instructional patterns.Teachers showed comparatively the most autonomy support during the student-centered group coaching pattern.Moreover, the teacher-student interactive patterns (dialogue with the class and student presentations) and the student-centered group coaching pattern showed comparatively the most teacher involvement.Concerning need-thwarting teaching behavior, teachers exhibited control mainly during the teacher-centered pattern of instruction and explanation of theory.They exhibited cold teaching mainly during the student-centered independent work pattern.

Types of need-supportive and need-thwarting teaching behaviors
The teachers displayed eight types of autonomy-supportive behaviors (see Fig. 2).A particular code could be assigned only once per instructional pattern.Autonomy support mainly aimed at promoting students' input, because promote students' input seems a more natural collocation.They did this by encouraging students to ask questions (during 40 % of the 166 observed instructional patterns) and by encouraging diverse responses (36 %).In addition, teachers provided autonomy support by fostering interest in learning (28 %), for example, by pointing out articles for students to do further reading themselves or websites where they could follow current developments relating to the profession.
The teachers displayed 12 types of structure-providing behavior (see Fig. 3).A particular code could be assigned only once per instructional pattern.They did so mainly through two new, inductively generated codes: providing an explanation (in 92 % of the instructional patterns) and asking questions to check understanding (72 %).In addition, teachers offered structure through positive feedback (64 %) in the form of "good" or "okay," but without further explanation.These three structureproviding behaviors often occurred together.When the lesson topic related to applying and demonstrating professional skills, they relatively often exhibited other teaching behaviors.Among these were providing a rationale (43 %) as to why the skill in question was crucial for developing one's professional conduct or giving feedback on where to go next (a new code; 23 %) or providing clear expectations (40 %).
The teachers displayed six types of teacher involvement behaviors (see Fig. 4).A particular code could be assigned only once per instructional pattern.The video recordings showed that teachers displayed this behavior primarily through respectful listening (in 74 % of the instructional patterns).The other types of teacher involvement behaviors were shown less often.For empathic behavior (40 %), teachers asked the students how they were.Teacher involvement was also shown by addressing students by their first name (36 %) and through the new, inductively  The teachers displayed six types of control behaviors and three types of cold teaching behaviors, albeit to a limited extent (see Fig. 5).A particular code could be assigned only once per instructional pattern.Cold teaching behavior was mainly demonstrated by not paying much attention to the students (in 28 % of the instructional patterns).The   behavior in the video recordings showed that teachers did not always notice that a student had raised their hand, for example, because they were looking at the information on the board.Teachers mainly showed direct control behaviors by using extrinsic motivation sources (in 19 % of instructional patterns), using controlling language (14 %), through stressing efforts needed to pass the exam (new code; 12 %), and through impatience (11 %).The new cold teaching behavior code offers irrelevant feedback (2 %) was barely seen.

Conclusion and discussion
The research question for this study was as follows: What needsupportive and need-thwarting teaching behaviors do university teachers in honors education demonstrate during their classes?We gained insight into these behaviors by using the observation tool.

