Modeling Different Variables in Learning Basic Concepts of Programming in Flipped Classrooms

Abstract

Learning the basic concepts of programming and its foundations is considered as a challenging task for students to figure out. It is a challenging process for lecturers to learn these concepts, as well. The current literature on programming training abounds with the examples of a wide range of methods employed. Within this context, one of the prominent approaches in programming training is flipped classroom (FC) model. This article has sought to illuminate the effect of cognitive flexibility, problem-solving skills (PSS), and flipped learning readiness (FLR) levels on students' programming achievements in programming training through FC model. A total of 149 freshmen computer science students studying in a state university in Turkey were recruited for this study. In this study, designed as a relational screening model, a personal form, an achievement test, and three different data collection instruments were employed to collect data. For the data analysis, structural equation modeling, a multivariate statistical analysis technique, was used to reveal a model explaining and predicting the relations between programming achievement and different variables. The findings clearly indicate that FLR is the most important predictor of the programming achievements of students in FC. Other important predictors were found as PSS and cognitive flexibility. The research model demonstrates that an increase or development in FLR, PSS, and cognitive flexibility levels in FC will enhance the achievements of students in programming.

Keywords

programming training App Inventor flipped classroom cognitive flexibility problem-solving skills university students

Introduction

Computer programming involves a great variety of skills, including problem-solving, computational thinking, abstracting, critical thinking, and analytic thinking (Gomes & Mendes, 2007; Gundurao, Manjunath, & Nachappa, 2010). At present, programming skills are important in terms of preparing computer science students for their future careers effectively (Vaca-Cárdenas et al., 2015). However, novice programmers encounter a number of challenges during programming learning in higher education (Topalli & Cagiltay, 2018). Learning programming for the first time requires having knowledge on not only semantic concepts but also syntax (Solmaz, 2014). As Gomes and Mendes (2007) confirm, a great number of problems during programming training derive from such concepts as variables in programming languages, cycles, ranges, functions, and syntax. On the basis of the evidence by a growing body of research currently available (Sáez-López, Román-González, & Vázquez-Cano, 2016), it seems fair to suggest that these complexities may evolve into drawbacks related to learning programming, considering it as a challenging process and decreasing interests and motivations of learners. Along similar lines, Rizvi, Humphries, Major, Jones, and Lauzun (2011) highlight that an interactive learning environment is sine qua non because there are myriad concepts that need to be learned by learners in limited programming process. In the literature, there are several examples of studies (Yildiz-Durak, 2018a) lending support to the needs of students as to easier programming software interface while learning programming and problem-solving skills (PSS). There is a vast amount of literature (Chen, 2014; Giannakos, Krogstie, & Chrisochoides, 2014; Herala, Vanhala, Knutas, & Ikonen, 2015; Horton & Craig, 2015; Karaca & Ocak, 2017; Lepp & Tonnison, 2015; Puarungroj, 2015; Tugun, Uzunboylu, & Ozdamli, 2017; Yildiz-Durak, 2018b) on flipped classroom (FC) model's convenience for increasing the interest, motivation, and academic achievements of students, by prodiving a better context, and thus enhancing practice opportunities, immediate feedback, and individualization of reflection in programming training.

Constructionists put forward that learners build knowledge in an active way when provided more practice and interaction opportunities (Papert, 1980). FC model adds different dimensions to the interaction of learner–content–learner–lecturer and provides a pattern backing up active engagement and interaction through tools presented online in programming training (Tugun et al., 2017). It is therefore thought that FC model is a convenient means of programming training. However, an increasing number of studies have emphasized that learning circumstances are associated with preliminary preparation before the class, technology use and communication skills, and learning motivations in FC model (Kim, Kim, Khera, & Getman, 2014). It is particularly important to take the students' readiness levels into consideration because there are also activities involving out-of-classroom tasks in FC model (Hwang, Lai, & Wang, 2015). On the other hand, as Gundurao et al. (2010) remind us, the most important skill employed in programming learning is the PSS. Robins, Rountree, and Rountree (2003) define “problem solving as a fundamental skill in programming training” (p. 151). When preparing problem solutions with computers, it is not enough to know only the rules of the programming language, but PSS are required for successful programming. Besides, good PSS need to be developed to learn how to write a computer program (Gomes & Mendes, 2007).

Cognitive flexibility is considered as one of the most preliminary individual differences that can guide during the regulations of the solution and output as well as assessment in problem-solving process (Alper & Deryakulu, 2008). Cognitive flexibility of the individuals are linked to the skills, including learning, regulating PSS, showing guidance, counseling, and checking (Zohar & Dori, 2012). Cognitive flexibility allows the student to employ the most effective learning strategies related to the topic being studied or to identify the steps to solve a problem, to find a solution, to control the learning process, and to control the products or self-regulating opportunity (Batting, 1979). In a nutshell, it can be argued that problem-solving and cognitive flexibility are those characteristics employed during programming training. To this end, this study sets out to assess the roles of PSS, cognitive flexibility, and flipped learning readiness (FLR) and the impact of these variables on programming achievement in FC model implementation in programming training.

The Purpose of the Study

This study seeks to determine the academic achievement levels of students in teaching basic programming concepts in FC model and to test if selected variables, “cognitive flexibility, PSS, and FLR levels,” considered as effective in programming learning, predict the academic achievement levels. In addition, it is also an attempt to reveal a model explaining and predicting the relations between programming achievement and different variables. On these grounds, the research question framing this study is “What are explanatory and predictive relations patterns between academic achievements of students and different variables in FC in programming learning?”

Conceptual Framework

FC Model

FC model is a kind of teaching and learning environment in which direct instruction is mostly provided outside and beforehand the classroom through videos. In this model, instructional time is employed for discussions on the topic, peer collaborations, and mentorship activities (Bishop & Verleger, 2013; Francl, 2014). FC model is used so as to achieve an active learning environment in a student-centered approach (Freeman et al., 2014). It makes possible to conduct active learning activities in the classroom, allowing teachers to afford more time for discussion activities. Students can develop their high-level cognitive skills and do more practice in this model (Bergmann & Sams, 2012; Davies, Dean, & Ball, 2013). It is therefore significant to note that using FC model in teaching activities such as programming is suggested to develop high skills of thinking and learning outcomes (e.g., Giannakos et al., 2014; Yildiz-Durak, 2018b). FC model enables teachers and learners to distribute students' learning materials to learners in both online and face-to-face environments and to achieve a well-balanced communication and interactions (Davies et al., 2013). Some previous research on FC model has shown that it makes possible to enhance student engagement, to boost students' learning experiences, to increase academic achievements, to deal with cognitive load, and to develop interaction of content–learner–teacher and their satisfactions as well as motivations (Awidi & Paynter, 2019; Chuang, Weng, & Chen, 2018; Connell, Donovan, & Chambers, 2016; Yildiz-Durak, 2018b).

Why Is FC Model in Programming Training?

Active engagement of the students is fundamental to programming process, and tools presented in online dimension of FC model support active engagement and interaction among students (Tugun et al., 2017). For example, FC model presents a broader instructional time and interaction prospects to teach abstract concepts of programming. It is possible for students to get support or feedback from their teachers and peers through messenger and discussions. In addition, the interests and developments of the students may vary according to individual differences. Therefore, as proposed by Lepp and Tonnison (2015), the fact that students can have a voice on learning steps individually in FC model will enhance their engagement to the lesson and motivations. This is due to the fact that FC model helps to revise the course content according to learners' wishes. Pawelczak (2017) notes that FC model is fundamental to programming training in that it personifies learning and backs up individual efforts. In addition, Lepp and Tonnison (2015) highlights that FC model in programming training serves for personalization of learning activities because it enables teachers to conduct different teaching approaches and methods. By the same token, in this model, students take more responsibility in following contents and programming assignments in and out of the classroom, thanks to preliminary preparations for the course (Pawelczak, 2017).

