Reflective Teaching And Metacognition In Romanian Teachers’ Practice. A Conceptual Delimitation

Reflective thinking

In the conceptualization of reflective thinking, researchers have highlighted the cyclicity of the process. It is the case of the theoretical models proposed by Kolb and Gibbs. For instance, the Experiential Learning Cycle developed by Kolb (1984) comprises four stages through which only an efficient learner could pass. In his vision, it is not frequent that people could arrive to gain mastery over the entire cycle (Kamis & Khan, 2019). Concrete experience is the first stage corresponding to the assigning of the task. The following stage, reflective observation, depicts the student reviewing the experience so far. The third stage refers to abstract conceptualisation, the student trying to understand the experience. The last stage is the active experimentation in which the student practices what it was learnt.

The other approach that reveals the cyclicity of the process is Gibbs' model of reflection (1988) illustrates a behavioural pattern that may be summarized as a previous reflection-based action. Gibbs' reflective cycle encompasses six components: description- the students retrace the details of the situation; feelings- the student determines the emotions and thoughts related to the events; evaluation- the students give positive or negative attributes to the experience; analysis- the student assesses the sense of the situation; conclusion: the student reviews other solutions for the event; the action- envisioning the plan for a future similar situation.

A different approach comes from Schön (1983), who delimitates between two types of reflection, namely reflection-in-action and reflection-on-action. Schön (1991) also states that experimentation, reflection and action are cyclic. However, in his view reflection does not occur only post-event. The reflection-in-action component stresses the continuity of the process, the previous events gathered knowledge animating and leading the path to new actions. The second level of the process, reflection-on-action, addresses the post-experience reasoning about the acquired knowledge and how it could be used in a future learning experience (Sellars, 2017).

Metacognition

Metacognition has been defined as cognition about one’s own cognition (Flavell, 1979) or the ability to actively control a variety of cognitive processes (Pintrich, 2000; Tobias & Everson, 2009; Moss, 2007). Flavell, one of the first researchers who focused on metacognition, argues that the ability to control the cognitive processes involved in learning is related to the metacognitive knowledge, experiences and skills that students possess. These three dimensions are essential to manage different cognitive processes. Metacognitive knowledge can be defined as beliefs about tasks, strategies and goals, metacognitive experiences comprise affective experiences related to cognitive processes and, lastly, metacognitive skills are the strategies used for controlling cognitive processes (Flavell, 1979; Desoete & Ozsoy, 2009).

In addition to the perspective developed by Flavell (1979), another perspective of metacognition is the one developed by Brown (1987). According to Brown’s perspective, metacognition consists of two dimensions: knowledge of cognition and regulation of cognition. Knowledge of cognition is defined as how students understand their declarative, procedural and conditional knowledge, while regulation of cognition involves strategies such as planning, evaluating and monitoring (Brown, 1987; Jacobs & Paris, 1987; Schraw & Moshman, 1995; Lee, Teo & Bergin, 2009; Spada, Georgiou & Wells, 2010). A student who knows that learning is easier when there is interest for the subject being studied, who becomes aware when he/she understands something and who uses different learning strategies according to the situation, might be described as a student who possesses knowledge about his/her declarative, procedural and conditional knowledge. Further, a student who thinks of several ways to solve a problem and who checks his/her work can be described as a student who regulates his/her cognition.

The first dimension of the model developed by Brown (1987) emphasizes the important role of declarative, procedural and conditional metacognitive knowledge. Researchers (Schraw, 1998; Schmitt, 2005; Harris et al., 2009) define declarative knowledge as the knowledge that a student holds about himself/herself as a learner (e.g. strengths, weakness, interests), about tasks and task-relevant strategies and about the factors that influence his/her performance. Procedural knowledge has been defined as knowledge about one’s cognitive processes, unlike declarative knowledge which emphasizes knowledge about one’s cognition (Misailidi, 2010). Procedural knowledge involves knowledge about how to apply different strategies to control the cognitive processes involved in learning. Declarative metacognitive knowledge can be described as the “knowing that”, while procedural metacognitive knowledge might be described as the “knowing how”. In addition to these two dimensions of metacognitive knowledge, Brown (1987) included a third dimension that deals with “knowing why and when” knowledge. The third dimension, conditional metacognitive knowledge, refers to students’ ability to explain their decisions concerning their memory actions (Schneider & Lockl, 2004). More specifically, conditional knowledge involves information about why and when to use declarative and procedural knowledge, helping students to efficiently allocate their resources according to specific tasks (Schraw, 1998).

