Teacher-student interaction in a special school for students with developmental disabilities in Chinese context

Abstract

This paper analyzes interactions between teachers and students with developmental disabilities in a special education school in China, which is still the main educational placement for the disabled in China. Video observation data collected from six students in second grade and one teacher, were coded by an improved Flanders Interaction Analysis System (iFIAS). Results indicate that the teacher played the dominant role in the class while her/his instruction was student-oriented, several evidence-based strategies were used to motivate students’ engagement, including questioning, timely feedback, reinforcement, game teaching, students’ self-determination. Few peer interactions happened due to homogeneous grouping. Besides, the frequency of interaction patterns between teachers and students with developmental disabilities (DD) of different learning ability was similar while the interaction patterns were different. There were more non-academic interactions between the teacher and students with most serious disabilities in the class than others, while there existed more cognitive extension-partial and full interactions between teachers and students with least serious disabilities in the class than others. The results shed light on how teachers interact with students who have DD, and how interactions can be differentially delivered to students with DD of various learning abilities.

1. Introduction

Although the idea of inclusive education is accepted by scholars (Chen 1996, Deng and Poon-McBrayer 2004, Deng and Zhu 2016, Ding 2007) and the government in China, segregated special schools are still one of the main education placements for children with disabilities in China. After more than three decades’ development of public special schools for children with developmental disabilities (DD)¹ in China, the demand of entrance opportunities for children with DD has been gradually met. By 2017, China had 2,107 special schools, which served 578,800 students with DD (Ministry of Education of People’s Republic of China 2018). Thereupon, more attention has been paid on the quality of education for students with disabilities. At the policy level, the Plan for Upgrading the Level of Special Education (2014–2016) and the 2nd Plan for Upgrading the Level of Special Education (2017–2020) proposed various strategies in order to promote the quality of instruction in special schools (General Office of the State Council of China 2014, Ministry of Education of People’s Republic of China 2017).

As a significant antecedent for the quality of education (Granic and Patterson 2006, Hagenauer and Volet 2014), teacher-student interactions in classrooms play an important role in contributing to teacher-student relationships and interpersonal adaptation (Pennings et al. 2018), ensuring a sense of belonging and comfort in classrooms (Englehart 2009), and improving students’ academic achievements (Allen et al. 2013, Downer et al. 2010, Hafen et al. 2015, Nugent 2009, Ryan and Patrick 2001, Slavin et al. 2009). In the process of instruction, teachers and students build their classroom via interacting with each other. However, research studies have shown that teachers react less favorably to low achieving students (Allen et al. 2013, Brophy and Good 1970, Empson 2003, Good 1970, Van Ewijk 2011). More negative teacher-student interaction patterns are related to less praise from teachers, more likelihood of social isolation, lower adjustment to school environments, and poorer academic performance (Birch and Ladd 1997). Apart from leaving students angry and resulting in more conflict in teacher-student relationships (Mejia and Hoglund 2016), those problems have continual influence on the future development of students, who may be in trouble from high risk to eventual school drop, and even negative adaptation in early adulthood (Caspi et al. 1987, Gest 1997). Students with DD usually have limited academic achievements and undesirable social adaptation related to intellectual disabilities, emotional or behavioral disorders (EBD) or other disorders, which negatively affect the quality of teacher-student interactions. Comparing to general education, there is no education quality evaluation from the government for education in special schools, which has led to the quality of special education in special schools being mist in China. In addition, the unique current social culture and educational environment in China may limit the generalizations of dominant Western research findings (Xiao et al. 2007). Therefore, it is significant to research interactions between teachers and students with DD in China, which will shed light on the quality of special education in segregated settings in the country.

1.1. Teacher-student interactions and the quality of teaching

Hagenauer and Volet (2014) found that frequency and the quality of teacher-student interactions are factors contributing to quality teacher-student relationships and the academic achievement of students. In class, teachers are viewed as the manager and guide, whose interactions with students will influence the classroom climate and students’ behavior to some extent (Buyse et al. 2008, Jennings and Greenberg 2009, Jussim and Harber 2005, Rimm-Kaufman et al. 2009). Cadima et al. (2010) have shown that higher quality teacher-student interactions, particularly with reference to classroom organization, have a positive effect on first grader’ literature skills including vocabulary and print concepts. And students with lower math skills (e.g. number identification, conceptual understanding) also benefit from high-quality teacher-student interactions.

The essence of teacher-student interactions is mutual actions between two humans with emotions. Academic guidance integrated with emotional support could facilitate student learning. In a classroom with high instructional and emotional support (e.g. concerning interests and initiative of students, providing challenging questions and critical issues, creating active social relationships) (Matsumura et al. 2008), students will be more actively engaged in classroom activities and have more interactions with their peers (Durlak et al. 2011, Hamre and Pianta 2005), perform higher math skills, and have high self-confidence in their academic abilities (Perry et al. 2007). Students will have high self-concept if teachers show close relationships with them (McFarland et al. 2016). Specifically, when students experience acceptance from teachers and peers, they are more likely to actively interact with teachers. Consequently, higher academic engagement (Hughes and Im 2016) and teaching quality are constructed. Teachers invest more emotion, energy and time in academic activities, and students will learn more vocabulary and decoding skills in their first grade (Connor et al. 2005). Similarly, Gehlbach et al. (2016) found that improved relationships prompted teachers to interact with students frequently and effectively who earned superior course grades five weeks later, narrowing the school’s performance gap by more than 60%.