On the other hand, Chen (2014) argues that students may lose their interest and motivations toward programming when encountering difficulties with content and learning. The underlying argument in favor of Chen (2014) is that students will not be able to learn and succeed if not motivated actively (Ryan & Deci, 2000). With regard to this, it can be suggested that several learning resources are presented in both online and face-to-face learning environments to deal with the difficulties encountered in programming training in FC model. Thus, the active engagements, interactions, and motivations of the learners are boosted. Furthermore, FC model may affect attitudes of the students in a positive way as it presents flexibility in instruction time and place. In this model, students take the responsibility of their learning in out-of-classroom activities and realize programming tasks to develop programming self-skills in a long period of time.

The data generated by several studies conducted on FC model in programming training (Herala et al., 2015; Horton & Craig, 2015; Karaca & Ocak, 2017; Puarungroj, 2015) appear to suggest that FC model develops students' learning performances. The data yielded by Yildiz-Durak (2018b) provide convincing evidence that students' interaction levels, engagement to the lesson as well as attitudes and self-efficacies have been developed in programming training process through FC model. What is more, FLR was found to be effective on these variables in the same study. From growing support found in the literature, it can be concluded that FC model is a convenient means of addressing the difficulties and learning requirements. Durak (2016) poses that the balance between theory and implementation in programming activities must be well-planned, but these situations are often ignored. In this regard, it can be noted that FC model paves the way this situation. To put it differently, it presents a good means of including theoretical perspective of the topic and allocating enough time for in-class activities when designed effectively. It is also possible to compensate for learning mistakes of the students in the classroom via FC model. At this precise point, teachers must attach great importance to students' practices. In essence, FC model can be regarded as a convenient means of satisfying learning needs of the students. In this study, FC model applications used in programming education are presented in Figure 1.

Figure 1.

FC model applications.

Why Was App Inventor Employed in Programming Training?

Programming tools can be classified into two categories: text-based and block-based (Yildiz-Durak, 2018c). Text-based programming language necessitates employing text-based characters based on syntax rules while coding, whereas block-based programming environments rest mostly on visual programming tools. A line of research has revealed that block-based programming environments present several conveniences for users, including easy coding without errors, or easy detection of errors, coding without syntax errors, easy debug, designing projects via multimedia tools, reducing cognitive load, and user-friendly interface (Portelance, Strawhacker, & Bers, 2016; Yildiz-Durak & Güyer, 2018).

App Inventor is one of the block-based programming tools. Developed by Google for smartphones, it enables its users to produce new applications in Android system. In the meantime, it was taken over by MIT Media Lab. App Inventor Blocks Editor has some similarities with Scratch in terms of operating system and interface and is claimed to be suitable for users from all age groups. What is more, several scholars have suggested that it can be employed during programming training at the beginning of the courses, thanks to convenient block-based system (Karakus, Uludag, Guler, Turner, & Ugur, 2012; Yildiz-Durak & Şahin, 2018), and is a useful tool for achieving a high level of motivation (Wolber, 2011). Figure 2 presents an example of App Inventor online editor.

Figure 2.

The Interface of App Inventor online editor (http://ai2.appinventor.mit.edu).

The Role of FLR in Programming Achievement in FC

According to a learning theory suggested by Thorndike (1932), the readiness level of students strongly affects the achievements level of the students. From moving this point, it can be argued that the readiness level of students may affect the students' performance in teaching process conducted via FC model. On the other hand, there has been compelling evidence in literature (Albert & Beatty, 2014; Burke, 2015; Chen, Wang, & Chen, 2014) that FC model is based on active learning theory and relied on constructivist approach. As noted by Hao (2016b), the students have their own control over the learning process, and the decisions about when and where to make the students' learning materials belong to the students in this model. FLR levels form driving force when it comes to increasing learning performance and efforts of students on online contents in programming training conducted through FC model (Saadé, He, & Kira, 2007). In addition, FC presents several learning resources (videos in online and face-to-face environments, resources on course contents, worksheets, and programming task instructions) and proper technologies (discussion environments, text message/messenger, online exam, online surveys and questionnaires, and instant feedback tools) to students through online learning environment. Hung, Chou, Chen, and Own (2010) put forward that students need to have technological self-efficacy to solve problems and encourage students to engage in online environments and interactions, to collaborate and use online learning tools properly, and to access students' learning materials. This is expected to affect academic performance of students related to programming. These authors assert that online communication self-efficacy is a necessary dimension to get rid of the limitations of online communication. FC model requires students to do preview before coming to the classroom (Peled, Blau, & Grinberg, 2015). It is therefore important to revise learning materials used out of classroom in FC model. In the FC model, extracurricular learning preparation reduces the perceived difficulty of a lesson and increases the perceived value of that curriculum (Stelzer, Gladding, Mestre, & Brookes, 2009). It is assumed that this will change negative attitudes into positive ones and thus contributing to academic performance. Lending support to the studies emphasizing the relation of academic performance and FLR and its subdimensions, in this research, the effect of readiness level in teaching conducted through FC model on academic achievement was investigated. In line with the main aim of the research, the following hypothesis has been formulated:

H1: FLR levels of students will affect their learning performance of programming concepts positively in FC model.

Subdimensions of the FLR

In a study conducted on e-learning readiness by Demir and Yurdugül (2015), it was concluded that there were several models on different dimensions of e-learning readiness level on student, teacher, and organizations. The authors investigated 30 readiness models on e-learning and found several patterns, such as technology use self-efficacy, self-learning, access to technology, and confidence in prerequisite knowledge. The authors concluded that self-confidence, motivations, and time management come into prominence. Based on an extensive literature review, it can be noted that there has been limited research on these patterns' adaptation for FCs. Patterns suggested by Hao (2016b) and adapted by Yildiz-Durak (2017) were employed in this study. The reason why these patterns were employed is that they encompass all trendy ones, especially doing preview factor that is specific to FC. In this study, five factors explain flipped readiness level. These factors are as follows: learner control and self-directed learning, technology self-efficacy, motivation for learning, in-class communication self-efficacy, and doing previews.

Learner control and self-directed learning

Self-directed learning is defined by Zimmerman (2000) as “an active engagement of learners to learning processes in cognitive, motivational and behavioral” (p. 85). Pintrich (2000) describes self-directed learning as “an active process in which learners' ability to control their cognitive, motivational and behavioral situations following determination of their needs, to manage and to supervise these” (p. 95). As it has been in the literature, individuals with self-directed learning skills are supposed to manage their learnings and to take responsibility on their learning needs. According to Yildiz-Durak (2018b), learners in FC model take an active role in learning process and are supposed to do preview on the course topic, to watch videos and to share the problematic issues with their teachers and peers. At this precise point, it is crucial for learners to have learning control and self-directed learning skills because they are required to make decisions on the time, the place, and the type of the material they will design. It has a well-known fact that efforts on programming training impact the effectives of teaching. Thus, it is suggested that in FC model in which programming training is conducted, these skills must be paid attention. In this regard, the following hypothesis has been formulated:

H1a: The control and self-directed learnings of students will affect their learning performance of programming concepts positively in FC model.

