Instructional clarity in physics lessons: Students’ motivation and self-confidence

Abstract This study examines the relationship between instructional clarity in physics lessons and students’ motivation and self-confidence to learn physics. We conducted a secondary analysis of data from the Trends in International Mathematics and Science Study (TIMSS) 2019. Using data from six countries, whose students’ science achievement scores were higher than the TIMSS 2019 scale centrepoint, this study explores the relevance of instructional clarity in physics lessons to students’ motivation and self-confidence to learn physics by using confirmatory factor analysis and latent multi-group structural equation modelling. We identify a statistically significant positive effect of instructional clarity on students’ motivation to learn physics. The findings corroborate expectancy value theory by establishing the mediating role of self-confidence in learning physics in the relationship between instructional clarity in learning physics and motivation to learn physics. Our findings support cognitive load theory by revealing the statistically significant effect of giving explanations in physics lessons on instructional clarity in learning physics. Physics teachers can reduce the heavy intrinsic load associated with the physics subject and increase instructional clarity in physics lessons by giving explanations and using signalling and redundancy.


Introduction
People in the 21st century live in an environment characterised by rapid technological changes.Science is the foundation of technology, so such technological advancements are dependent on scientific knowledge.However, students' motivation to learn science subjects remains low and ABOUT THE AUTHOR Palmira Pečiuliauskienė, is a professor at the Education Academy, Vytautas Magnus University of Lithuania.The research area is didactics of science education.Her research interests include motivation for learning science, innovation in science education, and digital technology in science education.During the last five years, the professor and her co-authors published two monographs: Science Teachers' Innovative Work Behavior: Factors and Actors; Motivation of New Generation Students for Learning Physics and Mathematics.This study is a permanent continuation of previously conducted research on the internal (personal) and external (technological) factors of students' motivation to learn science.This study is contributing to the existing literature about students' motivation to learn science and provide a basis for new research by identifying the multiple mediating roles of cognitive appraisals (i.e., science self-concept, self-efficacy, and value) between teaching and learning characteristics.
poses a risk to future technological progress (Astalini et al., 2019;Haverly et al., 2022;Lavonen et al., 2021;Wilson et al., 2020).Scholars have argued that students' motivation to learn physics and mathematics decreases particularly rapidly in middle school compared with their motivation to learn other subjects during the same period (Alexander et al., 2019;Krapp & Prenzel, 2011;Potvin & Hasni, 2014;Stankov et al., 2012).Furthermore, students' motivation to learn science decreases faster than that in other subject areas during the middle school years (Steidtmann et al., 2022;Woods McConney et al., 2013).
A decline in students' academic motivation has been linked to a decline in the quality of teachers' instructional behaviours (Alles et al., 2017;Maulana et al., 2016;Titsworth et al., 2015).These behaviours can be examined from various perspectives.Research on teachers' instructional behaviours has demonstrated the role of instructional clarity in students' academic motivation, cognitive load and achievement (Alles et al., 2017;Bolkan et al., 2016;Chan et al., 2021;Yagan, 2021).In neuroscience, cognitive load is associated with attention and memory systems in students, and the learning process is based on attention and memory (Tokuhama-Espinosa, 2021).These systems require considerable energy, which reflects the magnitude of the cognitive load.In line with this understanding, Sweller et al. (2011) argued that instructional clarity can be used to reduce cognitive load in students.
Teachers' instructional behaviours are determined by various factors including the content of the subject (Darmaji et al., 2019;Houichi & Sarnou, 2020).According to Titsworth et al. (2015, p. 410), "future research should investigate clarity within and across disciplines".The process of teaching physics content is based on instruction in which information is presented so that students can acquire new knowledge (Demkanin, 2018).Such instruction should be examined in detail by considering students' thinking while they process new information and should be enriched through the addition of new roles for physics teachers, namely, mentoring, tutoring, coaching and scaffolding (Chen & Lu, 2022;Demkanin, 2018).According to West (2023), a physics teacher is always in an unplanned space in terms of questions from students, and for the learning process to be successful, it is important that certain essential elements, including clarity, are present in physics education.
The Trends in International Mathematics and Science Study (TIMSS) provides the opportunity to analyse instructional clarity in physics education and middle school students' motivation.Research reviews using the TIMSS international dataset provide rich data on the influence of students' motivation on levels of achievement in science subjects (Steidtmann et al., 2022;Tsai & Yang, 2015;Zhang & Bae, 2020).The data also supplement existing knowledge on the influence of the subcomponents of motivation (i.e.self-concept, self-confidence, interest, value and motivational beliefs) on students' science outcomes, such as a deeper conceptual understanding, scientific reasoning and opportunities to participate in science (Çiftçi & Yıldız, 2019;House & Telese, 2017).
Until recently, few studies have focused on the relationship between instructional clarity in physics lessons, students' motivation to learn physics and students' confidence in learning physics.In this study, we therefore aimed to explore the role of instructional clarity in physics lessons on students' motivation to learn physics in direct and indirect ways with consideration of the mediating effect of students' self-confidence in learning physics.
The following research questions were formulated: (1) How does instructional clarity in physics lessons predict students' motivation to learn physics?
(2) How does instructional clarity in physics lessons predict students' self-confidence in learning physics?
(3) How is students' self-confidence in learning physics associated with their motivation to learn physics?
(4) What are the direct and indirect effects of instructional clarity in learning physics on students' motivation to learn physics?