1.2. Interactions between teachers and students with DD

Inclusive education has pervaded the world especially in the West since the 1980s. Most students with DD are placed in ordinary schools and inclusive classrooms which are considered to be the least restrictive environment (LRE) and social justice placement (Artiles et al. 2006, Waldron and McLeskey 1998, Yell 1995). However, special schools and special classrooms still exist and act as one of the main educational placements in the Chinese context nowadays.

Teachers’ perception of students strongly influences teacher-student interactions wherever in regular schools or special classes. Special education teachers hold lower expectations for the performance of deprived and disadvantaged children (Jones 1972, Salvia et al. 1973). Differential treatment of students might be viewed as a desirable characteristic of instruction when adapted to students’ levels of understanding (Walberg and Wang 1987). It has been claimed that students in special classes hardly receive instruction to fulfil their individual needs (Jordan et al. 1997). Although Cooper (1979) pointed out that teachers may have differential treatment when providing feedback based on their expectations of students’ ability, in the light of Dembo et al. (1978), no differential communication was found with various students from special education teachers. During classroom teaching, control techniques are more likely to be applied when teachers interact with students (Fink 1972, Harris 1982). Indeed, the nature of these interactions are composed of teacher commands and student compliance rather than pro-social behavior (Shores et al. 1993). Teacher-student interactions in special schools or special classes seem to be of lower quality overall.

With the fading of these separated settings, general teachers are responsible for educating students with DD. Researchers have compared alternative settings and found combing the educational placement of resources rooms or regular classes with individualized instruction were good for disabled students’ academic and social outcomes (Madden and Slavin 1983, Rich and Ross 1989). Many research studies were conducted to analyze interactions between teachers and students with DD, including learning disabilities (LD) (McIntosh et al. 1993), mildly handicapped (Alves and Gottlieb 1986), and EBD (Sutherland 2000).

Numerous studies have shown that teachers pay more attention to students with disabilities in classes. Cameron (2014), for example, found that inclusive classrooms were organized along traditional lines with the vast majority of interactions provided by general educators to the whole class, followed by brief intervals of one-to-one interactions directed towards individual students, most frequently children with disabilities. Teachers pay more attention to students with disabilities including adjustments of their instruction with the aim of students receiving effective teaching. Cook and Cameron (2010) found students with DD received more teachers’ concern than others.

In contrast, some studies have shown that the frequency of interactions between teachers and students with DD is low. For example, students with EBD probably exhibit behavior problems (e.g. higher rates of aggression, noise making) which can lead to negative patterns of interactions between teachers and students (Sutherland et al. 2008). Some regular teachers would regard LD as unchangeable so as to shrink interactions with them (Jordan and Stanovich 2001, McIntosh et al. 1993, Salvia et al. 1973). Moreover, Students with LD receive more teacher criticism and warnings due to their interactions (Chapman et al. 1979).

1.3. Interactions between teachers and students with DD in China

Sutherland et al. (2008) consider the classroom context has possibly shaped the complexity of interactions between students’ learning and behavior problems. The development of special education in China has its own foundation and characteristics (Deng et al. 2001, Yang and Wang 1994) due to its social, cultural and historical context. Although the existence and sustainable development of special education is accepted by the government in China, the shortage of qualified special education teachers (Wang and Mu 2014) still cannot meet the education requirements for the largest population of students with disabilities in the world (Mcloughlin et al. 2005). The class size in schools for students with DD is still not small (Li et al. 2010). Therefore, the traditional whole-class teaching model in Chinese special education is common.

Special schools in China primarily consist of schools for the blind, schools for the deaf and schools for students with DD. The research studies on interactions between teachers and students with disabilities in China are rare, and most research focuses on interactions between teachers and students with blindness and ID. Li (2015) analyzed the interactions between 54 third-grade students with ID and their teachers and found that interactions were characterized by single form, low frequency and limited content. Another study (Wu 2017) has parallel consequences which revealed that teachers dominate classrooms with lower students’ engagement. The form of interactions between teachers and students with disabilities is mainly question and answer (Q&A) mode that is expounded as ‘ask–answer–feedback’ (Huang 2017, Li 2014, Wu 2017). What is more, the frequency of interactions and asking requirements vary with grade level. Results from investigations are certainly mixed. There are some studies showing that teachers have more initiative interactions with students in special education classes (Shao 2016, Zhou et al. 2016). Ellsworth and Zhang (2007) found interactions between teachers and students with disabilities to be active with students responding positively and teachers giving timely feedback.

Empirical research studies related to interactions between teachers and students with DD are limited in China (Wu 2017). Teacher-student interactions in special education schools are neglected by researchers. The complete picture of interactions between teachers and students with DD in special schools remains unclear. Therefore, it is meaningful to analyze teacher-student interactions in special education schools in the context of China as well as evidence-based potential practical strategies. The result of the study cannot only give suggestions to improve teacher-student interactions in special schools, but also show teacher-student interactions in the Chinese context which is the biggest special education system in the world.

Two questions were addressed in this study, with the aim of improving our understanding of teacher-student interactions in special schools in China, revealing the influencing factors and interpreting their relevance in the contexts.

1. What are the characteristics of teacher-student interactions in special schools in China?

2. Whether there were differences of frequency and patterns of teacher-student interactions in different groups with different learning abilities?