Technology self-efficacy

Self-efficacy is identified as “belief of individuals as to their competency on a certain task if they can realize an aspiration successfully or not” (Bandura, 1977, p. 192). Çelen, Çelik, and Seferoğlu (2011) note that students can gain more achievements in online learning environments when they have high level of self-efficacy perceptions along with acquiring basic computing skills. Similarly, Hung et al. (2010) stress that determination of learners' technology self-efficacy levels yield fruitful outcomes in terms of effective utilization of online learning environment. Technology self-efficacy is crucial for FC model because lots of learning resources and tools are needed to be employed in online learning environments. Based on this, the following hypothesis has been proposed:

H1b: Technological self-efficacies of students will affect their learning performance of programming concepts positively in FC model.

Communication self-efficacy

Online communication self-efficacy can be described as “the perception of the individuals as to how they can understand the communication and culture in online environments and how well they can express themselves” (Demir, 2015, p. 33). On the other hand, Palloff and Pratt (1999) suggest that shy students tend to take part in online environments rather than traditional ones. Therefore, it can be noted that individuals with diverse communication tendencies may vary in benefitting from the learning opportunities in FC model. In this regard, based on the idea that communication self-efficacy is of great importance in programming training, the following hypothesis has been proposed:

H1c: Communication self-efficacies of students in the classroom will affect their learning performance of programming concepts positively in FC model.

Motivation for learning

Motivation is a driving force that empowers individuals to engage in an activity in a voluntary way intrinsically or extrinsically. The fact that learners have high levels of motivation has a profound impact on their learning performances (Ryan & Deci, 2000). The dimension motivation for learning in FC model is a driving force for students so as to engage learning activities as they wish and to improve their learning performance. Alsancak-Sirakaya (2015) suggests that understanding of students' motivation level toward learning and their tendencies is vital for planning educational resources and improving the effectiveness of teaching. In this context, the following hypothesis has been formulated:

H1d: Learning motivations of students will affect their learning performance of programming concepts positively in FC model.

Doing previews

FC model requires learners to take responsibility outside the classroom, as well. In this regard, this model necessitates doing preview before the instruction in the classroom (Peled et al., 2015). By the same token, this dimension is considered as fundamental to FC model. The instructional activities will not yield expected results unless learners do previews before the course action. Therefore, it is very important to revise the resources before coming to the classroom in FC model (Hao, 2016b). In this context, the following hypothesis has been formulated:

H1e: Preliminary preparations of students will affect their learning performance of programming concepts positively in FC model.

The Role of Cognitive Flexibility in Programming Achievement in FC

Batting (1979) defines “cognitive flexibility as a skill which uses effective learning strategies related to topic and determines the problem solving steps in learning process” (p. 546). According to Spiro, Collins, Thota, and Feltovich (2003), achieving a high-level learning in a learning environment depends on the easiness of learning transfer and addressing to students with different levels of cognitive flexibility. On the other hand, Andrade and Coutinho (2016) proposed a model in which cognitive flexibility can develop in FC model. In their model, Bergmann and Sams' (2014) model, based on Bloom's taxonomy, was cited as a reference. In FC model, it has been suggested that the balance between theory and practice must be achieved in and outside the school time.

FC programming training requires students to employ their PSS, determining the best means of problem-solving, thereby using their cognitive flexibility levels. In this regard, as Cooper, Dann, and Pausch (2000) remind us, many students are not ready to think algorithmically when entering the first programming courses and cannot create an algorithm for solving the problem when faced with programming problems. This indicates that the student does not have enough cognitive flexibility to solve various problems because algorithmic thinking and problem-solving are influenced by individual levels of cognitive flexibility (Alper & Deryakulu, 2008). In FC model, the cognitive flexibility levels of students are thought to be influential on programming performance. In this context, the related research hypothesis is listed in the following:

H2: Cognitive flexibility levels of students will affect their learning performance of programming concepts positively in FC model.

The Role of PSS in Programming Achievement in FC

PSS contributes to developing creativity, productivity, critical thinking, and cognitive flexibility (Klegeris & Hurren, 2011; Lazakidou & Retalis, 2010). Roach (2014) emphasizes that FC model presents better prospects in terms of time and space. This is due to the fact that students can focus on the problems in their homes through online platforms; can do several learning activities, projects, and assignments; and can practice in their school time. Allocating more time for practices in schools adds to PSS positively (Çakiroğlu & Öztürk, 2017). In addition, this enables students to take their own responsibility in learning, to engage courses actively, to determine their own targets, and to improve their self-control skills (Yildiz-Durak, 2018b). A line of research has suggested that FC model makes important contributions for programming training as it has been considered that the abstract concepts of programming is difficult to understand (Sáez-López et al., 2016). Models that can present time and space flexibility are needed in programming training. In addition, programming is related to problem-solving (Kelleher & Pausch, 2005). Problem-solving is regarded as one of the milestones of programming learning (Apiola & Tedre, 2012).

H3: PSS levels of students will affect their learning performance of programming concepts positively in FC model.

PSS, Cognitive Flexibility, FLR, and Structural Relations Among These Variables in Programming Achievement

Cognitive flexibility theory (Spiro, Coulson, Feltovich, & Anderson, 1988) emphasizes that students are required to benefit from authentic perspectives and flexible thinking structures so as to achieve a deep learning (Spiro et al., 2003). A line of research focusing on designed learning environments has revealed that cognitive flexibility supports the development of thinking skills (e.g., problem-solving) and helps learners have positive emotional changes during learning activities (Cheng & Koszalka, 2016). As indicated by Savery and Duffy (1995) and Jonassen (1997), learning environments and tasks must be complex, and learners must be allowed to display their PSS. Therefore, focusing on PSS and cognitive flexibility in which high-level thinking skills are needed provides information on effectiveness of teaching and interaction of learners with the content (Jacobson & Spiro, 1994). On the other hand, cognitive flexibility is defined as “adaptation of cognitive strategies and competency to deal with unexpected or changing conditions” (Dennis & Vander-Wal, 2010, p. 242). Therefore, cognitive flexibility accounts for self-regulation of the individuals so as to deal with new circumstances. The existing literature has focused on self-efficacy, awareness, attention, self-tendency, objectives, and so forth (Spiro et al., 1988).

Canas, Quesada, Antolí, and Fajardo (2003) and Fitzgerald (1997) note that individuals perform better when they include cognitive systems and environments in e-learning environments. At this point, cognitive flexibility and problem-solving play a significant role in solving problems faced in e-learning environments. On the other hand, Yildiz-Durak (2018b) highlights the importance of balance between theory and practice in programming training, and FC presents important advantages in this perspective. The author suggests that programming training is related to FLR in FC model. Several learning resources and technologies are presented to students in online learning environments in FC. According to Hao (2016b), students need access to students' learning materials, employment of tools in line with the aim of the course, interaction with teachers and peers, and FLR to solve problems in FC. Talbert (2015) notes that students accustomed to traditional means of teaching may have difficulty in adapting to FC. In summation, in FC that is considered as important in programming training, cognitive flexibility and PSS are significant in both adapting to FC and programming achievement. By the same token, those variables have been included in the present study.