Instructional clarity and motivation
Scholars have described instructional clarity in education as the verbal, nonverbal and structural behaviours of teachers that lead to students' understanding of the course content.Furthermore, a teacher's capacity to deliver classroom instruction clearly and concisely (Maulana et al., 2016) may help increase students' ability to process learning information (Bolkan et al., 2016;Wackermann et al., 2010).Teacher clarity has a strong effect on students' affective and cognitive learning (Bolkan et al., 2016).Instructional clarity guarantees instructional engagement, instructional scaffolding, maximises students' opportunities to learn and establishes a supportive environment, that encourages students to actively participate in classroom activities (Chen & Lu, 2022;Demkanin, 2022;van de Pol et al., 2010).
Cognitive load theory (CLT) states that instructional clarity manifests in education when the teacher provides concrete and worked examples and explanations and applies signalling, vagueness, redundancy and coherence (Alles et al., 2017;Bolkan et al., 2016;Mayer & Moreno, 2010).To apply signalling, the teacher provides summaries, links new content to previous course content, stresses key words and draws students' attention to essential course content (Mayer & Moreno, 2010).Vagueness means the teacher helps create a mild tone, fills lexical gaps, provides the appropriate amount of information and deliberately withholds information in the classroom, while coherence means that the teacher may provide material that is not essential for learning, and redundancy refers to the teacher giving repetitive information (Bolkan, 2017).
CLT addresses the essence of instructional clarity in education from a psychological perspective.According to CLT, there are three forms of cognitive workload: germane load (students' mental resources), intrinsic load (the difficulty of the material to be learned) and extraneous load (poor instructional design) (Paas & Ayres, 2014;Sweller et al., 2011).Instructional clarity reduces students' intrinsic and extraneous loads, ensures that their limited working memory capacity is not exceeded and facilitates the learning process (Bolkan et al., 2016;Mayer & Moreno, 2003, 2010;Tokuhama-Espinosa, 2021).This is relevant in teaching physics because this subject has a heavy intrinsic load.As a discipline, physics requires the application of mathematical methods to explain natural phenomena and requires that students understand equations, diagrams and graphs (Darmaji et al., 2019;Demkanin, 2018;Ornek et al., 2008;Redish, 2022;Wilson et al., 2020).From a neuroscience perspective, the heavy intrinsic load can be reduced if teachers consider the fact that the adolescents are distracted by the social world, and they want to learn about this world (Lieberman, 2013).To reduce the intrinsic load in the physics classroom, teachers "must be trained in each of the basic dimensions-knowledge, abilities, and relationships" (Demkanin, 2018, p. 4).
Teacher-student interaction based on tutoring, scaffolding and mentoring promotes students' learning motivation (Demkanin, 2022).In terms of neuroscience, "motivation is a complex cognitive, affective, and behavioral phenomenon that is likely mediated by multiple neural structures and processes" (DiDomenico & Ryan, 2017, p. 146).The motivation to learn physics depends not only on internal factors (interest, knowledge and abilities), but also on external ones (teaching methods by teachers and the educational subject) (Astalini et al., 2019;Fischer & Kauertz, 2021;Hong et al., 2017).A physics teacher can motivate their students to learn physics by using instructional materials while reducing their intrinsic and extraneous loads."In studying physics, students usually carry out investigations, both in class and in the lab, examples of observing and concluding information, classifying data, predicting, measuring, and questioning, interpreting and analysing data" (Darmaji et al., 2019, p. 16).Instructional clarity allows students to spend more time clarifying ideas in a lesson, highlighting the relationships between concepts and relating these to prior knowledge (Demkanin, 2018;Fischer & Kauertz, 2021;Fisher & Horstendal, 1997).
A reduction in intrinsic and extraneous loads has been positively associated with cognitive and affective learning (Mazer, 2013;Titsworth & Mazer, 2010;Titsworth et al., 2015).Bolkan et al. (2016) investigated the relationship between instructional clarity and students' academic motivation and test results and found a complex relationship between these variables.They observed that clear instruction does not increase students' test scores when their motivation is low, but it increases these scores when their motivation is high.