Need-supportive and need-thwarting teaching behaviors
The critical finding concerning the need-supportive teaching behaviors is that teachers were twice as likely to exhibit structureproviding behaviors as autonomy support and teacher involvement.Given the presence of a much larger arsenal of structure-providing behaviors compared to those seen in secondary and secondary vocational education, we can conclude that these honors education teachers differed in their focus on supporting their students' need for competence.Teachers in honors education supported the need for autonomy and relatedness equally, in contrast to previous studies in secondary schools and secondary vocational institutions, where proportionally less autonomy support and more teacher involvement were shown (Cents-Boonstra et al., 2020;Van den Berghe et al., 2013).One explanation may be that the teachers in this study worked with a mix of instructional patterns that focused on students' input and direction regarding their learning process.It is not known how need-supportive and need-thwarting teaching behavior are related to instructional patterns in secondary and secondary vocational education, as our study is the first to instructional patterns as the unit of analysis.
As in previous studies (Van den Berghe et al., 2013), we saw in our study that the teachers were much more likely to show need-supportive teaching behavior than need-thwarting teaching behavior.Need-thwarting teaching behaviors occurred to a limited extent, and only in combination with one or more need-supportive behaviors.An explanation for the limited use of control approaches, in particular, could be found in the wide range of structure-providing behaviors that might render control unnecessary.Cold teaching by not paying much attention to the students was especially common in independent work, which seems logical, as students were working by themselves.
In this study, we identified some other teacher behaviors that were not found in previous observational studies in other types of education (secondary schools/secondary vocational schools).Based on the observations we made using the observation tool, we also noticed that the relationships between the different dimensions of honors education teaching behaviors differed from those in other types of education.The present study is, therefore, an addition to existing studies.

Constructing the observation tool
One outcome of this study was the development of a valuable tool to observe need-supportive and need-thwarting teaching behaviors in honors education.For such a tool to have content validity, it must be based on tools from previous observational studies, supplemented by findings from experimental, self-report, and lab studies (Patton, 2015).The tool's reliability was supported by an intensive inter-coder process in which codes were deleted, tightened, and supplemented for the honors education context.This resulted in tightening 26 existing codes and identifying nine new codes for teaching behaviors in honors education.
The supplementary and tightened codes for honors education mainly show broadening of the arsenal of structure-providing behaviors that teachers exhibit compared to previous observation tools.Two structureproviding behaviors not described in previous studies (in secondary schools and secondary vocational institutions) are noteworthy: providing an explanation and asking questions to check understanding.This may have been because some studies in secondary schools and secondary vocational institutions have continued to work with the existing validated list developed by Haerens et al. (2013) or Van den Berghe et al. (2013).The number of structure-providing codes was also limited in a secondary school study by Jiang et al. (2019), in which a new, validated observation schedule was developed.We made our observations in a completely different context, adding new structure-providing codes where necessary.
Another salient feature was the further breakdown of feedback that teachers in this study provided both positive feedback and feedback on where to go next (feed-forward).However, this feature was not reported in previous observational studies of teaching behaviors.Observational studies in secondary schools and secondary vocational institutions only distinguished positive feedback (praise) in the form of compliments and confirmations (Haerens et al., 2013;Jiang et al., 2019;Morgan, 2006;Van den Berghe et al., 2013), and constructive and non-constructive feedback as extremes on a bipolar scale (Jang et al., 2010).This might be caused by a change within higher education from a test culture to a feed-forward culture to prepare students for learning how to deal with the increasing complexity of the labor market (Sluijsmans & Segers, 2018).Feed-forward provides students with information about how they performed a task and how they can improve further (Hattie & Timperley, 2007), rather than whether or not the task has been accomplished.
The other new codes for structure-providing behavior involved different ways of explaining concepts.The video recordings were made with first-and second-year students at the end of semester 1, when they were still in their program's initial phase.A cognitive basis is often laid in the early years of a program.Structure-providing behaviors support the need for competence.This might explain why teachers exhibited many (different) structure-providing behaviors in the observed lessons.
The teaching behaviors aimed at autonomy support and teacher involvement were also seen in observational studies in secondary schools (Haerens et al., 2013;Reeve et al., 2004;Van den Berghe et al., 2013, 2016) and secondary vocational institutions (Cents-Boonstra et al., 2020).The only newly generated code for teacher involvement was shares personal experiences in the lesson.On the other hand, previous self-report studies in secondary schools (Aelterman et al., 2019;Vansteenkiste et al., 2012) and higher education (Vermote et al., 2020) often did not include teacher involvement.
A new control behavior, i.e., stresses efforts needed to pass the exam, was added to the need-thwarting behaviors.We did not find this code in secondary school and secondary vocational studies.This may have been because the recording occurred at the end of the semester.Previous studies have also identified teaching behaviors that contributed to chaos (Cents-Boonstra et al., 2020;Van den Berghe et al., 2013).However, we did not encounter this type of teaching behavior in the present study.Perhaps the broad arsenal of structure-providing behaviors prevented chaos from occurring.