Method

In this study, it was aimed to reveal a model explaining and predicting the relations between programming achievement and different variables as well as to examine the relations between programming achievements of university students and different variables. This research was designed as a relational screening method because it intends to reveal the relations among some variables. The relational screening model is the one in which two or more variables are measured to seek for the change altogether and the degree of so-called change. In this model, the relations among variables are not evaluated as a cause-effect relation, but determination of a variable enables researchers to make predictions on other variables (Karasar, 2005). This present study, designed as a relational screening model, encompasses the following procedures: identifying the problem, recruiting participants, implementation process, data collection, data analysis, and results. First of all, an extensive literature review has been conducted to see whether the relations among PSS, problem-solving, and cognitive flexibility levels are worth to investigate for measuring programming achievement. Following this, the purpose of the research has been identified, and sampling method has been decided. For this purpose, convenient sampling method has been chosen to solicited participants in the study. This sampling method is the one suggesting that participants who are suitable for the purpose of the study and easy to reach should be included in the study. Then, validity and reliability tests on data collection instruments have been checked in the previous research. Course contents, outline, and course materials have been formed, afterward. Course content and resource sharing have been planned as online and face-to-face practices in a balance. The implementation procedure has been completed successfully. Participants have been surveyed to collect data. Next, the obtained data were analyzed via structural equation modeling (SEM), and descriptive statistics were conducted. Finally, the findings based on the analysis were discussed. Figure 3 summarizes the implementation procedure.

Figure 3.

Implementation procedure.

Research Model

Figure 4 summarizes research model.

Figure 4.

Default research model.

From the model based on the hypotheses presented in Figure 4, the relation between programming achievement in FC and FLR levels as well as PSS and cognitive flexibility levels can be seen. A hypothesis for each dimension has been formulated because it is believed that focusing on each dimension separately will provide more detailed results.

Study Group

A total of 149 freshmen computer science students studying in a state university in Turkey during academic year in 2017–2018 were recruited for this study. The study group was sampled through convenience sampling method. The participants had a full-year preparation class of English in their higher education. When the existing literature is examined, it can be noted that language is one of the most significant barriers in block-based programming training (Durak, 2016). Therefore, it was thought that it would be necessary to get information on language levels of the participants. In the sample, 42.3% of the participants were female, while 57.7% were male. Participants were between 18 and 24 years of age.

Data Collection Instruments

In this study, a personal information form, an achievement test, and three different data collection instruments were employed. The information on data collection instruments and their items are demonstrated in Table A1 in appendix.

Personal information form

This form was developed by the researcher. Personal information and the data on information technology use were collected through this instrument, including 10 items. Two field experts were consulted while developing this form. Items in the questionnaire are in Likert-type scale even though they differ according to the questions.

Flipped Learning Readiness Scale

Developed by Hao (2016b), this scale was adapted to Turkish language for secondary school students by Yildiz-Durak (2017). The original form of the scale was developed by Hao (2016b). The original scale included 27 items and 5 subdimensions in total. These subdimensions are as follows: learner control and self-directed learning (FLR-1), technology self-efficacy (FLR-2), communication self-efficacy (FLR-3), motivation for learning (FLR-4), and doing previews (FLR-5).

In this study, the scale was attempted to adapt for university students as the study group included 90 different university students. During adaptation process, two field experts were consulted to ensure validity. According to the opinions of experts, an item was rewritten. Exploratory and confirmatory factor analyses were conducted after content and language validity was ensured. The scale included 26 items in 5 subdimensions as a result of adapting implementation. The Cronbach alpha coefficient was calculated as .921. The Cronbach alpha' coefficients related to subdimensions in adapting process can be listed respectively: .928, .901, .801, .892, .718. Internal consistency coefficients in the study correspond with the original scale development research. In the original scale development (total and subdimensions), Cronbach alpha coefficients were calculated as respectively .899, .889, .810, .812, .715, and .700.

Cognitive Flexibility Scale

This scale was developed by Martin and Rubin (1995) and was adapted to Turkish language by Çelikkaleli (2014). Responses were on a 6-point, Likert-type scale that includes 12 items and one dimension. The scores can be obtained from the scale range from 12 and 72. The high scores in the scale also mean high level of cognitive flexibility. In this study, Cronbach alpha coefficient was calculated as .775.

Problem-Solving Skill Scale

This scale was developed by Heppner and Peterson (1982) and was adapted to Turkish language by Sahin, Sahin, and Heppner (1993). Responses were on a 6-point, Likert-type scale that includes 35 items. In this study, Cronbach' alpha coefficient was calculated as .890.

Achievement test for teaching the basic concepts of programming

The academic achievement test was developed by the researcher to determine learning levels of the students related to programming concepts. While developing the test, an indicator chart based on the content that will be provided during implementation was developed. The themes included in the content can be listed as follows: basic concepts of programming (variables, fixed, flow control blocks/condition structures, operators); data representation, parallelism; synchronization; user interactivity; flow control; and interface components of App Inventor. A question data with 40 items were formed to evaluate the behaviors in each level on the indicator chart before developing the achievement test. While developing the pools of items, the questions were constructed according to knowledge, comprehension, application, analysis, synthesis, and evaluation steps. The achievement test was checked by two experts to ensure validity and reliability. Furthermore, a pilot implementation including 25 computer science second-grade students (who have been provided with programming training before) was conducted. The achievement test was developed, by reducing the items in the pool as 20 questions, and the final test was scored as maximum 100 points. Fifty minutes was determined as the exam practice time. According to the feedbacks from the experts and the scores of the students, the values of the item strength were evaluated, and necessary changes were made.

The reliability of the Kuder–Richardson 20 test was calculated, and the reliability coefficient was .77. This value exceeds the recommended R–B flashiness index (Nunnally, 1967). The difficulty index of the items in the test ranges from .29 to .88. The discriminatory index of the items in the tests is between .23 and .56.

Implementation Process and Data Collection

A 14-week procedure was conducted for teaching basic programming concepts in FC in this present study. The programming training tool was selected as App Inventor (http://appinventor.mit.edu).

The lecturer informed the students on the achievements of the course, the implementation process, and the expectations from the students. The students were also provided with an online outline of the course. In addition, in the first lesson, the students were asked to enroll in Edmodo (edmodo.com) and to download it as a mobile application to their smartphones. Thus, students were enabled to get immediate information about the courses and were able to engage courses better. Thanks to this, they were able to do preview before coming to classroom. On the other hand, students staying in pensions did not have PC or tablet computers, while all of them are mobile phone owners.

Throughout the implementation process, the students were asked to watch each week's video before coming to classroom. Resources used in the process can be listed as follows: the lectures themselves and videos prepared by the researcher and videos about the topic on YouTube. The videos pertaining to each week's topic were shared with the students through the Edmodo before that week. The length of the videos varies between 4 and 15 minutes. Preview studies of the students were revised, and the previews were supported by short lectures. Some of the activities performed in the classroom were carried out as group activities. In addition, in-process group discussions and project work were conducted. Table 1 summarizes contents, implementations schedule, procedures, individual activities, group activities, resources, and tools.

Table 1.

Contents and Practices Done Through App Inventor in FC Weeks.