2. Method

2.1. Video analytics

Video has been generally applied in the research of anthropology, sociology, medical and education for a long time (Collier 2001, Liu et al. 2018, Makdissi and Davis 2016, Stigler et al. 2000). In the field of education, the trend toward analyzing classrooms by means of video continues to heat up (Jacobs et al. 1999, Seidel et al. 2011). Classrooms full of teacher-student interactions are complex and dynamic settings. The teaching process is available to visualize when video recording; thus, storage, playback, and analysis have become technically convenient to implement. Due to its nature, the quality of video files does not diminish over time or from usage. In addition, numerous neglected information in live classes can be found within these objects. The merit of video analysis as a research method has been identified in scores of literature studies (e.g. Barron 2007, Erickson 1986).

The video material used for analysis could consist of selected ‘edited’ sections and ‘raw’ data from lessons (Stigler et al. 2000). With the characteristics of space and time, videotape taken by a machine is deemed as a bridge for observation indirectly and analysis directly. The application of video technology means repeated presentation of observable behavior, as researchers are approaching real time (Stigler and Hiebert 2000). Further, during teaching and learning, given that words and actions of teachers and students are not necessarily consistent, video analysis will then assist in distinguishing discrepancies between saying and doing (Ruhleder and Jordan 1997). In summary, with the aid of video analysis, it is possible to see the sight of interactions between teachers and students with DD in classrooms.

2.2. Participants and context

A second-grade demonstration class called ‘My Snacks’ from a special school for students with DD (school named as AH) was selected. AH located in **, the capital of China, which is a first-tier city with rich educational resources. AH represented the high development level of special education. Guided by the Compulsory Education Course Setting Experiment Plan for the Special Schools for DD issued by the Ministry of Education, AH had delved into integrated curriculum several years ago, which is a popular form of curriculum in special schools for students with DD in China. Integrated curriculum is a model that minimizes instruction in isolated academic disciplines by combining goals and objectives from a variety of areas (Reisberg, 1998). Special education teachers using an integrated curriculum may select a subject for instruction and include related topics from other areas in the unit plan. Students with special needs often require instruction in life/independent living, social, and academic skills Fu et al. 2016, Sun 2004, Yu 2016). By using the integrated curriculum model, special teachers can infuse skills from these areas into the typical content focused on in the classroom. With the framework of integrated curriculum, activities are created to design professional supports that are responsive to students’ needs. Lonning et al. (1998) found it was of more meaning when course contents taught by teachers are integrated compared with separate concepts. Other evidence based on cognitive science and educational research have discovered that supporting an integrated curriculum can bring out comprehensive promotion in brain development, academic and social performance (Fantuzzo et al. 2011, He 2009, Loepp 1999).

The lesson My Snacks was selected since it was a demonstration class, displayed on a seminar of curricula and the teaching reform of special education in China in May 2018. The class was applied package-shift mode in which 3 teachers were responsible for the daily teaching and training, routine management and student life management of a class (Li et al. 2010). The students in the class included five children with moderate or severe ID, one child with severe cerebral palsy and two children with severe autism spectrum disorder (ASD). The two children with severe ASD were pulled away from the classroom for one-on-one training rehabilitation.

Participants of the lesson My Snacks included a teacher and 6 students with DD in the classroom. The teacher had a master’s degree in special education and taught students with DD in AH for five years. During the teaching process, the students who were between eight and ten years old were divided into three groups according to the their learning performance and the principle of homogeneous grouping: (1) students with moderate ID in group A (n = 2) had the best cognitive abilities in all groups with voice resolution ranged from 46% to 65%, who could have simple conversation while be difficult to express long sentences and could complete learning tasks under the occasional guidance of the teacher. (2) Students with severe ID in group B (n = 2), having capability of partial verbal communication with voice resolution ranged from 26% to 45%, who needed more verbal and behavioral assistance from the teacher. (3) A student with severe ID, accompanied by his mother, whose voice resolution ranged from 11% to 25%, and a student with severe cerebral palsy, accompanied by a caregiver, having no verbal communication belonged to group C (n = 2), who needed frequent, continuous and high-intensity assistance.

Students diagnosed with moderate or severe developmental disabilities were: (1) nominated by their teacher as having moderate or severe developmental disability and (2) diagnosed with moderate or severe developmental disabilities by the public hospitals in Beijing.

2.3. Videotape coding system

The Flanders Interaction Analysis System (FIAS) was originally designed by Flanders (1963) aiming to observe and code the verbal interchanges between a teacher and his/her students. From the initial use of data collection to subsequent application to improve teacher effectiveness (Flanders 1970, Flanders and Simon 1969), FIAS remains an influential tool for exploring classroom interactions through continuous innovation and localization (Fang et al. 2012, Freiberg 1981, Wen and Cui 2016). Goronga (2013) used FIAS to explore the nature of classroom interactions in a primary school and pointed out that teachers dominated classroom verbal interactions. Nurmasitah (2010) adopted FIAS directly to analyze a 45-minute geography immersion class; besides the similar conclusion that teacher-controlled activities had more direct influence than indirect influence, it also revealed that students were active enough in teacher-student interactions. In addition to diagnosing teaching aspects, modified FIAS can also be used to improve teaching behavior in medical education (Lyon et al. 2014).

However, typical FIAS exposes some limitations such as insufficient attention to student class performance and inadequate adaptability to classrooms with information technology, while information technology has been widely used in classroom teaching, consisting multimedia, computer, network technology, and even digital technology such as the application of mobile devices (Fang et al. 2012). To preserve the traditional analytical function of FIAS and deal with existing problems, we used a coding system stemmed from an improved Flanders Interaction Analysis System (iFIAS, see Table 1) adapted by Fang et al. (2012). The major improvement of FIAS added two categories, including silence and technology. The former one clarified beneficial silence and helpless silence in class, of which, beneficial silence meant silence being conducive to teaching, such as teachers allowing students to think independently or to complete the study of quiet. The later one assisted classroom analysis in typical digital environments, including teacher manipulating technology and student manipulating technology. IFIAS can encode teacher language, student language, silence and judgment of the teaching model through an analysis matrix. In addition, application to information technology, distinguishing the type of teacher questioning and students’ initiative behavior are the innovations of iFIAS. The reliability has been demonstrated by studies for classroom interaction analysis (Li 2015, Wang 2017, Xue 2015).