Self-confidence in learning and motivation
In this study, we aimed to determine the internal driving force that draws students to the study of physics.Motivation can be defined as "a driving force or forces responsible for the initiation, persistence, direction, and vigour of goal-directed behaviour" (Colman, 2015, p. 278), and has been analysed using different theories (Eccles & Wigfield, 2002).Zhang and Bae (2020) performed a large systematic literature review of the motivation theories used in the TIMSS and provided a conceptual overview of the dominant theoretical frameworks: expectancy value theory (EVT) and self-determination theory (SDT).In this study, we use EVT to examine students' motivation and self-confidence in learning and applied SDT to study students' academic motivation.
SDT describes motivation as a continuum from amotivation and extrinsic motivation to intrinsic motivation (Deci & Ryan, 2002), where intrinsic motivation is considered the highest level of motivation.Intrinsically motivated individuals engage in an activity freely and are directed by feelings of interest and enjoyment (Byman et al., 2012;Loukomies et al., 2013;Núñez & León, 2019).For instance, most students engage in physics when they feel fully successful in physics and when their endeavours are fruitful (Guido, 2013).
The findings of the TIMSS provide the opportunity to analyse both intrinsic and extrinsic motivation in the study of science.Leong et al. (2018) examined the effects of intrinsic and extrinsic motivation on student science achievement using the TIMSS 2011 data and found that students from Western cultures value extrinsic motivation more than intrinsic motivation in science learning.Meanwhile, Lay and Chandrasegaran (2016) performed a secondary analysis of the TIMSS 2011 data of students from Malaysia and Singapore and found that students' academic motivation and confidence in learning science were positively and statistically significantly associated with their achievements in the subject.
According to EVT, the motivational beliefs of students depend on their goals, self-concepts and subjective task value (Eccles & Wigfield, 2002;Rosenzweig et al., 2019;Trautwein et al., 2012).The TIMSS instruments provide an opportunity to explore the EVT components of intrinsic value, utility value, self-concept and self-confidence (Zhang & Bae, 2020).Notably, researchers have treated the same TIMSS items differently, and confusion exists regarding the meaning of the terms self-concept, selfconfidence and self-efficacy (Bong & Clark, 1999).For example, the item "I usually do well in science" has been treated as both a self-concept item (Liou, 2017;Wang & Liou, 2017) and a confidence item (Akilli, 2015;House & Telese, 2017;Lay & Chandrasegaran, 2016;Liu & Wang, 2019).In our study, selfconcept refers to a person's perception of themself (Shavelson et al., 1976).While self-confidence is a person's belief in their own ability to excel in an activity (Akbari & Sahibzada, 2020;Bong & Clark, 1999;Maclellan, 2014), self-efficacy is a person's belief in their own knowledge and themself in a practical situation (Bandura, 1997;Jansen et al., 2015).Accordingly, self-efficacy can be conceptualised as casespecific self-confidence (Çiftçi & Yıldız, 2019;Gonca Usta, 2017).School students may indicate that they are confident and can solve physics tasks well (i.e.academic self-confidence in solving physics tasks) but feel less efficacious about their ability to solve highly complex physics tasks (i.e.academic self-efficacy).The academic view of students' perceptions of their own subject knowledge and abilities is defined as academic self-confidence (Everingham et al., 2017;Sheldrake, 2016), which is demonstrated via their academic self-concept and academic self-efficacy (Jansen et al., 2015).
A large body of research has examined students' academic self-confidence.For example, researchers have addressed the impact of self-confidence on the learning process (Akbari & Sahibzada, 2020;Stankov et al., 2012;Tridinanti, 2018), the self-confidence of female and male students in open-access courses (Atherton, 2015), the role of self-confidence in achievement in interdisciplinary science subjects (Everingham et al., 2017), the impact of self-confidence on the academic achievements of elementary students (Verma & Kumari, 2016), students' selfconfidence in mathematics (Nurmi et al., 2003), the associations between self-confidence, gender and locality and the achievements of adolescents (Fatma, 2015), the role of drama education in the self-confidence and problem-solving skills of primary school students (Palavan, 2017) and selfconfidence and critical thinking (Hong et al., 2021).Scholars who analysed the role of selfconfidence found that while academic self-confidence influences students' motivation to learn, a lack of self-confidence can result in students' lack of motivation to learn and negative attitudes towards learning (Benabou & Tirole, 2002;Palavan, 2017).Considering SDT, EVT and the results of the aforementioned studies on academic self-confidence, we hypothesised: H 2 : Self-confidence in learning physics directly affects students' motivation to learn physics.