Weeks	Contents of programming training	Learning activities in FC model
Weeks	Contents of programming training	In-class learning activities, tasks, and tools	Out-of-class learning activities, tasks, and tools
Week 1	Mobile programming: Introduction to mobile app invention Programming process Introduction of basic concepts of programming Visual and object-based programming tools and features Introduction of mobile application (App Inventor, Android studio, etc.): Investigation of the structure in App Inventor mobile app creating tool	Registration to online dimension of the course (edmodo.com) and introduction of the tool Presentations Official websites of the apps	Investigation of online dimension of the course (edmodo.com) Notes on basic concepts of programming (PDF) YouTube video introducing mobile programming tools
Week 2	Mobile programming App Inventor setup Essentials of App Inventor usage App Inventor usage in PC and mobile devices (smartphone/tablet)	Introduction of basic concepts' features in mobile app invention tool (interfaces, media, actions, layers, database, resources, services, receivers, and content providers, etc.) Worksheet on discovering commands related to parts	Notes on App Inventor Setup (PDF) Listing the basic components for setup of App Inventor in discussions Creating an account in App Inventor website (http://appinventor.mit.edu) Investigation into sensors in mobile devices and their operating systems Discussion environment
Week 3	Mobile programming-design stage Drawing of algorithms and flow diagrams Usage of emulators	Worksheet on drawing algorithm and flow diagram Practices of emulator usage	Programming task: Reflection of daily tasks in algorithms (this task will be sent to the instructor from homework section of Edmodo homework part.
Week 4	Introducing interface properties of mobile App creating program	Course presentation on the introduction of App Inventor properties (ppt)	Sample practices with App Inventor interface usages—YouTube videos
Week 5–7	Mobile programming-design stage Forming steps in mobile app creating Planning designing of interface in mobile app to be created Designing the mobile app according to principles	Forming procedures of mobile app via group work Designing the interface and arrangements according to principles	Sample practices with App Inventor interface usages –YouTube videos Designing a simple paint interface in App Inventor Discussing design principles to make mobile app more effective
Week 8–10	Mobile programming-coding stage Employment of variables during mobile app creating Employment of condition, decision, and cycle in mobile App Inventor tool	Worksheet Programming tasks	Video on usage of App Inventor variables Video on condition/decision/cycle of App Inventor variables
Week 11–12	Project preparation Debug	Developing an app for a specific purpose via App Inventor Detection of errors in a developed mobile app	Reporting on errors in mobile app, interface objects, and interaction supervision Sharing reports via Edmodo messenger application
Week 13–14	Presentation of projects Sharing projects Conducting data collection tools	Presentation of the purpose, target population, interface properties in the projects Practice on inclusion of the mobile app invented in app stores Filling the data collection tools by the participants	Sending a report on self-projects after watching all presentations via messenger (students were asked to introduce their own projects in terms of usability, functionality, originality, interface design, interactions, and error detection).

Note. FC = flipped classroom.

The students were reminded to do previews before the course through Edmodo. Worksheets and programming exercise tasks were employed so as to allow students to do more practice on applications. Figure 5 summarizes examples of materials in the course.

Figure 5.

Examples of materials in the course.

Project development activities were conducted between 11th and 14th weeks of the procedure. Also during the 14th week of the implementation process, the data collection instruments were applied to the study group in an online environment.

Data Analysis

SEM was employed to reveal a model determining, explaining, and predicting the relations between programming achievement levels of university students and different variables. There have been several experimental studies on the effect of programming training on PSS. However, very few researchers have investigated into FC model in programming training. On the other hand, the relation between e-learning readiness and programming achievement has not been addressed yet. This study is to fill this gap in the literature, by focusing on all these variables altogether, and SEM was employed so that the study can be more original one. Linear Structural Relations (LISREL) 8.51 program was used in data analysis. LISREL program enables SEM, confirmatory factor analysis, path analysis, and so forth to be conducted. Chi-square goodness-of-fit test (χ²), root mean square error of approximation (RMSEA), goodness-of-fit index (GFI), comparative fit index (CFI), normed fit index (NFI), and nonnormed fit index (NNFI) were employed to be able to demonstrate the compatibility level of the related patterns in the proposed model. Goodness-of-fit coefficients and acceptable range of values were demonstrated along with their reference values. In this study, RMSEA value of the model was found as 0.016. Because the calculated RMSEA value is in the range of 0 < RMSEA < 0.05 (Schermelleh-Engel, Moosbrugger, & Müller, 2003), this indicates a good model fit. Similarly, because calculated NFI value is 0.95, which is in the range of 0.95 ≤ NFI ≤ 1, and calculated NNFI value is 0.95, which is in the range of 0.95 ≤ NNFI ≤ 0.97 (Tabachnick & Fidell, 2007), this indicates a good model fit. As calculated GFI value is 0.97, which in the range of 0.90 ≤ GFI ≤ 0.95 (Tabachnick & Fidell, 2007), this indicates an acceptable fit. In addition, the fit indexes of the model (chi-square, RMSEA, NNFI, NFI, CFI, and GFI) were found as significant at .05 level.

Findings

Descriptive Statistics

In this part of the study, the findings and interpretations were included. The descriptive statistics values of the students' programming achievement, cognitive flexibility, PSS, and FLR levels are demonstrated in Figures 6 and 7.

Figure 6.

Descriptive statistics on students' programming achievement, cognitive flexibility, and PSS.

Figure 7.

Descriptive statistics on students' FLR levels.

When Figure 6 is examined, the academic achievement levels of the students were assessed based on 100 points. It was found that programming achievement scores of males (M = 78.54, SD = 18.77) were higher than that of females (M = 77.11, SD = 21.45). The general programming achievement score mean of the students is 77.75. When it comes to cognitive flexibility scores of the students, it was found that cognitive flexibility scores of females (M = 51.36, SD = 8.52) were higher than that of males (M = 48.22, SD = 8.57). The general cognitive flexibility score mean of the students is 49.96. PSS scores of males (M = 137.27, SD = 21.97) were higher than that of females (M = 134.93, SD = 22.23). The general PSS scores mean of the students is 135.98.

When Figure 7 is examined, it was found that FLR scores of females (M = 73.22, SD = 14.21) were higher than that of males (M = 70.67, SD = 14.05). The general FLR level scores mean of the students is 72.08. With respect to the subdimensions of FLR, it was found that technology self-efficacy subdimension (M/k = 2.82) was higher than other subdimensions, whereas motivation for learning subdimension (M/k = 2.64) was lower than other subdimensions.

Correlation and Path Analyses

Correlation values are demonstrated in Table 2.

Table 2.

Correlations Matrix.

	1	2	3	4	5	6	7	8	9
1. Programming achievement	1
2. Cognitive flexibility	.168*	1
3. PSS	.272*	.356**	1
4. FLR	.290*	.224*	.204	1
5. FLR-1	.147	.161	.140	.802**	1
6. FLR-2	.314*	.349*	.302*	.912**	.552**	1
7. FLR-3	.175*	.168*	.109	.803**	.458**	.764**	1
8. FLR-4	.168*	.186*	.123	.448**	.147	.386**	.404**	1
9. FLR-5	.291*	.231*	.257*	.525**	.175*	.465**	.489**	434**	1

Note. PSS = problem-solving skills; FLR = flipped learning readiness; FLR-1 = learner control and self-directed learning; FLR-2 = technology self-efficacy; FLR-3 = communication self-efficacy; FLR-4 = motivation for learning; FLR-5 = doing previews.

Correlation is significant at the .01 level (two-tailed).

Correlation is significant at the .05 level (two-tailed).