Table 1.

The improved Flanders Interaction Analysis System (iFIAS).

Categories		Code theme		Specific description
Teacher talk	Indirect influence	1. Accepts feeling		Accepts and clarifies the feeling of students in a non-threatening manner. Expressing a positive or negative attitude toward students’ responses clearly
		2. Praises or encourages		Praises or encourages action or behavior of teacher with subjective judgement. For example, ‘Anan observed very carefully, everyone should learn from him’. Nodding head, applauding or saying, ‘um’ are included.
		3. Accepts or uses ideas of students		Repeating, clarifying, building or developing ideas suggested by a student.
		4. Asks questions	4.1 Asks opened-ended questions	Asking questions with multiple possible answers based on contents.
			4.2 Asks closed-ended questions	Asking questions with certain answers which may be ‘yes’ or ‘no’.
	Direct influence	5. Lecturing		Giving facts of opinions about content or procedures, expressing his/her own ideas; asking rhetorical questions.
		6. Giving directions		Directions, commands or orders with which students are expected to comply to make classroom more orderly.
		7. Criticizing or Justifying Authority		Statements intended to change student behavior from non-acceptable to an acceptable pattern; stating why the teacher is doing what he is doing; extreme self-references.
Student talk		8. Response		Responding to questions raised by the teacher passively.
		9. Initiation	9.1 Active response	Responding to questions raised by the teacher actively. Raising his/her hand and name answered by the teacher.
			9.2 Active asking	Initiating to ask questions and express opinions.
		10. Peer Discussion		Discussing and exchanging ideas with their peers.
Silence		11. Helpless confusion		Continuous noise or confusion deviating from class. Observers cannot judge the quietness or confusions, using this category.
		12. Beneficial silence		Silence or quietness for successful teaching. Maybe thinking problems or doing exercises of students.
Technology		13. Teacher manipulating technology		Using multimedia technology to present teaching content and explanations, like showing a PowerPoint for a long time.
		14. Student manipulating technology		Using technology for presentations.

Note. Adapted from Flanders, N. A. (1963). Intent, action and feedback: A preparation for teaching. Journal of Teacher Education, 14(3), 255. Copyright 1963, The American Association of Colleges for Teacher Education and Fang, H. G. (2012). Improved Flanders Interaction Analysis System and its application. China Educational Technology, 10, 110. Copyright 2012, The National Center for Educational Technology.

According to the iFIAS code table (Table 1), the researchers coded the speech acts in the demonstration class. The observers wrote down the category number of the interactions they observed and recorded these numbers in sequence in a column every 3 s. In this study, a total of 885 original codes were obtained. Then, the two adjacent original codes were combined into one coding sequence pair, and each code formed a set of code sequence pairs with the previous and the last code respectively; thus, 885 original codes formed 884 code sequence pairs. Finally, the 884 encoded sequence pairs were formed by matching the original code and filled into the matrix for subsequent analysis and interpretation.

Teachers’ language included direct influence and indirect influence. The teacher’s speech and behavior represented by codes 1–4 (Table 1), which is the indirect influence, consisted of the acceptance of feelings, praise or encouragement, acceptance or using ideas of the students and asking questions. While codes 5–7 constituted the direct influence, including teacher’s lecturing, giving directions, criticizing or justifying authority.

For the interactions between the teacher and students with disabilities, the conversation interactions were classified first into topics that were academic or nonacademic. The academic interactions were then categorized under one of the three levels of cognitive extension: comprehension monitoring, partial cognitive extension, and full cognitive extension. The nonacademic interactions were categorized according to whether the teacher initiated organizational, management, status, or personal questions and statements (Jordan and Stanovich 2001, Jordan et al. 1997). Combining the classification with the coded themes of iFIAS, the coded themes of 1 and 2 were secondary coded into comprehension monitoring, the coded themes of 3–6 and 8–9 were secondary coded into cognitive extension-partial and full, and the coded theme of 7 was secondary coded into nonacademic interaction.

2.4. Analytic approach

FIAS is a reliable instrument that is descriptive rather than evaluative for analyzing teacher-student interactions (Campbell and Barnes 1969). In the analysis process of this research, two observers divided the teaching process of the classroom into different teaching links in accordance with the order of the teaching time and described the activities of the teacher and students, which was the coding process. Then, the observers combined the codings into a coordinate for every two digits to form the coordinate, and frequencies of 884 encoded sequence pairs were entered into the suited cells by the researchers. Coding and matrix can quickly discover the characteristics of speech acts in teaching. For the data in the matrix table, some frequency analysis can be performed or the ratio of the data of two items can be calculated. The matrix of classroom interactions between the teacher and students with DD (Table 2) was obtained. The row and column within the matrix both represent the coding sequence number. As shown in Table 2, the frequency of first category named accepts feeling was 36, the frequency of first category named praises or encourages was 94, and the rest were in a similar fashion.

Table 2.