The second dimension of Brown’s model, the regulation of cognition, has been subdivided into two components. The first component is the monitoring of cognitive processes and involves knowledge about the effectiveness of strategies (“is this strategy efficient?”). The second component, the ability to regulate cognition to improve effectiveness refers to the strategies employed for enhancing the efficacy of the learning strategies (Williams & Atkins, 2009). These two functions of metacognition - monitoring and control - are also found in Flavell’s model. The three dimensions of metacognition considered essential in monitoring different cognitive processes (metacognitive knowledge, experiences and skills) are believed to influence the student’s learning through two functions: monitoring and control. The monitoring function expresses itself through metacognitive knowledge and metacognitive experiences. In contrast with the monitoring function, the control one is represented by metacognitive skills (Efklides, 2006).

An integrative approach is proposed by Zimmerman (2000) through his Process-Oriented Model of Self-regulated Learning. The three phases of this socio-cognitive cyclic model comprise: 1) forethought, referring to the goals, planning of the task and self-motivation beliefs; 2) performance, focusing on self-control and self-monitoring; 3) self-reflection involves a criteria-based evaluation of the inquiry task in the search for developing methods of obtaining better outcomes.

Participants and sampling method

The sample of the study has 258 pre-service and in-service teachers who attended university classes at Babes-Bolyai University in Cluj, during the 2^nd semester. Participants were chosen from the training course in which they were enrolled, referred to earlier on, and were asked to answer to a questionnaire. Considering that courses at the university level were held online during the 2^nd semester of the year 2019-2020, teachers were asked to fill in the questionnaire during an online evaluation.

Instruments

The instruments used to gather the information were two scales that have already been used in the literature and found to be effective in assessing both metacognitive skills and self-reflection skills in teachers. The first one, the Reflective Thinking for Problem Solving Scale was developed by Kizilkaya and Askar in 2009 and has 14 items grouped in three sub-scales: questioning (5 items), reasoning (4 items) and evaluating (5 items). The items are measured on a 5-point Likert scale: 1=Never, 2=Rarely, 3=Sometimes, 4=Usually and 5=Always and has an internal consistency of 0.84. Considering validity, Kizilkaya and Askar have run a confirmatory factor analysis and got the following results: CFI= 0.95, RMSR= 0.08, RMSEA= 0.071. It seems that the Reflective Thinking for Problem Solving Scale is an adequate instrument through which we can measure the reflective thinking skills of pre-service and in-service teachers. About the second scale applied in this research, we considered that the Teacher Metacognition Inventory would be appropriate to assess the metacognitive skills of teachers. The Teacher Metacognition Inventory was developed by Jiang, Ma and Gao in 2016 and the final version have 28 items measured on a 5-point Likert scale (1-Strongly disagree, 5-Strongly agree). The inventory has six sub-scales grouped into (1) Teacher metacognitive experience, (2) Metacognitive knowledge about pedagogy, (3) Teacher metacognitive reflection, (4) Metacognitive knowledge about self, (5) Teacher metacognitive planning and (6) Teacher metacognitive monitoring. For the present study, we have used items of the following subscales: teacher metacognitive experience, metacognitive knowledge about the self, teacher metacognitive planning, and teacher metacognitive monitoring. To check for the validity and reliability of the scale, Jiang, Ma and Gao run a series of analysis (item discrimination index, exploratory factor analysis, convergent validity analysis, parallel analysis, and reliability analysis) which shows that the Teacher Metacognition Inventory is well suited to be used in our research. Demographic information was gathered through the demographic form which has a series of questions developed by researchers and comprises gender, teaching degree, teaching experience, and the level of teaching.

Demographic characteristics of teachers

First, we have run the descriptive statistics analysis to examine the demographic background of our participants. Nearly all of our participants are female teachers (97%), with no didactic degree (73%) and with less than 1 year of experience. The demographic characteristics of the participants are illustrated in Table 01 .