As it can be seen in Table 2, the correlation values between programming achievement levels and other variables can be listed as follows: programming achievement—cognitive flexibility (r = .168, p < .05), programming achievement—PSS (r = .272, p < .05), and programming achievement—FLR (r = .290, p < .05). It is considered that correlation coefficients show a high-level relation when they are between .07 and 1.00, medium-level relation when they are between .70 and .30, and low-level relation when they are between .30 and .00 (Büyüköztürk, 2009). The data appear to suggest that there is a statistically significant positive low relation between programming achievement and other variables.

When the SEM model is examined in Table 3, it can be seen that chi-square goodness-of-fit test is significant (p = .000). When the variables related to programming achievement are assessed, it can be seen that “FLR” is the highest variable with linkage coefficient (γ = .36), and it is also the most important one, followed by “PSS” (γ = .22) and “cognitive flexibility” (γ = .17). When the linkage coefficients are examined, the relative important order of predictive variables on FLR can be listed as follows: FLR, PSS, and cognitive flexibility. When the linkage coefficients between programming achievement and FLR subdimensions are assessed, it can be seen that “technology self-efficacy” is the highest variable with linkage coefficient (γ = .38), and it is also the most important one, followed by “doing previews” (γ = .38), “communication self-efficacy” (γ = .16), “learner control and self-directed learning” (γ = .14), and “motivation for learning” (γ = .11). When the linkage coefficients are examined, the relative important order of FLR predictive variables on programming achievement can be listed as follows: “technology self-efficacy, doing previews, communication self-efficacy, learner control and self-directed learning, motivation for learning.”

Table 3.

Hypothesis Acceptance/Rejection Table.

Way/ relation		Stand. Coefficients	t value	Acceptance/ rejection
H1	FLR → Programming achievement	.36	14.60	Accept
H1a	FLR-1 → Programming achievement	.14	2.80	Accept
H1b	FLR-2 → Programming achievement	.47	16.41	Accept
H1c	FLR-3 → Programming achievement	.16	3.84	Accept
H1d	FLR-4 → Programming achievement	.11	2.03	Accept
H1e	FLR-5 → Programming achievement	.38	15.56	Accept
H2	Cognitive flexibility → Programming achievement	.17	3.85	Accept
H3	PSS → Programming achievement	.22	4.90	Accept

Note. FLR = flipped learning readiness; PSS = problem-solving skills; FLR-1 = learner control and self-directed learning; FLR-2 = technology self-efficacy; FLR-3 = communication self-efficacy; FLR-4 = motivation for learning; FLR-5 = doing previews.

As it can be seen in Table 3, eight hypotheses were attempted to test in this study. The data yielded by this study provide convincing evidence that all hypotheses presented were supported (p < .05). The direct effect of FLR and its subdimensions on programming achievement were found as significant (p < .05). The hypotheses between FLR and its subdimensions and programming achievement (H1, H1a, H1b, H1c, H1d, and H1e) were accepted. The direct effect of cognitive flexibility on programming achievement was found as significant (p < .05). The related hypothesis (H2) was accepted. The direct effect of PSS on programming achievement was found as significant (p < .05). The related hypothesis (H3) was accepted. Accordingly, returning to the hypotheses posed at the beginning of this study, it is now possible to state that FLR and its indicators, cognitive flexibility, and PSS level are significant in the academic achievement in programming training in FCs.

In this present study, the explanatory power of the model was found as 0.58. Approximately 60% of variance in academic success scores are explained by current predictive variables.

A Mediation Analysis of the Impact of Past Information and Communications Technology Usage Experience on Academic Achievement

To check the mediating effect of the variables, a new SEM was followed (Mueller & Hancock, 2008). A measurement model was conducted to examine the relation between variables in the study. The measurement model formed via all variables showed a good model fit. The coefficients indicate that the model has a good model fit (p < .001, RMSEA = 0.036, NFI = 0.99, NNFI = 0.97, GFI = 0.94). Following this, SEM was conducted to test the effect of mediating variables related to effect of PSS, FLR, and cognitive flexibility on programming achievement of the students. A review of the literature has revealed that past Information and Communications Technology (ICT) usage experiences has been considered as a variable that can affect teaching in FC even though it has not been employed as a direct variable (Kong, 2014; Yildiz-Durak, 2018b). Although investigation of past ICT usage experiences is not one of the purposes of this present study, its mediating effect has been investigated because it has been thought that it may have been effective as an external factor. Coefficients of SEM are shown in Figure 8.

Figure 8.

Mediating effect analysis of past ICT usage experiences on academic achievements. The arrows in bold indicate that t value is significant.

When the SEM is examined in Figure 8, it can be seen that chi-squared model goodness-of-fit test was found to be significant (p = .000). The correlation coefficient between programming achievement and past ICT usage experience was found to be γ = .26. Based on the analysis in this article, it can be concluded that past ICT usage experience is a significant predictor of programming achievement. When the variables related to past ICT usage experience are examined, the highest correlation coefficients were found as ‘technology self-efficacy’ (γ = .30) and ‘communication self-efficacy’ (γ = .25), respectively. In summation, it can be concluded that past ICT usage experience has a direct and mediating effect on programming achievement.

Discussion

The findings of this study revolve around predictive variables that have an impact on programming achievement and achievement scores of the students in programming training in FC. This present study provides confirmatory evidence that FLR is the most significant predictor of programming achievements of the students in FC. On these grounds, it can be argued that other significant predictors are PSS and cognitive flexibility orderly. The model of the research validates the view that any increase/development in FLR levels, PSS, and cognitive flexibility levels of the students will enhance programming achievements of the students in FC.

Programming Achievements of the Students in FC

On the basis of the evidence currently available, it seems fair to suggest that programming achievements of the students during programming training process in FC model were at good level. At this precise point, on logical grounds, it can be argued that FC model affects programming achievements of the students positively. Within this context, feedback was provided to the learners by the lecturer in online environment to enhance the engagements of the students in FC model. Students had the opportunity to ask questions not only to the lecturer but also to the peers in the discussion part as well as messenger part of Edmodo. It is believed that studies on strengthening the student-content–student-lecturer interaction are positively reflected in the programming success. On the other hand, the presentation of extracurricular videos and course content on the FC model allows for more time for the students to practice and active learning. Learning programming concepts outside the classroom and the progress of the course by watching videos on applications affect learners' attitudes toward difficult tasks in a positive way, and this affects the programming success, as well. Similarly, in a study on FC model employed in programming training at university level by Puarungroj (2015), it was found that students' engagement and motivation levels have increased. Besides, it was added that the problems experienced beforehand disappeared, and students' academic achievements and PSS have developed. The author stated that the instructor must attach great importance for determination of the selection of the video content and online books so as to enhance the effectiveness of programming training in FC. In addition, the online environment is paid attention so that there will not be any technical problem. Online quizzes have been found to be effective in increasing engagement of the students in the study. Pawelczak (2017) compared the differences between programming training in FC model and traditional methods and concluded that the motivation level of the participants increased. Students reported that activities for academic purposes are much more than the traditional methods. On the other hand, it was found that theoretic part of the teaching was reported as weak when compared with traditional methods. An another study at K-12 level, by Tugun et al. (2017), compared FC model and traditional model in block-based programming environment and concluded that programming achievement was found to be much more than the traditional method. In addition to this, the motivations of the students were found to be higher in FC model. It was finally suggested that an integration model as to how FC model can be used in programming training must be developed. Elmaleh and Shankararaman (2017) investigated FC model in programming training at college level based on three factors: learning levels, final exam scores, and competency and feedback levels. The findings revealed that there are more students who got higher scores in the final exams in FC model, and their competencies have increased. Besides, FC model provides students with feedback opportunities.