Instructional clarity and self-confidence in learning
We investigated the mediating role of academic self-confidence on the relationship between instructional clarity and students' motivation to learn by drawing on two theories: EVT and CLT.According to EVT, students look for answers to two questions when performing a task: "Am I able to do this task?"(expectancy component) and "Why am I doing this task?"(intrinsic value component) (Pintrich, 2003).The expectancy component is related to students' cognitive appraisals (i.e.self-concept and self-confidence).Maclellan (2014) noted that while two students may have the same belief or idea, one may have considerably greater cognitive appraisals of that belief or idea than the other.Moreover, students who believe they are able to do well are much more likely to be motivated (Maulana et al., 2016).Cognitive appraisals-self-concept and self-confidence-impact students' final judgments and task performance results.Maclellan (2014) established that self-confidence is not a stable capacity, and people can "lose" or "gain" it (p.60).
The literature on teachers' instructional behaviours suggests that instructional clarity influences students' self-confidence (Alles et al., 2017;Maclellan, 2014;Maulana et al., 2016).Teachers' instructional behaviour involves both instructional clarity to achieve the learning goal and classroom management to maximise students' opportunities to learn.Researchers have distinguished between three forms of teachers' instructional behaviour: strong control (teacher-led), shared control (student-led) and loose control (co-student-led) (Maulana et al., 2016).Teachers' instructional behaviours maximise students' opportunities for self-regulation and responsibility and are positively associated with the expectancy component (Alles et al., 2017;Maulana et al., 2016).
Instructional clarity does not always work because of students' individual differences in thinking about course material (elaboration of information) and their motivation to learn (Bolkan et al., 2016).Studies have confirmed the role of instructional clarity in students' self-concept and selfconfidence.For instance, Boekaerts and Rozendaal (2010) studied the role of instructional methods and the specifics of mathematics tasks in students' confidence and found that instructional methods influence students' confidence.Chen and Lu (2022) investigated the relevance of classroom management and instructional clarity to students' enjoyment and achievement.They noted the mediating role of students' mathematics self-concept in the relationship between instructional clarity and students' academic emotions and achievement (Chen & Lu, 2022).However, research on the role of instructional clarity in students' self-concept and self-confidence is lacking, especially with regard to the specifics of the subject (Maclellan, 2014;Maulana et al., 2016).Considering this, we proposed the following hypotheses:

