Revisiting Peer Classroom Observations as A Teacher Professional Development Protocol: A Critical Theoretical Review of Global Practices

Peer observation is among the popular tools to evaluate teacher's performance in the classroom. As education evolves and becomes more eclectic, the focus of peer observation is not merely to assess teachers but also to facilitate them in pursuing their professional development. Characterized by the SALSA (Search, Appraisal, Synthesis and Analysis) framework of Grant and Booth (2009), this critical review searches, appraises, synthesizes, and analyzes the literature in contemporary models of peer observation globally, based on which the author adopts two peer-observation models left out by other researchers. The article then critically discusses the questions regarding the benefits and concerns, as well as proposes holistic solutions to enhance peer observation of teaching. Finally, the author points out the research gaps and offers further pedagogical recommendations for applying peer observation of teaching as a professional development protocol from both managerial and professional perspectives.

 12  to revise more up-to-date and more full-coverage models of observation, investigate perceived benefits and limitations, and propose solutions according to theoretical themes. Gosling(2002) believed that the evaluative model (Column A) is also the management model that will benefit the institution as the observation's outcome is judgmental or descriptive. Within this model, 'power' is the mere aspect that determines the relationship between the observers and the observed staff. The institutions may utilize this model in their appraisal, promotion, or teacher assessment process (Carroll & Loughlin, 2013). As the observation of teaching as an evaluative means aims at appraisal or promotional purposes, they are usually single shots rather than a continuous linear. On the other hand, developmental observation (Column B) resembles a form of mentoring in which a more experienced teacher, an education developer, or a headteacher will observe the less experienced teacher (Gosling, 2014). This form of observation focuses on developing the novice teacher's teaching ability or testing their competencies for recruitment. The developmental observation can also occur when a teacher is underperforming and needs monitoring.

Contemporary Peer Observation Models in Practice
However, the initial model by Gosling (2002) fails to include the reversed version of the developmental model that takes place contemporarily in many countries such as Vietnam or the United States. Within this reversed model, an expert teacher gives a demonstration lesson, and the novice/preservice teachers observe (Column C). This form of observation is of demonstrative purpose for the novice teacher(s) to gain more experience from the observed teacher. In a demonstrative observation section, the observed teacher functions as an experienced facilitator. At the same time, the observers are the guest teachers who are willing to learn and connect to the observation experience (New York State Education Department, 2014). This observation model may also include instructional coaches or program department consultants working closely with the demonstrating teachers to help the observers (Grierson & Gallagher, 2009). While the observed teachers may gain creditability for their expertise, this form of observation emphasizes the development of novice teachers. As they observe the lesson, they can extend their teaching repertoire, transfer suitable methods to practice, learn the program content, and contribute feedback to the demonstrative teacher (Putnam, 1985).
Previously, it is also argued that there should be an emphasis on collaborative peer observation, for this model is more encouraging in terms of mutual development, non-judgmental, more autonomous, and more dialogic (Sachs & Parsell, 2014). This observation form is a ground on which other peer-observation models develop, including the peer-reviewed and peer-supported/ peer-developmental observation model (Purvis et al., 2009) (Column D). However, peer-reviewed observation has a significant limitation as it may become unfocused at times (Gosling, 2002). Byrne et al. (2010) adapted the models suggested by Gosling (2002), adding the peer-developmental model as a model that evolves from the peer-reviewed observation (Column E). They claimed that this model allows the teachers to identify the aspects of mutual interest and focus on other aspects besides traditional teaching practices. There are three noticeable differences between peer-reviewed and peer-developmental observation as a peerdevelopmental observation: (1) Allow the professional development to happens through exploratory dialogues in addition to the observation sections; (2) Allow the teachers to specify what aspects they should focus on instead of reviewing all the aspects of teaching; (3) Open to a broader scope of interest beyond teaching practices. Table 1. Models of peer-observation (adapted from Gosling (2002) and Byrne et al. (2010))