Based on the extensive literature review, it has been revealed that FC model increases the achievement level in programming courses. This present study corroborated with the existing literature. Taken together, it can be noted that online learning environment should be clear and understandable in programming training via FC model. It is necessary to attach importance for videos, and e-resources should be simple and easy to understand. In addition, employing online surveys, quizzes, and messenger applications will add to students' engagement.

The programming achievement was found to be higher in males rather than females. This is consistent with previous results (Boechler, Dragon, & Wasniewski, 2014) in that such variables as gender and age have an impact on the performances of the individuals in computer-based environments. It can be argued that this circumstance will make a difference in programming performance either directly or through other means because lots of cognitive and affective skills differ among individuals based on gender. Accordingly, in tasks where these skills need to be exercised during many skillful programming processes, performance will also vary depending on personal characteristics and variables.

FLR Levels and Programming Achievements of the Students

The scores of computer science students in FLR and its subdimensions in FC were examined. According to the findings of this study, the participants had the highest level of readiness level in technology self-efficacy subdimension, followed by learner control and self-directed learning, communication self-efficacy, and doing previews. The lowest FLR level found in the data is motivation for learning subdimension. FLR levels of the students were found to be medium.

The fact that the readiness levels of the students were high in technology self-efficacy subdimension in FLR and this dimension has a strong effect on programming achievement means that technological self-efficacy skills need to be developed in programming training. This is corroborated by Yildiz-Durak (2018b) who argues that students with a high degree of technology self-efficacy are equipped with the skills required for online learning on the FC model, such as the use of computers, online information search, and the execution of basic software functions, and these skills enable students to be ready to take course content. This contributes to students' programming skills in a positive way. On the other hand, Durak (2016) proposes that students with high technology self-efficacy tend to have a higher programming learning tendency and need to improve the technology self-efficacy of students to achieve success in programming learning. For this reason, it is suggested by Hung et al. (2010) that lecturers should prepare specific guidance activities for online learning or give special support to students who need training. In this regard, providing online instruction in programming training in the FC model and providing technical training and support to the students in the online environment will both increase the readiness level to the learning process and reduce the possibility of encountering future technical difficulties. In addition, in this study, students must be given immediate feedback on technical issues to prevent different technology self-sufficiency levels from being an obstacle to teaching.

Findings clearly indicate that doing previews in readiness subdimension was ranked second in terms of predicting programming achievement in FC. In the FC model, revising extracurricular materials is indispensable for an active learning (Peled et al., 2015). Students who do not do any previews before the lesson are less likely to actively participate in learning activities during the course. This is thought to negatively affect the performance of the students' programming. For this reason, it is important to investigate the out-of-class resources first to ensure programming achievement in the FC model. However, each student may not have necessary conditions to do preview. For example, some students stay in pensions and do not have Internet access. Therefore, it must be taken into consideration that this situation may adversely affect programming achievement. On the other hand, there is a great deal of responsibility for the lecturer to make the students' extracurricular preparations. First of all, it is important that the prepared materials are interesting, that they are of ideal length, that immediate feedback is given to prevent students from leaving the online system, and that student activities are monitored on the online system. In addition, the lecturer may give notifications to the students at certain intervals to prepare the course contents online. In this study, Edmodo mobile application was used to provide extracurricular studies, and notifications were sent to the students 1 day ago by the lecturer. Yildiz-Durak (2017) mentions that learner control and self-directed learning subdimensions FLR were very important due to the existence of extracurricular activities in the FC model. This thinking is hardly distinguishable from Poole (2000) who demonstrates that in online learning environments, self-regulatory learning and student control are important to ensure effective instruction.

In online learning environments, factors such as online games, social media can easily distract student attention, and students may be prone to cyber loafing during the course. In this case, students must have self-discipline so as to engage in extracurricular activities. The FC model includes both online and face-to-face learning dimensions. For this reason, communication self-efficacy of students influences the level of simultaneous or asynchronous interaction with course contents, peers, and lecturers in both online and face-to-face settings. Interaction is considered as a preliminary condition for learning. McLaughlin et al. (2013) note that FC model increases the interaction between the students and learning resources. In consistent with this, Roper (2007) proposes that students with higher online self-efficacy were found to feel more comfortable with active participation in the course.

In this study, it was found that motivation subdimension of FLR is less effective than others in programming achievement in FC model, and the readiness level of the students was found as low level. This issue was addressed by Ryan and Deci (2000), concluding that the fact that students have motivational orientation toward learning has significant influence on their learning performance. Motivation for learning subdimension in FC model is a driving force that affects learners' efforts to learn online content with their own and to improve learning performance. Understanding the level of motivation and preferences of learners to learn is very important to improve the planning, production, and implementation of educational resources (Hao, 2016a, 2016b).

It is considered that the students who are taking programming education in the present study are less open to new learning applications and that the basic characteristics of FC model and the level of active learning activities performed in class are effective on the result obtained. It is therefore possible to suggest that more collaborative project studies should be included to increase students' motivation for learning. As noted by Day and Foley (2006), active learning is one of the most effective ways to increase motivation. In the FC model, it is thought that devoting classroom time to interactive events and the ability of lecturers to present materials in different genres and address them to individual differences motivate students to take their own learning responsibilities. Similarly, it may have motivated the students to reach the learning resources of the students whenever they want and to advance according to their individual speed. On the other hand, Hamdan, McKnight, McKnight, and Arfstrom (2013) suggest that in the flipped class model of the students, the previews outside the class before the lesson positively reflect the motivation of the students. Enfield (2013) recommends that weekly quiz type of exams should be conducted to increase the learning motivation in FC model so that the students will have a strong motivation to watch videos.

PSS Levels of Students and Programming Achievements

It is not enough to know and apply only programming codes when preparing problem solutions in programming processes. PSS should be developed so that programming processes can be learned. As Gomes and Mendes (2007) remind us a reason for this, programming learning is considered as a process involving problem-solving stages. The FC model is also a method that contributes to the development of PSS. The FC model provides continuous access to the information background required for problem-solving with a structure that is adaptable to individual differences (Bishop & Verleger, 2013). In the study conducted by Wiginton (2013), it was suggested that FC model increases learners' self-efficacy beliefs related to themselves.

This situation enables students to overcome challenging tasks and to solve problems. What is more, face-to-face and online lessons in the FC model make a positive contribution to the development of basic computer skills and to writing, running, and debugging programming codes (Yildiz-Durak, 2018b). This affects programming self-efficacy in a positive way, and therefore, it is thought that it will increase the programming achievement. In addition to this, the self-efficacy beliefs of the learners who actively produce their products are also developing. Enfield (2013) concurs well with this and puts forward that FC model allows learners to find more application opportunities, to get feedback in different forms, and to improve learning performance by improving students' beliefs about themselves.