The matrix of classroom interactions.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	Total
1	3	6		6	5	4		3	4		1	4			36
2	4	42			17	5	2	4	6		1	10	1	2	94
3		1													1
4	2	2		6	7	7	1	4	5			12			46
5	2	13		7	57	26	1	18	5		2	18	1	4	154
6	8	11		4	19	64	4	5	6		5	44	1	1	172
7	1	3			1	5	14		1		1	3		1	30
8		4		7	7	7		11	5			5	2		48
9	1			8	6	9		4	10			8	1		47
10												1			1
11	2	1		1	1	3		1			2	2			13
12	11	10	1	8	19	40	7	8	4	1		109			218
13	1	1			1			1				2			6
14	1				3	2	1				1			10	18
Total	36	94	1	47	143	172	30	59	46	1	13	218	6	18	884

Note. Rows 1–14 and columns 1–14 represent the 14-code theme of improved Flanders Interaction Analysis System shown in Table 1. Zone A is the lightest gray area, Zone B is the lighter gray area and Zone C is the darkest gray area.

Besides, certain areas in the matrix were useful for describing specific verbal behaviors of the teacher and students. In the matrix of iFIAS (Table 2), the area where the rows 1–3 intersect the columns 1–3 was named the active integration grid and denoted as Zone A. Rows 1–3 were positive categories in the teacher-student interactions, including teachers receiving emotions, teachers praising or encouraging teachers, and teachers adopting students’ opinions; therefore, the area where rows 1–3 intersect columns 1–3 was positive, and marked as Zone A (Flanders 1970). The greater data located in this area, the more harmonious the relationship between the teacher and students. Similarly, the codes of rows 6–7 were teacher instruction, teacher criticism or maintenance of teacher authority, which were relatively negative, thus, the area where the rows 6–7 intersect with the columns 7–8 was defined as a defect grid, and designed as Zone B. Besides, due to category 4 referring to asking questions and category 5 lecture, the cross region formed by the intersection of columns 1–10 and rows 4–5, columns 4–5 and rows 1–10 was named the content cross region (denoted as Zone C), which was the ratio of the teacher’s explanation and teaching behavior to the whole teaching behavior (Flanders 1970). The larger data located in the zone indicated the gap between the teacher and the students with DD. The data analysis intensity of the active integration grid and defect grid could reflect the emotional atmosphere in the class (Wu et al. 2016). Furthermore, we analyzed the classroom structure and style through matrix analysis. Each of the categories and variables of the classroom interactions could be calculated by the formula in Table 3 (Flanders 1970, Gao 2007); specifically (1) teacher talk ratio refers to the percentage of teachers’ speech time during the teaching period, (2) student talk ratio denotes the proportion of students speaking time during the teaching period, (3) silence ratio means the percentage of silence time during the class, (4) technology applied ratio refers to the proportion of using technology during the class, (5) active integration area-to-defect area ratio means the ratio of the teacher’s use of indirectly affected speech time to the use of directly affected speech time, calculated by the ratio between Zone A and Zone B, and (6) Content cross ratio refers to the proportion of the teacher’s explanation during the class, calculated by the ratio of Zone C to the total.

Table 3.

The ratio statistics of interaction behavior in classroom teaching.

Variable	Formula	Ratio (%)
1. Teacher talk ratio	$\left[{\sum }_{i=1}^7\mathit{Row}\left(i\right)\right]\mathrm{*}100/\mathit{Total}$	59.16
1.1 Indirect-to-direct influence ratio	$\left[{\sum }_{i=1}^4\mathit{Row}\left(i\right)\right]\mathrm{*}100/{\sum }_{i=5}^7\mathit{Row}\left(i\right)$	51.59
1.2 Positive-to-negative influence ratio	$\left[{\sum }_{i=1}^3\mathit{Row}\left(i\right)\right]\mathrm{*}100/{\sum }_{i=6}^7\mathit{Row}\left(i\right)$	64.85
1.3 Teacher questioning ratio	$\mathit{Row}(4)\mathrm{*}100/{\sum }_4^5\mathit{Row}(i)$	24.74
1.3.1 Asking opened-ended question ratio	$\mathit{Row}(4.1)/\mathit{Row}(4)$	65.96
1.3.2 Asking closed-ended question ratio	$\mathit{Row}(4.2)/\mathit{Row}(4)$	34.04
2. Student talk ratio	$\left[{\sum }_{i=8}^{10}\mathit{Row}\left(i\right)\right]\mathrm{*}100/\mathit{Total}$	11.99
2.1 Student passive response ratio	Row(8)*100/Total	6.67
2.2 Student active response ratio	$\mathit{Row}(9)\mathrm{*}100/\mathit{Total}$	5.20
2.3 Student peer discussion ratio	$\mathit{Row}(10)\mathrm{*}100/\mathit{Total}$	0.11
2.4 Student initiative speech ratio	$\left[{\sum }_{i=9}^{10}\mathit{Row}\left(i\right))\right]\mathrm{*}100/{\sum }_{i=8}^{10}\mathit{Row}\left(i\right)$	44.34
3. Silence ratio	$\left[{\sum }_{i=11}^{12}\mathit{Row}\left(i\right)\right]\mathrm{*}100/\mathit{Total}$	26.13
3.1 Confusion ratio	$\mathit{Row}(11)\mathrm{*}100/\mathit{Total}$	1.47
3.2 Benefit silence ratio	$\mathit{Row}(12)\mathrm{*}100/\mathit{Total}$	24.66
4. Technology applied ratio	$\left[{\sum }_{i=13}^{14}\mathit{Row}\left(i\right)\right]\mathrm{*}100/\mathit{Total}$	2.71
5. Active integration area-to-defect area ratio	Zone A / Zone B	215
6. Content cross ratio	Zone C / Total	30.77

Note. For example, Zone A/Zone B = the total frequency in Zone A/the total frequency in Zone B = 56/26 = 2.15 = 215%.