The Perceived Benefits of Peer Observation among Teachers
There has been an extensive body of literature investigating peer observation benefits, especially the contemporary peer-supported/ collaborative model for both in-service and preservice teachers (Day, 2015;Do, 2012;Yon et al., 2002). The first benefit is the improvement in terms of both subject-matter knowledge and pedagogic content. Besides, peer-observation sessions can initiate the reflective process among teachers (Day, 2015). Peel (2005) described her own experience with peer observation, considering it an opportunity to reflect on the substantive knowledge. Besides, peer observation of teaching heightened her awareness of physical presence in the classroom. Another study by Kohut et al. (2007) also concluded that peer observation could familiarize teachers with the course's disciplinary content. This is in line with Putnam's (1985) table of perceived benefits of peer observation. However, teachers may prefer more feedback on the pedagogic strategies that they use in class. Day (2015) showed that the student-teachers who took part in peer-observation realized they should reduce teacher talking time and the cliché that teachers usually use such as good, well-done, or excellent. They could also learn each other's strategies as well as getting inspired by others' passion. By watching other teachers, observers can get new ideas and reflect on their teaching (Bell & Mladenovic, 2008). The study by Bell and Mladenovic (2008) concluded that evaluative and developmental peer observation work is accompanied by collaborative peer observation since it makes the teachers focus consecutively on the student's experience of the tutorial, the reflective practice, and assessment. Another benefit is the de-isolating effect peer-observation brings about (Bailey et al., 2001). Forty-eight percent of the teachers who participated in a survey by Adshead et al. (2006) agreed that peer-observation could have de-isolating effects on teachers. In the study, even when some teachers did not feel isolated in the first place, peer observation of teaching gave them a greater sense of connection to the teaching community.

Concerns Related to Peer Observation of Teaching
There are many concerns related to peer observation, from a managerial scale to a personal scale. First, a lack of trust in the observed novice teachers' ability may discourage the observers from receiving the mutual benefits necessary for the development of the teaching staff (Day, 2015). It is also noteworthy to address the teaching staff's challenges to specify what good teaching practices are. Even when a vast body of literature sought to list characteristics determining good teaching practice, many of them are high interference (Ramsden, 2002). These high interference characteristics are significantly subjective or even unobservable.
Consequently, this provokes doubts regarding the validity of teaching observation if it is linked to the institutional or departmental decision-making process (Kohut et al., 2007;Rinehart, 1993). Even when there are characteristics to describe a teacher's performance, the stability of their performance and effects on student academic performance is very low (Good & Lavigne, 2015;Konstantopoulos, 2014;Morgan et al., 2014). The performance of a teacher is under the influence of many different aspects related to the classroom contexts. Therefore, summative observation

VIETNAM JOURNAL OF EDUCATION
 14  sessions can coincide with the teacher performance's parachute drop, which may demotivate and inappropriately assess their teaching performance (Stodolsky, 1984;Yon et al., 2002).
There are also concerns about the formalization and the institutionalization of some aspects of the peerobservation process, for example, the discussion part (Purvis et al., 2009). This problem becomes more significant when there is a lack of institutional support and training because of the limited time allocation (Kohut et al., 2007). Additionally, many teachers attend the observation sessions unwillingly because it is merely the institutional policy rather than a real benefit. Deeming the peer observation process unauthentic (Santos, 2016), the observers may believe that they cannot benefit from others' teaching practices. In order words, peer-observation is usually onesidedly beneficial (Gosling, 2014).
The teachers also have a problem with being polite and giving constructive feedback as they do not have a structured process to follow. They are afraid of losing face and confronting their colleagues in terms of expertise (Bailey et al., 2001). As peer observation may be face-threatening, teachers also raise questions about avoiding Hawthorne effect or the observer's paradox (Thampy & Kersey, 2015). Indeed, it is common for workers, particularly teachers, to alter their behaviors compared to their daily practices when they are under observation (Rosenthal et al., 2009). In brief, there are six main concerns about peer observation: (1) Validity; (2) Formalization and institutionalization; (3) Willingness and mutual interest; (4) Politeness and confrontation; (5) The stability of teacher performance; (6) Hawthorne effect.

Duration and Frequency
Peer observations can be either a summative assessment or a formative sequence of observations in terms of duration and frequency. According to O'Leary (2012), summative peer observation is the snapshot approach to observation. Yon et al. (2002) claimed that summative peer observation may lead to a biased decision if the observed teacher is in his/her 'parachute drop'. This problem heightens when teachers undergo evaluative peer observation sessions, which puts them under institutional or departmental pressure.
Peer observation should reflect teachers' professional formative and developmental growth (Drew et al., 2017) while summative peer observation can only address the performance of a teacher at an isolated point in time. Moreover, formative peer observation can yield benefits when teachers need to familiarize themselves with various teaching situations to reach their professional goals (Knight et al., 2006). In sharp contrast, summative observation cannot thoroughly investigate the development of the teachers. Besides, formative peer observations can usually evoke more supportive feedback and constructive comments as the observer have to follow up after their initial session rather than solely give feedback (Lomas & Nicholls, 2005). This continuation of observation and experience exchanges can foster the culture of criticisms in the feedback-giving process (Amrein-Beardsley & Popp, 2012).
All in all, formative peer observation sessions can function more effectively in comparison with summative sessions as they can (1) address the continuity of teacher professional development process, (2) allow both observed and observing teachers to acclimatize with different teaching situations that enlarge their teaching knowledge repertoire, and (3) contribute to a more civilized culture of giving and receiving observation feedback. In other words, the formative peer observation of teaching can be the solution to teacher performance instability. Therefore, instead of implementing peer observation as a summative activity, the school faculty should conduct repeated peer observations or formative peer observation of teaching (Yon et al., 2002).