Study location and sample size
We initially aimed to conduct a secondary data analysis of all the countries with student achievement scores above the TIMSS 2019 scale centre point (Mullis et al., 2020) and found that the achievements of the students in 18 countries were higher than the TIMSS 2019 centre point.We were interested in the data from three questionnaires: BSBP 38 i , BSBP 39 i and BSBP 40 i (BSBP, variable name for the TIMSS 2019 Student Questionnaire).However, we were unable to locate the data (BSBP 38 i , BSBP 39 i and BSBP 40 i ) for some countries in the TIMSS 2019 database.For this reason, we performed a secondary analysis of the data from only six countries with student achievement scores above the TIMSS 2019 scale centre point: Finland, Lithuania, Hungry, Sweden, Portugal and the Russian Federation (Table 1).We downloaded the primary data from the TIMSS 2019 database (http://www.timss.org/)and removed questionnaires hat had not been completed, so the sample was reduced (Table 1).Our final study sample was therefore smaller than the initial samples for each country (Table 1).We merged the databases of the different countries into one database, and the total sample was 21,029 subjects.

Research model and instrument
In this study, students' motivation to learn physics, their confidence in learning physics and instructional clarity in physics lessons were explored by performing a secondary analysis of data from three validated TIMSS 2019 questionnaires, respectively: BSBP 38 i , BSBP40 i and BSBP 39 i .All questions (BSBP 38i, BSBP 39i and BSBP 40i) have four ordered categories: 1 -agree a lot, 2 -agree a little, 3 -disagree a little and 4 -disagree a lot.
We analysed the data from BSBP 39 i .Based on this questionnaire, we created a latent variableinstructional clarity in physics lessons (ICPL).In our structural model (Figure 1), the unobserved latent variable (ICPL) was measured with seven observed variables (BSBP 39 i ) and was considered a latent factor.The TIMSS 2019 BSBP 40 i measures how self-confident students are about their ability in physics in terms of their level of agreement with eight statements about learning physics.This scale consists of positive and negative statements about students' confidence in learning physics.In this study, we used only positive statements (BSBP 40A, BSBP 40D, BSBP 40E and BSBP 40F) about the students' selfconfidence in learning physics.Based on these positive statements, we created a latent variableself-confidence in learning physics (SLP).SLP encompasses four observed variables (Figure 1).
The TIMSS 2019 BSBP38 i consists of nine items about motivation to learn physics.Two statements on this scale correspond amotivation: BTBS 38B ("I wish I did not have to study physics") and BTBS 38C ("Physics is boring").Other statements express extrinsic and intrinsic motivation to learn physics.Based on these extrinsic and intrinsic motivation statements, we created a latent variable-motivation to learn physics (MLP) (Figure 1).
We calculated the average variance extracted (AVE) and composite reliability (CR) of the latent variables: ICPL, SLP and MLP (Table 2).We tested the internal consistency of latent variable items by calculating their Cronbach's alpha coefficients (Table 2).The convergent validity and composite reliability scores of our latent variables are appropriate (AVE > .50;CR > .70)(Fornell & Larcker, 1981), and the Cronbach's alpha coefficients confirm the good internal consistency of latent variable (ICPL, MLP and SLP) items (>.650) (Table 2).
The normality of the data was checked.We hold those values for asymmetry (skewness and kurtosis) between − 2 and + 2 are considered acceptable to prove a normal univariate distribution (George & Mallery, 2010).Asymmetry coefficients indicated that the data satisfy the condition of normality.We calculated the multivariate normality of the data (critical ratio [c.r.] = 4.326 < 5.0).The critical ratio of kurtosis indicated the multivariate normality of the data (Bentler, 2006).We checked for the multicollinearity of the variables and found that the variance inflation factor (VIF) values met the requirements (VIF <4.0).
The structural model (Figure 1) was analysed using confirmatory factor analysis (CFA) and structural equation modelling (SEM).The fitness of the items of these latent variables and the structural model showed a sufficient fit and confirmed the structure of the latent variables (Table 3) (Bentler & Bonett, 1980).

Figure 1. Structural model: instructional clarity in physics lessons (ICPL), self-confidence in learning physics (SLP) and motivation to learn physics (MLP).
Confirmatory factor analysis enabled us to reveal the relationships between the unobserved variables (ICPL, MLP and SLP) and observed variables (BSBP 38 i , BSBP 39 i and BSBP 40 i ) (Figure 1).