In addition, iFIAS proposes that teachers’ questions can be divided into open-ended and close-ended, and Question and Answer Model (QAM) and Innovative Inquiry Model (IIM) as well. QAM in the classroom was used to express the situation that the teacher encouraged the students with DD to answer close-ended questions, which included the categories and combination of the teacher asking close-ended questions (category 4) and students responding passively (category 8), represented by the cells (4, 4),² (4, 8), (8, 4), (8, 8). The QAM reflected the degree of training-based questioning. IIM reflected the extent of the teachers’ innovative questioning, adopting and accepting the viewpoints of students with DD, and guiding students to have initiative talk, including the categories and combination of students active response (category 9), accepting or using the ideas of students (category 3), students passively answering and being accepted by the teacher (the combination of category 8 and category 3), teachers asking close-end questions and students actively responding (the combination of category 4 and category 9), students responding and raising their hand to answer (the combination of category 8 and category 9), the teacher asking questions and using the ideas of students (the combination of category 4 and category 3), are represented by the cells consisting (9, 9), (9, 3), (3, 3), (3, 9), (8, 3), (4, 9), (8, 9), (4, 3).

To assess interobserver agreement (IOA), two independent observer encoded all facts. All records were compared, and each record was coded as an agreement (both observers concurred on correct or incorrect) and disagreement. Then, the IOAs were calculated by dividing the total number of agreements by agreements plus disagreements and converting the quotient into a percentage. The IOA of the two observers’ video coding records was 93.33%.

3. Results

3.1. Classroom structure and emotional atmosphere of the class

The iFIAS divided the classroom interactions into four dimensions that contained teacher talk, student talk, silence and technology. The frequency of the four types of behavior shown in Table 3 were 59.16%, 11.99%, 26.13%, and 2.71% respectively, which could indicate the structure of the classroom. The ratio of the teachers’ and students’ verbal behaviors in this classroom was about 4.93:1 (59.16%:11.99%), indicating that the teacher played the dominant role in the classroom. In addition, the proportion of classroom silence was 26.13%, which contributed to 24.66% of teaching, much higher than the confusion proportion (1.47%). Combined with the video observations, it was found that the teacher spent much time patiently waiting for the students with DD to answer questions in the class. At length, the technology application frequency was 2.71% and the time of the high ratio of technology application was mainly in creating a situation by the teacher.

As shown in Table 3, the frequency of the active integration grid was 56 while the defect grid was 26, revealing more praise or encouragement when the teacher interacted with her students. The ratio of the active integration grid to the defect grid was 2.15:1. which indicated that the teacher-student relationship in this class was harmonious to some extent.

3.2. Teaching style and questioning techniques of the teacher

3.2.1. Teaching style of the teacher

The ratio of indirect to direct ratio and positive to negative influences on the students with DD could be used to analyze the teacher’s teaching style. According to the coding, the frequency of the teacher’s indirect and direct language was 51.59%, which meant that the teacher’s language in the class took half of the time and gave many chances for the students to respond. The teacher always asked the students questions, encouraged them to answer, and accepted or praised the responses of students with DD.

Moreover, the proportion of the teachers’ positive and negative impacts on the students with DD was 64.85%. The negative behavior of the teacher was mainly raised when the students left their seats and went to the front of the blackboard without her permission and other irregularities. And the positive behavior of the teacher included patiently listening to the students, and timely approving and praising those with DD.

3.2.2. Questioning techniques of the teacher

Questioning techniques of the teacher could be analyzed by the teacher’s questioning ratio, including the ratio of open-ended questioning and close-ended questioning and the ratio of QAM and IIM.

As shown in Table 3, the frequency of asking questions of the teacher in the teaching process was 24.74%. And the proportion of open-ended questions and close-ended questions that the teacher used were 65.96% and 34.04% respectively.

Furthermore, we analyzed the pattern of questions. Table 4 shown the cell (8, 8) appeared the most frequent in the QAM with the total number of 11 times, indicating that the students had more consecutive responses. The next were teacher questioned-student answered and teacher continuous asking, respectively appearing six times in the cells (4, 4) and (4, 8) seven times. For ‘the teacher continued to ask questions for the responses of students’, this was reflected by cell (8-4) the least (four times). The above illustrates the teacher-led questioning session in this class, showing a typical question and answer mode in China’s special classes (Q&A, Li 2014, Wu 2017). And in the IIM (Table 5), only the continuous speech and active answering questions of students with DD appeared in the cells (9, 9) and (4, 9) respectively, and there were no other sequence pairs representing the continuous adoption of the student’s viewpoints by the teacher. Overall, the IIM was relatively single in the classroom, indicating that the students with DD had less doubt and lacked creativity (Gao 2007).

Table 4.

Data on the degree of QAM.

	Question and answer mode (QAM)
Sequence pair	(4, 4)	(4, 8)	(8, 4)	(8, 8)
Frequency	6	7	4	11

Table 5.

Data on the degree of IIM.