The Win-Win Situation
Some studies claim that observation sessions will benefit only the less experienced teachers observed (Byrne et al., 2010;Gosling & Mason, 2009). However, opposing studies on teachers' perception show that, while the focus is on the novice teachers, the expert teachers themselves notice that they also improve their teaching through the reflection of their observation (Ahmed et al., 2018;Kohut et al., 2007). As the expert teachers will observe and comments on the less experienced teachers (Gosling & Mason, 2009), the observers usually take on the roles of guiding and improving the observee's development and additionally helping them improve their teaching practices by reminding themselves of how they have developed through times. Kohut et al. (2007) conducted a study with 530 teachers using two different survey sets to explore their perception of peer observation. In this study, Kohut et al. noticed that the observers even felt peer observation improved their teaching more than the observation. Therefore, the relationship between the observers and the observed staff based on their expertise can be reciprocal (Richards & Lockhart, 1991).

VIETNAM JOURNAL OF EDUCATION
 15  The application of collaborative peer observation of teaching can further foster the balance in interests between both the observers and the observed teachers. In , 25 peer-review observation sessions took place within 12 months. After these sessions, the participants reported feeling the mutual benefits inherent in this observation form. The teachers believed that they could share their experience to curb isolation and learn from other teachers how to handle the classroom. This mutual benefit also takes place in the study of Mento and Giampetro-Meyer (2000). Collaborative peer observation of teaching can also work in the appreciative inquiry framework (Fileborn et al., 2020). In the form of a peer-developmental observation, this framework suggests that the focus of the reviews of teaching should focus on each other's strengths to understand, anticipate, and unearth the hidden potential within their teaching practices. According to Byrne et al. (2010), the friendly and non-judgmental environment that peer-developmental reviews of teaching create through dialogues and conversations makes the teacher feel the process more meaningful and essential. The teachers in the interviews also believed that the collegiate nature of peer-developmental observation allows them to develop as a group rather than as an individual. In other words, this collegiality can create both trust and mutual ownership of teacher's professional development in an institution.

The Self-Selected Observer: A Critical Friend
It is believed that the observation partners should be 'critical friends' as the observation process should rely on trust, mutual respect, and meaningful reflection-on-action (Farrell, 2001;Shortland, 2010). Francis (2001) claimed that there is little attention paid to the complexity of interpersonal communication in peer observation. This lack of communication may stem from the fear of losing face, bringing about hesitation and non-compliance in the early period of peer-observation process (Shortland, 2010). Therefore, the selection of observation partners may, in turn, curb the problems related to politeness and confrontation as well as Hawthorne effects. Teachers can feel more relaxed if they have built a good rapport with their observation partners (Hendry & Oliver, 2012).
Regarding the evaluative nature of observations that constitute the appraisal process, there must be a transparent and unbiased judgment of teacher's performance. Hence, the random pairing of observers and observers who have a neutral working relationship by the management may be appreciative and appropriate. However, a self-selected partner may yield more benefits when it comes to teacher professional development. Exploratory qualitative research by Carroll and Loughlin (2013) investigated motivations, challenges, and outcomes of 10 first-time peer-observation users in Higher Education institutions in Ireland. The study concluded that self-selected partners might increase the level of negative feedback endurance. The self-selected partners bring about comfort, and the participants felt that the criticism would be confidentially treated among the observers and the observes themselves. Besides, self-selected partner selection can also prevent confrontation and allow both parties to exchange constructive criticisms feeling defensive. Bell and Mladenovic (2008) also adopted the same approach in their study to increase the collegiality and confidentiality of the peer observation process. This is in line with the suggestion proposed by Shortland (2004), stating that informal collegiality can be more fruitful in teacher professional development.
While self-selected partners may help limit peer observations' adverse face-threatening effects, participants need to be careful with their choice. The teachers participating in Carroll and Loughlin's (2013) research claimed that it was challenging for them to strike a balance between politeness and openness. Noticeably, some observers were also under departmental pressure to give positive feedback if the observation results are related to the judgmental and assessment purposes of their friends or colleagues. Shorthand (2010) warned that some productive relationships between peers can be jeopardized in the case of conflicts of interest within the process of peer observation. Therefore, it is recommended that if the peer-observation of teaching is for the sake of the departmental or institutional evaluation, random pairing should be prioritized.