CFA results: analysis of latent constructs
Confirmatory factor analysis shows that the latent variable ICPL is statistically significantly related to all observed variables of ICPL (p < .001)(Table 4).
To compare the effects of the different predictors of BSBP 39i on the outcome (ICPL), we studied the magnitudes of the standardised beta (β).The following items have a strong relationship with ICPL (Table 4): "The teacher is good at explaining physics" (β = .876,p < .001),"The teacher has clear answers to my questions" (β = .872,p < .001),"The teacher is easy to understand" (β = .817,p < .001)and "The teacher does a variety of things to help us learn" (β = .853,p < .001).The highest standardised coefficients are found between ICPL and the following explanation activities: "The physics teacher is good at explaining physics", "The physics teacher has clear answers to my questions" and "The teacher is easy to understand".
The coefficient of determination (R 2 ) indicates the percentage of variation in the ICPL model explained by the observed variables of BSBP 39 i (Table 4).It shows that the data fit the regression model (ICPL).The values of R 2 vary from 38.6% to 76.8% (Table 4).
We investigated the internal structure of the latent variable MLP (Figure 1).As noted earlier, the fitness of this variable's items revealed a sufficient fit and confirmed its structure (Table 3).The CFA results indicate a statistically significant relationship between the unobserved variable (MLP) and the observed variables of BSBP 38 i (p < .001)(Table 5).
The magnitudes of the standardised beta (β) of the variables corresponding to intrinsic motivation, "I like physics" (β = .940,p < .001)and "Physics is one of my favourite subjects" (β = .849,p < .001),confirm the significant effect on the outcome (MLP).
The coefficient of determination (R 2 ) indicates the percentage of variation in the MLP model explained by the observed variables.It should be noted that the R 2 of the variable "I like physics" is the smallest among all variables (Table 5) but is statistically significant (R 2 = .388> .200).This means that 38.8% of the data fit the regression model (MLP).
We also examined the internal structure of the latent variable SLP.The CFA results confirm the statistically significant relationship of positive statements with the unobserved variable (SLP) (p < .001)(Table 6).The CFA results show that the standardised beta (β) is higher for mastery experience items: "I learn things quickly in physics" (β = .896,p < .001)and "I am good at working out difficult physics problems" (β = .832,p < .001).Meanwhile, the standardised beta (β) for social persuasion item "My teacher tells me I am good at physics" (β = .794,p < .001) is slightly lower (Table 6).
The CFA results confirm that the coefficient of determination (R 2 ) varies from .631 to .803.This means that 63.1 percentage points should fall within the regression line of the variable "My teacher tells me I am good at physics", and 80.3 percentage points should fall within the regression line of the variable "I learn things quickly in physics".The coefficients of determination (R 2 ) of  the other variables show a high percentage of variation in the SLP model explained by the observed variables (Table 6).

SEM results: analysis of the structural model
We used SEM to determine the association of instructional clarity in physics lessons with students' motivation and self-confidence to learn physics (Figure 1).We performed SEM using the software Amos 24 to test our four hypotheses (H 1 -H 4 ).SEM indicates the possibility of relationships between the latent variables (ICPL, MLP and SLP).It includes two components: a measurement model (essentially CFA) and a structural model (Figure 1).Our model consists of exogenous (ICPL) and endogenous (MLP and SLP) variables (Figure 1).
We find that all direct and indirect paths are significant in the structural model (Table 7).The results of SEM (p value) show that instructional clarity in physics lessons directly and positively affects students' motivation to learn physics (β = .266,p < .001),(R 2 = .764)(Table 7).It also directly and positively affects students' self-confidence in learning physics (β = .564,p < .001),(R 2 = .617).We find that students' self-confidence in learning physics directly and positively affects their motivation to learn physics (β = .874,p < .001),(R 2 = .764).Therefore, we confirm the hypothesis that instructional clarity in physics lessons indirectly and positively affects students' motivation to learn physics (β = .493,p < .001)(Table 7).
The values of the standardised beta (β) confirm the significant effect on the outcomes (MLP and SLP) (Table 7).The values of the coefficient of determination (R 2 ) indicate that the model is a good fit for the data and show a strong linear relationship between the variables (ICPL, MLP and SLP) (Table 7).