	Innovative inquiry mode (IIM)
Sequence pair	(9, 9)	(3, 3)	(3, 9)	(8, 3)	(8, 9)	(9, 3)	(4, 9)	(4, 3)
Frequency	10	0	0	0	4	0	8	0

3.3. Features of verbal behaviors of students with DD

This study classified students’ speech behaviors into three categories: passive response, initiative speech and peer discussion. The characteristics of students’ verbal behaviors were classified according to the percentage of various verbal behaviors and the ratio of initiative speech behaviors of students with DD. Table 3 showed that the percentages of the three verbal behaviors of students with DD were 6.67%, 5.20% and 0.11% respectively. The percentage of passive responses and initiative speeches of the students were similar; the percentage of student peer discussion was much lower than those, which showed the students with DD interacted a lot with the teacher while rarely interacted with their peers. Notably, there were no student initiatives to ask questions in the class, and student-to-student communication was only once.

3.4. The patterns of teacher-student interaction in different groups

Even in the special class with six students, the teacher divided the students with DD into three groups based on their cognitive competence. As shown in Table 6, the overall frequency of teacher-student interactions in the three student groups was similar, while the results yielded a pattern of differences in the three groups related with students’ learning abilities. As the column for non-academic interaction shows, the teacher used a higher frequency of non-academic interactions for the students in group C than those in groups A and B. For academic interactions, the teacher tended to use more comprehension monitoring for students in group B than those in groups A and C, while using more cognitive extension-partial and full for students in group A than those in groups B and C.

Table 6.

Frequency by student group of teacher-student interactions.

Student group	n	Non-academic interactions	Academic interactions		Total no. of interactions
			Comprehension monitoring	Cognitive extension-partial and full
Group A	2	2	30	115	147
Group B	2	1	44	97	142
Group C	2	18	30	82	130

4. Discussion

In this paper, we have focused on the interactions of speech acts between the teacher and students with DD in a special school in China, aiming to analyze the characteristics of the interactions between the teacher and students with DD and figure out whether there were differences of frequency and patterns between the teacher and students in the three groups with different learning abilities. Therefore, we have organized discussions from three aspects of teacher classroom speech acts, student classroom speech acts and teacher-student interactions in different groups.

4.1. Teacher classroom speech acts

First, the teacher dominated the demonstration class, but the core of instruction was still student-orientated. The teacher’s language ratio was 59.16%, which was much higher than the student’s language ratio (11.99%), indicating that the teacher played a dominant role in the classroom. Meanwhile, the ratio of silence (26.13%), indirect and direct language (51.59%), and positive to negative influence (64.85%) all showed that the teacher provided adequate time for students to think and inspired them to respond positively. This is similar with other scholars’ findings that teachers played the control role in the classes in general and special schools (Zhao 2010, Wu 2017). This might be influenced by the traditional teaching style embedded in traditional collectivist culture and Confucianism. Chinese students usually obey the authority of their teachers forming the teacher-centered learning environment (Cheung 2012).

Although the rhythm of the class was controlled by the teacher, the students were still the center of the classroom. Many scholars consider learner-centered teaching can promote classroom practices effectively which has been echoed in many cases (Dole et al. 2016, Opdenakker and Damme 2006). Learner-centered classrooms can improve teachers’ teaching effect and students’ active participation (Chen 2019, Zhu et al. 2018). While Cheung (2017) found it was a more effective teaching practice that provided a balance between teacher-directed and student-centered. For the classes of special schools, teachers should play the guiding role and consider students’ physical and mental characteristics at the same time. In inclusive education classrooms, teachers need to pay more attention to students with DD and give them more academic and behavioral guidance (Cameron 2014, Kemp and Carter 2002, Nelson and Roberts 2000). Consistent with the teacher’s dominated role and previous research studies (e.g. Avramidis and Norwich 2002, Kemp and Carter 2002), teachers strongly believed that students with disabilities required more attention.

Secondly, the teacher gave the students timely feedback. As shown in Table 2, there were many evaluations from the teacher. She gave timely verbal affirmations and reinforcement to support students’ good behaviors such as answering questions actively and completing tasks according to requirements. That was in line with relevant research studies which have pointed out that in the management of students, teachers need to design appropriate and clear rules about classroom behavior and reinforcements for good behavior or give negative effects for bad behavior (Evertson and Emmer 1982, Stage and Quiroz 1997). The large proportion of disapproval in classrooms will not provide an effective learning environment (Thompson et al. 1982), while a positive environment where frequent praise is given for appropriate behavior is much more conducive to student achievements (Brophy and Good 1970, Brophy and Evertson 1976). Kurt and Tekin-Iftar (2008) also found that timely feedback can improve teaching efficiency. Moreover, many studies have pointed out that Positive Behavior Support (PBS) should be provided to students (Hieneman et al. 2005, Horner et al. 1990, Mesibov et al. 2002, Morris and Horner 2009), especially for children with DD (Spooner et al. 2012).

Thirdly, the teacher used questioning skills frequently, especially open-ended questions. In this study, the teacher’s questioning ratio was 24.74%. She asked the students with DD many times questions related to the main teaching concepts of dried seaweed, potato chips and haw flakes, which aimed to extend the students’ learning content. Questioning is a good way to focus students’ attention on a particular concept and check whether they have mastered certain knowledge that can be used as a means of review (Brown 2014, Cooper and Irizarry 2014). The types of questions teachers ask during classroom activities can influence students’opportunities for incorporating new information and developing new meanings in teaching activities (Hackenberg 2005, Harrop and Swinson, 2003). The percentage of the teacher’s questioning in this study was similar with that of in general classes (26%) (Li 2014), while it was lower than the questioning ratio of 32.56% in a classroom with blind students in China (Zhu et al. 2018). Many foreign studies have shown that in classrooms, 50%−80% of teachers’ questions are close-ended, and 20% of teachers’ questions are able to lead students to thinking deeper (Black 2001, Gall 1984, Kawanaka and Stigler 1999). And some studies point out that question-asking is a skill that can be learned better by doing and experiencing (Korkmaz and Yesil 2010, Martin and Cuban 2005).