The Professional Dialogue: Formal or Informal?
Gosling (2014) claimed that dialogue-based peer-developmental observation is paramount to help teachers overcome the pedagogic assumptions that teachers usually take for granted. Rather than focusing only on either the observers or the observed teachers, the application of peer-reviewed and peer-developmental observation can bring about the conversations and dialogues that liberate the teachers and allow them to voice their opinion about teaching practices (Brookfield, 2005).
The feedback session can be informally arranged among the teachers to discuss whatever areas they think they should focus on. The feedback session can also be a compulsory formal session in which teachers have to follow the strict guidelines of the institution. A study showed that institutionalized feedback is suitable for evaluative purposes (Mento & Giampetro-Meyer, 2000). These formal feedback sessions can prompt teachers to speak more and

VIETNAM JOURNAL OF EDUCATION
 16  seriously consider peer observation of teaching as a part of their actual professional development (Purvis et al., 2009). Also, formal feedback is more suitable and time-saving if the observation group has more than two members, particularly if it demonstrative observation where one experienced teacher mentors novice teachers. Additionally, formal feedback sessions are also beneficial if the teachers have limited experience with peer observation or are unwilling to work with a peer through self-reflection (Gosling, 2014).
On the other hand, while some institutions consider formalizing the feedback exchange and discussion, the teachers participating in the study by Purvis et al. (2009) believed that this could bureaucratize the professional development process. They claimed that discussion revolves around their teaching practices could have been kept informal. This informality preference is in line with the idea that professional development should rely on informal and non-formal learning to identify the gaps that cannot be revealed by informal learning (Misko, 2008). Although informal learning is invisible, it is a part of the developmental process, especially when initiated as a part of the learning's choice (Callanan et al., 2011;Eraut, 2004). In a case study by Shortland (2010), informal conversations between the observers and the observed teacher strengthen the trust and deepen the understanding between them. The informality of the professional dialogues creates an atmosphere that is less stressful and more conducive to development (Hammersley-Fletcher & Orsmond, 2004). From the studies' results, the author suggests the following table of feedback sessions suitable for each peer observation model (Table 2). formal → informal formal informal informal Feedback should be formal for an evaluative session as transparency is required, especially if the observation is for appraisal purposes such as promotion or salary consideration. On the other hand, developmental observation, in which an expert observer gives training to a less experienced observed teacher, should start as formal sessions to familiarize the observed with the process and foci of their teaching procedures. After that, as novice teachers develop, the feedback session becomes less formal to reduce pressure and save time. Likewise, for demonstrative peer observation, the feedback exchange should also be formal to channel a larger audience as there is usually one experienced teacher who will demonstrate what they do in class to the group of guest teachers. In sharp contrast, as regards peer-reviewed and peer-developmental observations, the feedback session should be informal as the outcome of the process focuses on the teacher's development rather than a departmental report. The formative nature of these collaborative observations would make formal discussions time-consuming and bureaucratic. Instead, short informal but focused exchange would allow the teachers to reflect on the specific areas to initiate changes with more minor degrees of interval.

Hawthorne Effect
Theoretically, Hawthorne effect is a phenomenon when the participants' awareness in an experiment, created by the experiment conditions, interacts with the independent variables and subsequently facilitates the effects on the dependent variables, thus leading to ambiguous results (Cook, 2015). Regarding education, Hawthorne effect happens when the experimental or observation environment changes the participants' (teachers and students) behaviors. In other words, the behaviors change when the teachers and students know that they are observed. Not all researchers agree with the existence of Hawthorne effect. Franke and Kaul (1978), Jones (1980), and Wickstrom and Bendix (2000) believed that Hawthorne effect is insignificant and controversial or hard to happen in different contexts. Levitt and List (2011) failed to replicate the original study that confirmed Hawthorne effect's existence. Therefore, whether Hawthorne effect occurs in peer observation of teaching requires further research. A study by  showed the mixed findings of Hawthorne effect. In this study, while some teachers changed their behavior to include the observers in their lessons, others refrained from acting differently from what they usually did.
Assuming that Hawthorne effect exists and wears off with time (Leonard & Masatu, 2006), peer observation should be a formative process so that after several sessions, the observation can reflect teachers' actual behaviors (Evans, 2014). In sharp contrast, Cook (2015) claimed that Hawthorne effect is consistent over time. In this case, the author suggests that formative collaborative observation, the observer's self-choice, and informal feedback sessions can allow the observed teachers to start in their comfort zone and express themselves more freely. In brief, whether Hawthorne effects in peer observation of teaching is a fundamental matter to consider or just a myth still needs further

VIETNAM JOURNAL OF EDUCATION
 17  investigation. However, teachers can take on peer observation of teaching with more comfort and willingness with the aforementioned strategies.