Analytical lens
Previous observational studies frequently opted for the timesampling method with observations at 5-minute time intervals (Cents-Boonstra et al., 2020;Haerens et al., 2013;Van den Berghe et al., 2013) or during a single specific lesson phase (e.g., the beginning of the lesson; Van den Berghe et al., 2016).These studies did not link their observations to specific types of instruction.Observational studies of whole lessons are also scarce (Lemke, 2007).In our study, we observed whole lessons and used instructional patterns as the unit of analysis within those lessons.This approach can avoid the fragmentation that characterizes other observational studies.Our approach sheds light on the combinations of need-supportive and need-thwarting teaching behaviors within the authentic, meaningful context for these behaviors (i.e., specific types of instruction as part of whole lessons).Identifying a behavior within recognizable instructional patterns allows teachers to gain better insight into it.We recommend that subsequent observational studies analyze teaching behaviors in instructional patterns within whole lessons to do justice to the richness of the instructional context.

Strengths and limitations
A strength of this study is that it has produced an observation tool to shed light on need-supportive and need-thwarting teaching behaviors in honors education.A further strength is the choice to use instructional patterns as the unit of analysis.The tool we developed offers concrete elaborations of teacher behavior in honors education and can therefore be used as a point of reference in the professional development of honors education teachers regarding need-supportive and need-thwarting teaching behavior.This can be viewed as an extension of the emphasis on knowledge dissemination in training, and the potential for a more student-centered or learner-oriented approach to teaching.
The number of observations and their situation within a specific higher education context, i.e., honors education, is a limitation of this study.The four teachers who were observed each taught three lessons at the end of the first semester to relatively homogeneous groups of firstand second-year honors students at a single institute.The lessons were on subjects with a cognitive orientation, focusing on knowledge and application.Therefore, it is impossible to generalize the results to all higher education programs.However, the results do suggest directions for follow-up studies in other higher education contexts.Follow-up research with the observation scheme that was developed could be conducted with different groups of higher education students, during different semesters or academic years, and in lessons with different types of higher education learning objectives (such as forming judgments, communication, and learning to learn).This would allow broader conclusions to be drawn about the specific teaching behaviors of teachers in higher education and the typical relationships between the various need-supportive and need-thwarting dimensions.The observation tool and analysis based on instructional patterns could be used for this purpose, and, where necessary, expanded.

Concluding remarks
In this study, we presented qualitative research on need-supportive and need-thwarting teaching behaviors.Although there is a vast body of knowledge based on self-reports about teaching behavior in different teaching fields, observational studies in higher education based on video recordings are scarce.This study is an important addition to the selfreport studies and the observation tool that we developed can help to train honors and higher education teachers for their role.

Fig. 1 .Fig. 2 .
Fig. 1.Occurrence of the use of the dimensions of need-supportive and need-thwarting teaching behavior during each type of instructional pattern (N = 166).

Fig. 3 .
Fig. 3. Types of structure-providing behaviors and number of instructional patterns in which they occurred (N = 166).

Table 1
Basic needs and dimensions of need-supportive and need-thwarting teaching behaviors.

Table 3
Observation schedule for types of need-supportive teaching behaviors.

Table 3
(continued ) a adjusted code.Numbers refer to empirical studies.Descriptions without a letter or a number were defined by the authors.1 Haerens et al. (2013). 2an den  Berghe et al. (2013). 3 Aelterman et al. (2019)2013).2Aeltermanetal.(2019).3Jiang et al. (2019) a adjusted code.Numbers refer to empirical studies.Descriptions without a letter or a number were defined by the authors.1