Discussion
We tested four hypotheses about the role of instructional clarity and self-confidence in learning physics in students' motivation to learn physics.The first hypothesis testing aimed to determine whether instructional clarity in physics lessons (ICPL) directly affects students' motivation to learn physics (MLP).A large body of research based on Cognitive load theory (CLT) revealed the associations between instructional clarity and students' motivation (Bolkan et al., 2016;De Loof et al  2019; Hong et al., 2021;Titsworth & Mazer, 2010;Titsworth et al., 2015).According to CLT, instructional clarity manifests in education when the teacher provides concrete and worked examples, gives explanations.Confirmatory factor analysis shows that the latent variable ICPL is statistically significantly related to the explanation activity of the physics teacher ("My teacher is good at explaining physics", "My teacher has clear answers to my questions") (Table 4).The CFA results also show the role of perceived immediacy ("I know what my teacher expects me to do") in instructional clarity in physics lessons (Table 4).These are in line with the findings of other studies in which perceived immediacy positively correlates with instructional clarity (Kelly & Gaytan, 2020;Violanti et al., 2018).Cognitive and emotional support are positively related to students' interest in learning physics from grades 5 to 6 and from grades 6 to 7 (Steidtmann et al., 2022).
The researchers argue that autonomy support, structure (feedback, coherent explanations), and involvement promote students' motivation (Chan et al., 2021;Chen & Lu, 2022).Chan et al. (2021) analysed the perceived instructional behaviour of teacher educators and pre-service teachers' learning motivation.Researchers revealed the direct effect of instructional clarity on pre-service teachers' intrinsic motivation (β = .261,p < .001)and did not reveal statistically significant effect on extrinsic motivation (Chan et al., 2021).Our results of SEM show that instructional clarity in physics lessons directly and positively affects students' motivation to learn physics (β = .266,p < .001).Surprisingly, the very similar result was obtained for learners of different age groups and learning experience.
The second hypothesis testing aimed to determine whether self-confidence in learning physics directly affects students' motivation to learn physics (Figure 1).The CFA results show that the mastery experience items "I learn things quickly in physics" and "I am good at working out difficult physics problems" are strongest predictors of students' self-confidence, whereas social persuasion item "My teacher tells me I am good at physics" show the weakest prediction (Table 6).This finding is consistent with expectancy value theory (EVT), which highlights the significant interplay of learners' cognitive appraisals (self-concept and self-confidence) with the learning environment and motivation for learning (Rosenzweig et al., 2019;Trautwein et al., 2012).Students' selfconfidence seems to function as a substantial factor in motivating them to learn physics.The results of the second hypothesis testing are consistent with those of another study about the role of students' self-concept in their enjoyment of learning mathematics (Chen & Lu, 2022).Furthermore, our second hypothesis testing confirmed the suggestion that physics teachers seeking to promote students' motivation to learn should focus primarily on enhancing students' selfconfidence in learning (Gonca Usta, 2017).
In the third hypothesis testing, we found that a direct path (ICPL→ SLP) is significant in the structural model (Table 7).Our results indicate the significant and favourable effects of instructional clarity on students' self-confidence in learning physics.Other researchers have investigated the relationship between teachers' instructional clarity and students' cognitive appraisals (selfconcept and self-confidence) (Avtgis, 2001;Chen & Lu, 2022;Maclellan, 2014).Avtgis (2001) found that clarity is positively correlated with students' attributional confidence.Chen and Lu (2022) established the positive and statistically significant role of instructional clarity in students' mathematics self-concept.Maclellan (2014) examined the role of teachers in the development of students' self-confidence and identified two opportunities: enhancing emotional well-being and strengthening domain knowledge.The results of our research do not contradict the findings of the aforementioned studies.
Consistent with our fourth hypothesis, we found an important mediating role of students' selfconfidence in physics learning in the relationship between instructional clarity in physics lessons and motivation to learn physics.This is in line with the results of a study about the mediating role of students' mathematics self-concept in the relationship between students' perceptions of instructional clarity and academic emotions (Chen & Lu, 2022).
Summarising the discussion, we can conclude that the findings of our research might enrich certain aspects of physics didactics and physics teaching practice.It appears that research on instructional clarity in learning physics has a positive bias, assuming that clarity is positively associated with students' motivation to learn physics.We identified a statistically significant positive effect of instructional clarity in physics lessons on students' motivation to learn physics.Our findings support EVT by establishing the mediating role of cognitive appraisal-self-confidence in learning physics-in the relationship between instructional clarity in learning physics and motivation to learn physics.Our findings also support CLT by disclosing the statistically significant effect of giving explanations in physics lessons on instructional clarity in learning physics.Therefore, physics teachers can reduce the heavy intrinsic load associated with the physics subject and promote students' self-confidence in physics learning by giving more explanations, augmenting perceived immediacy and using signalling and redundancy.
According to Mayer and Moreno (2010), all signalling methods in the lesson are important: linking new content to previous course content and stressing key words.A physics teacher could encourage students to discover key words, list of them in each physics topic, and create concept maps based on the content of several physics' topics.This would enable students to see the connection between the new content and previously covered physics content.The key concepts acquire a new meaning when we apply programming in physics lessons.Basic physics concepts or their symbols become part of the program code.
When using multimedia and interactive screens during lessons, the teacher should keep in mind the principle of coherence and redundancy because these tools work best if there is no competition for the learner's attention between text and graphics, and narration (Clark & Mayer, 2011).
As already mentioned, students' motivation to learn physics decreases particularly rapidly in middle school (Alexander et al., 2019;Byman et al., 2012).By performing a secondary analysis of TIMSS 2019 data, we investigated middle school students' (eighth graders') motivation to learn physics.We hope that our findings will encourage physics teachers working with middle school students to consider the mediating role of self-confidence in learning physics in the relationship between instructional clarity and motivation to learn physics.