Teacher’s responses to the frequency of using reflective thinking skills

The first question of the research aimed to identify to what extent teachers use reflective thinking skills in their practice. The descriptive statistics presented in Table 2 show how frequent teachers use reflective thinking skills in problem-solving. Responses have been grouped in 3 categories: rarely / never, sometimes, and always / most of the time. It seems that participants use quite frequently reflective thinking skills in problem-solving. Most of the participants take into consideration the methods they use when solving a problem (94.20%) and focus on the needed information to solve a problem (95.3%). When failing to solve a problem, 88% of teachers ask themselves questions to understand why they were not able to solve it. Furthermore, questions are also addressed to better understand the information given in the text of a problem (87.2%). The lowest percentages were for the items of the evaluation sub-dimension of the scale: "I question the solutions of my colleagues, and try to find better ones" (55.8%) and "After completing a problem, I make comparisons between the solutions of my colleagues, and assess my own solution accordingly" (47.2%).

Table 02 clearly illustrates the high percentages of teachers who employ self-reflective thinking skills in problem-solving. Besides the percentages presented in Table 02 , we continued with the analysis and determine the means, standard deviations, and the minimum and maximum scores for each of the three subscales of the Reflective Thinking for Problem Solving Scale. These are illustrated in Table 03 presented below.

The minimum scores of Questioning, Evaluation and Causation range between 4-6 points and the maximum between 20-25. The Questioning sub-scale has a mean of X̄ =21,21 (SD=3.15) while evaluation has one of X̄ =19.31 (SD=3.46) and Causation one of X̄ =16.60 (SD=2.77). Considering that the maximum score is 25, we can see that the mean of the Questioning sub-scale is the closest to the maximum score. This means that pre-service and in-service teachers use their questioning in problem-solving. The lowest mean is the one for the Causation dimension but it is still a high one considering that the minimum score for this dimension is 4. Concerning the means presented above we can conclude that student teachers use their reflective thinking skills in problem-solving quite frequently.

Participants’responses to the items measuring metacognition

Besides reflective thinking skills in problem-solving, metacognitive skills are also used to a high degree by teachers who participated in the study. Participant’s responses for this variable are illustrated in Table 04 .

As can be easily observed, high percentages of participants seem to feel relaxed when successfully completed a classroom teaching task (98.10%), set specific teaching goals for lessons (86.4%), are well aware of their weaknesses in teaching (82.5%) and check their teaching progress periodically to figure out whether it meets their expectations (77.5%). The means of the four dimensions of The Teacher Metacognition Inventory are represented in Table 05 . All the means of the subscales are very high, being close to the maximum score of 5. The mean of the metacognitive experience dimension is X̄ =4.53 (SD=0.61), of the knowledge about self is X̄ =4.20 (SD=0.85). For the metacognitive planning sub-scale the mean is X̄ =4.07 (SD=0.68) and for metacognitive monitoring is X̄ =4.03 (SD=0.74). These results show that students teachers who participated in our study use metacognitive skills in teaching.

Relationships between thinking skills and metacognitive skills

To examine the relationships between the two variables introduced in the study we have run a correlation analysis and presented the results in Table 06 . The Pearson correlation coefficient indicates a positive and moderate correlation between reflective thinking skills and metacognitive skills: r(258)=0.446 p<0.01. This means that as metacognitive thinking skills increase so do the reflective thinking skills.

Since the correlation between reflective thinking skills and metacognitive skills has a moderate intensity, we tested to see if there are differences in metacognitive skills in students with low, medium, and high levels of reflective thinking skills. We assumed that there are significant differences between these two variables and run an ANOVA analysis. Results are presented in Table 7 and indicate the existence of significant differences between our groups of interest, with F value of F(2, 258)=31.675 (p<0.001). To identify which groups are different (low, moderate and high levels of reflective thinking skills) we have used the Tukey post-hoc test (Table 08 ). Tukey’s test indicates that student teachers with medium reflective thinking skill differ from both those with lower (p=0.17) and higher levels (p=0.000). Interesting, the differences between those with low and those with high levels of reflective thinking skills are not significant (p=1.0).