Pedagogical Recommendations and Research Gaps
As teaching becomes more eclectic, the purposes and function of peer observation of teaching also evolve. Rather than focusing exclusively on the evaluative nature, peer observation of teaching is now implemented mainly in five different models: evaluative, developmental, demonstrative, peer-reviewed, and peer-developmental. While different models suit different needs and specific teaching and learning contexts, this critical review theoretically supports developmental collaborative peer observation as it contributes to the teacher's development in various aspects. Technically, developmental collaborative peer observation can reduce fear, balance mutual interest, and reduce the observer effects.
School and institution administrators should come to terms with the eclectic nature of teaching in the new areas, thus allowing flexibility in the observation process. In the foreseeable future, it is suggested that even when the observation is for assessment purposes, there should still be room for teachers to move from conventional teaching steps. The school management should advocate developmental collaborative peer observation, allowing teachers to focus on their topics of concern. These topics do not need to be limited to teaching practices but can be about other areas of professional development. In the long run, the leaders of institutions should create and foster a professional community network where teachers can pair and participate in different peer observation types. The schools should also provide teachers with basic guidelines for conducting peer observation; however, the teacher should be informed that the observational experience should go beyond what is written in the documents (Gosling & Mason, 2009). These communities can be the platform for teachers to share their reflection-on-teaching or conduct action research.
Teachers should participate in peer observation as a reflective, collaborative, and developmental process. Moreover, teachers should distinguish between summative and formative peer observation. For formative observation, the goal is to develop their professional skills, not to judge or assess. Therefore, when practicing formative peer observation, teachers should behave as they usually do rather than put on a pretense to please the observer (Evans, 2014). If they are allowed to select their observational partner, teachers should point out particular areas they want to focus on and conduct the observation. When attending to other teachers' observations, teachers should be objective as a 'critical friend' rather than avoid conflicts (Megawati, 2018;Thampy & Kersey, 2015). They should also avoid comparing their teaching styles with other teachers but rather reflect on ideas, strategies, and approaches to teaching as different teaching contexts require different sets of skills.
Although there is a shift from using peer observation as an evaluative tool to a professional development protocol, most research works are small-scale and localized. There is a call for more empirical research on the effectiveness of peer observation in professional development. Research has mainly focused on teachers' perception; however, other aspects, such as the effects of peer observation on value-added academic measures or class management, are also worth consideration. There should also be research on how students perceive themselves as a part of the observation. As mentioned above, there is still limited research into other psychological aspects related to peer observation of teaching, such as the Hawthorne effect and teacher comfort zone, which opens up more in-depth studies. Additionally, although the author does not advocate the use of peer observation as the only evaluative tool for appraisal, future research can also investigate what aspects should be focused on in other to assess the teacher's performance more objectively and accurately. The effectiveness of the application supporting tools and devices such as an audio recorder or video recorder to reduce observation anxiety, fear, and stress can also be investigated. Besides, the development of computer-assisted teaching and learning and the outbreak of Covid-19 increase the demand for online learning and teaching flipped classrooms and blended learning. This new learning trend may require a search for other approaches to peer observation, which is another potential area for research.

CONCLUSION
This critical review investigates five different models currently popular in the world: evaluative, developmental, demonstrative, peer-reviewed, and peer-developmental. Subsequently, the author analyzes the merits of the peer observation of teaching, which benefit the understanding of disciplinary content and pedagogic strategies, initiate and maintain reflection of teaching practices, and de-isolate teachers. On the other hand, the article also reviews six main concerns related to peer observation of teaching, ranging from the duration, validity, the choice of observational partner, formality, and other psychological aspects. While it is virtually impossible to thoroughly outline all the concerns and criticism against peer observation of teaching, the author evaluates what aspects practitioners of peer observation should prioritize. Careful choices of conducting peer observation should be made according to each

VIETNAM JOURNAL OF EDUCATION
 18  specific teaching context rather than strictly follow cut-and-dried procedures so that peer observation is truly a blessing in professional development.