Limitations and directions for future research
As with any other empirical research, our study has some limitations.We consistently showed the positive effect of instructional clarity on students' motivation to learn physics but did not analyse instructional clarity in physics lessons from a teacher-student communication (adaptive instruction) perspective.Clarity in physics lessons is a process of communication in which the meanings of physics concepts, relationships and procedures are discussed.Therefore, future research about instructional clarity in learning physics must delve more deeply into questions about how instructional clarity, classroom communication and motivation to learn physics are interrelated.Considering teacher-student classroom communication, we can move closer to the constructivist aspect (teacher-student shared control) of clarity while obtaining a deeper understanding of instructional clarity as a negotiation process in physics lessons.
This study used the combined TIMSS 2019 databases of countries whose students' achievement scores were above the TIMSS scale centrepoint.Therefore, conducting a study about the effect of instructional clarity in physics lessons on the motivation to learn physics using TIMSS 2019 data from countries whose students' achievements are below the TIMSS 2019 scale centrepoint would be appropriate.Correlating the students' views on instructional clarity with what a teacher does will likely require a different kind of study.Qualitative research could perhaps reveal more about the role of instructional clarity in the motivation to learn physics by considering the individuality of students and the range of their cognitive abilities of their cognitive abilities.
We revealed the role of instructional clarity in physics lessons in students' motivation to learn physics at a low inference clarity level.This study thus calls for testing other mathematical models at the intermediate inference clarity level by identifying the multiple mediating roles of organisational dimensions (intermediate inference).Testing new models at this level would enrich the discussion about the impacts of a minimally guided approach (teacher-student shared control) and an instructional guidance approach (teacher-centred instruction) during physics education.

H 3 :
Instructional clarity in physics lessons directly affects students' confidence in learning physics.H 4 : Instructional clarity in physics lessons indirectly affects students' motivation to learn physics.

Table 3 . Fitness of the items of the latent variables: model ICPL, MLP, SLP and the structural model
Notes: χ 2 -absolute/predictive fit Chi-square; RMSEA-root mean square error of approximation; GFI-goodness-of-fit index; IFI-incremental fit index; TLI-Tucker-Lewis index; CFI-comparative fit index.

Table 5 . Standardised and unstandardised coefficients of the latent variable MLP Question code Items about motivation for learning physics
2 -coefficient of determination; B-unstandardised coefficients; SE-standard error for the unstandardised beta; β -standardised beta; p -probability.

Table 6 . Standardised and unstandardised coefficients of the latent variable SLP Question code Items about confidence in learning physics R 2 B S.E. β p label
.,