The reason for the low proportion of questioning in the class was there were 65.96% open-ended questions which needed much time for the students with DD to think and respond. Further analysis of the content of the open-ended questions revealed that in this study, they mainly included food which the students wanted to eat and buy, which indicated that the teacher gave the students with DD more opportunities for self-determination. Many studies show that students with DD have low levels of self-determination and lack knowledge and skills of self-determination (Carter et al. 2009, Shogren et al. 2016). While self-determination is of great significance to students with DD, which can promote people’s mental health, affect transition and help them to better socially integrate (Agran et al. 2002, Shogren et al. 2015, Wehmeyer 1997). The participant teacher in the research was of the awareness to give students with DD more opportunities for self-determination.

4.2. Student classroom speech acts

First, the results of this study show the importance of developing learning activities with the discipline of operability and practicality for students with DD. We observed that the time taught by the teacher in the lesson was about 10 min, while the time for students to read and write the exercises was about twice the teaching time. It may be related to children with DD having less attention and working memory than ordinary children (Deutsch et al. 2008, Jarrold et al. 1999). Besides, students with DD have slower perception and information processing skills than general students (Nugent and Mosley 1987, Stanovich 1978). Matrix analysis showed that the students’ passive responses were slightly higher than the active responses.

In addition, the study shows there were few peer interactions in the class, which indicated a lack of communication and counseling between peers in the classroom. Students in the peer tutoring condition performed better than students receiving traditional instruction (McDuffie et al. 2009). This may result from the teacher dividing the students into homogeneous groups in the class, so they could not help each other since they were at the same learning level.

4.3. Teacher-student interaction in different groups

The results in this study show that the total frequency of interactions between the teacher and students in different groups to be relatively similar, including 147 times with group A, 142 times for group B, and 130 times for group C. This finding is close to previously reported similarities. Owing to most of students with DD being educated in inclusive settings in Western countries, many studies on teacher-student interactions have been concentrated in inclusive education classrooms. Cameron (2014) found that teachers interacted similarly in frequency with the students who had varying degrees of disability.

In addition, the patterns of teacher-student interactions in this study showed the following two characteristics. Firstly, the teacher used higher frequency of non-academic interactions for students in group C than those in groups A and B. Relevant studies in inclusive education classrooms show more non-academic interactions between teachers and students with DD than other students, because they are more prone to distraction, which require more warnings and criticism (Cook, 2001). In this study, the students in group C had more non-academic interactions with the teacher because of their low learning ability and frequent behavior problems, which needed the teacher to remind them about their attention and behaviors. Secondly, the teacher tended to have more cognitive extension-partial and full interactions with the students in group A than the other two groups. Some studies have pointed out that teachers will be more inclined to engage in more academic interactions with students with stronger learning abilities (Jordan et al. 1997, Schulte et al. 1990, Wallace et al. 2003). This would be consistent with the result of Jordan et al. (1997) that there are more students with excellent performance in the interaction of cognitive extension.

5. Conclusion, limitation, and implications

5.1. Conclusion

Overall, this study posited the characteristics and patterns of interactions between the teacher and students with DD in a special school in China. Firstly, the teacher played the dominant role in the class while instruction was student-oriented, using lots of evidence-based strategies to motivate students’ interest and to participate in activities, including questioning, timely feedback, reinforcement, game teaching, and students’ self-determination. However, the homogeneous grouping in the class led to the students only interacting with the teacher and few peer interactions were apparent in class. Secondly, a striking feature of the interaction patterns between the teacher and students with DD of different learning abilities were more non-academic between the teacher with students in group C than others, and higher frequent cognitive extension-partial and full interactions between the teacher with students in group A than the other two groups.

5.2. Limitations

Some limitations remain in this study. Firstly, it is a case study, and all the participants were from one classroom at one school in **, while the development of special education in China is uneven. Thus, it is not suitable to generalize teacher-student interactions in special schools within the whole country. Secondly, the analyzed class in the study was conducted by one teacher, which makes it difficult to explore the personal factors of teachers affecting teacher-student interactions in special schools. Further work with more teachers is needed, which may contribute to special education teachers’ training in practice. Thirdly, due to the specific education placement in the study, the differences of general teachers’ and special education teachers’ interaction with DD is not analyzed, which is of great importance for the education quality for students with DD. Future attention should be paid in that topic.

5.3. Implications

The present research has important implications about segregated education placement for students with DD in the Chinese context in terms of effective teaching.

Firstly, teachers should appropriately use open-ended questions, which can trigger active thinking, questioning and dialog among students with DD. Proper questioning not only can assist in focusing students’ attention, but also can be used as a good way to check and review knowledge points (Cooper and Irizarry 2014). Secondly, teachers need to increase the proportion of positive feedback and initiation and decrease neutral and disapproving feedback. Praise and encouragement from teachers towards students’ appropriate behavior, accepting students’ emotions as much as possible and affirming students’ opinions, will help to establish good teacher-student relationships, and timely feedback will improve teaching efficiency (Kurt and Tekin-Iftar 2008). In addition, teachers should balance the frequency of interactions with different groups of students, and a better balance between academic and procedural activities should be achieved. Finally, teachers should focus more on teaching independent work habits to students so that the disproportionate amount of procedural time that is now spent in classrooms can be reduced (Thompson et al. 1982).

Funding Statement

This work was supported by International Joint Research Project of Education, Beijing Normal University grant ICER201904.

Disclosure statement

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