Cognitive Flexibility Levels of Students and Programming Achievements

This present study supported the hypothesis that cognitive flexibility level is a predictor of programming achievement of students. Even though there has not been any research focusing on the role of cognitive flexibility level in programming training in current literature, Alper and Deryakulu (2008) argued that cognitive flexibility is a predictor on academic achievement and retention in web-based learning environments. Timarová and Salaets (2011) lend support to this and put forward that cognitive flexibility is effective on the stress experienced during learning process and those with high level of cognitive flexibility experience less stress than others. Accordingly, this situation affects learning attitudes, motivations, and academic achievements of the students indirectly. In the FC model, it is thought that the students with high level of cognitive flexibility will have higher interaction, can lead their own learning process, and therefore can make maximum use of FC. This is in complete agreement with Martin and Anderson (1998) who highlight that self-efficacy perception related to a problem encountered is linked to cognitive flexibility. Çelikkaleli (2014) also suggested that being cognitively flexible allows us to see cognitive, affective, and behavioral alternatives in problem-solving and to generate new paths for solution. This situation also affects programming achievement positively. Martin, Anderson, and Thweatt (1998) draw attention that cognitive flexibility allows the individual to establish a balanced relationship with other people. On the other hand, in the learning environments, it is thought that the cognitively flexible individuals will use the interaction ways more effectively, and they will be positively reflected in the learning success.

Mediating Effect

The findings of this study clearly indicate the mediating effect of past ICT usage experience. Several studies in the literature have concluded that ICT usage experience, especially in digital classrooms, contributes to classroom activities (Kostaris, Sergis, Sampson, Giannakos, & Pelliccione, 2017). Kong (2014), on the other hand, suggests that this contribution is two-way. That is to say, ICT usage experience in the beginning in digital classrooms affects the activities in learning process, while digital classrooms develop ICT usage experiences. As noted by Johnson and Renner (2012), ICT usage experience enables student to take responsibility of learning, increasing the perception of ease of use. It also decreases leaving the classroom due to technical problems and need of technical support after the course.

Implications

The results have further strengthened our confidence in hypothesis that FLR and its subdimensions, cognitive flexibility, and PSS level are significant predictors of the academic achievement in programming training in FC. This has provided strong evidence for the assertion that FLR, PSS, and cognitive flexibility levels of the students must be taken into consideration, and they are attempted to develop so as to enhance the programming achievement of the students and FC model in FC. The findings of this present study puts forward that technology self-efficacy should be attached great importance for the programming achievement of the students in FC. The lecturer may be supposed to implement activities that are directed to boost technology use, knowledge, and attitudes of the students to increase the success of programming training in FC model. The increase in FLR levels of the students in programming training in FC means that this will affect the programming achievement of the students in a positive way. From moving this point, it can be suggested that further research in this area may include studies on the development of FLR levels before the instruction.

Limitations and Recommendations

We are aware that our work clearly has some limitations. The participants may have had different levels of preliminary knowledge on programming and different levels of attitudes to programming depending on the high school types. Therefore, programming achievements can be examined through an experimental research design, by determining the preliminary readiness of the students as moderator variable in FC model. The present study has employed only Edmodo mobile application. To put it differently, the effects of the mobile e-learning environment have been overlooked. Future work can be compared with experimental studies in which the programming success is differentiated when FC model is used with mobile FC. Based on the level of cognitive flexibility, the participant's engagement in FC can be examined in future studies on how interaction preferences change and how this reflects programming success. The current study was limited by quantitative data collection. In other words, in online and face-to-face learning in FC, qualitative research can be done on variables that are thought to be influential in programming success and subdimensions of FLR.

Footnotes

Authors' Note

Our data are not yet available online in any institutional database.

Ethical Statement

The research was conducted in a school in Turkey and approved by the school administration. Participation was voluntary and anonymous. Informed consent was obtained from all participants.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Hatice Yildiz Durak is currently an assistant professor in the Department of Computer Education and Instructional Technology at Bartın University. She has started to work in the Department of Computer Education and Instructional Technology at Bartın University since 2016. Her research interests are as follows: programming training in early child education, e-learning, professional development, problematic internet usage.

References

Albert

Beatty

B. J.

(2014) Flipping the classroom applications to curriculum redesign for an introduction to management course: Impact on grades. Journal of Education for Business 89(8): 419–424. doi:10.1080/08832323.2014.929559.

Alper

Deryakulu

(2008) The effect of cognitive flexibility on students' achievement and attitudes in web mediated problem based learning. Education and Science 33(148): 49–63.

Alsancak-Sirakaya, D. (2015). The effect of flipped classroom model on academic achievement, self-directed learning readiness and motivation (Unpublished doctoral dissertation). Gazi University, Ankara.

Andrade, M., & Coutinho, C. (2016, November). Implementing flipped classroom in blended learning environments: A proposal based on the cognitive flexibility theory. In E-learn: World conference on e-learning in corporate, government, healthcare, and higher education (pp. 1115–1125). Waynesville, NC: Association for the Advancement of Computing in Education.

Apiola

Tedre

(2012) New perspectives on the pedagogy of programming in a developing country context. Computer Science Education 22(3): 285–313. doi:10.1080/08993408.2012.726871.

Awidi

I. T.

Paynter

(2019) The impact of a flipped classroom approach on student learning experience. Computers & Education 128: 269–283. doi:10.1016/j.compedu.2018.09.013.

Bandura

(1977) Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review 84(2): 191–215. doi:10.1037/0033-295X.84.2.191.

Batting

W. T.

(1979) Are the important “individual differences” between or within individuals?. Journal of Research in Personality 13: 546–558. doi:10.1016/0092-6566(79)90015-1.

Bergmann, J., & Sams, A. (2012). Flip your classroom: Reach every student in every class every day. Washington, DC: International Society for Technology in Education. Retrieved from http://books.google.com/books?id=nBi2pwAACAAJ.

10.

Bishop, J. L., & Verleger, M. A. (2013, June). The flipped classroom: A survey of the research. ASEE National Conference Proceedings, Atlanta, GA, 30(9), 1–18.

11.

Boechler

Dragon

Wasniewski

(2014) Digital literacy concepts and definitions: Implications for educational assessment and practice. International Journal of Digital Literacy and Digital Competence (IJDLDC) 5(4): 1–18. doi:10.4018/ijdldc.2014100101.

12.

Burke

D. D.

(2015) Scale-up! Classroom design and use can facilitate learning. The Law Teacher 49(2): 189–205. doi:10.1080/03069400.2015.1014180.

13.

Büyüköztürk, Ş. (2009). Sosyal bilimler için veri analizi el kitabı: İstatistik, araştırma deseni, SPSS uygulamaları ve yorum (9. baskı). Ankara: Pegem Yayınları.

14.

Çakiroğlu

Ü.

Öztürk

(2017) Flipped classroom with problem based activities: Exploring self-regulated learning in a programming language course. Journal of Educational Technology & Society 20(1): 337–349.

15.

Canas

Quesada

Antolí

Fajardo

(2003) Cognitive flexibility and adaptability to environmental changes in dynamic complex problem-solving tasks. Ergonomics 46(5): 482–501. doi:10.1080/0014013031000061640.

16.

Çelen, F. K., Çelik, A. & Seferoğlu, S. S. (2011, September 9–10). Online learning in higher education: Problems faced in the system and possible solutions to solve these problems. Paper presented at the 1st Educational Sciences Graduate Students Research Symposium, Erciyes University, Kayseri.

17.

Çelikkaleli

Ö.

(2014) The validity and reliability of the cognitive flexibility scale. Education and Science 39(176): 339–346.

18.

Chen

E. Z.

(2014) Flipped classroom model and its implementation in a computer programming course, Lärarlärdom: Högskolepedagogisk Konferens 2014.

19.

Chen

Wang

Chen

N. S.

(2014) Is FLIP enough? Or should we use the FLIPPED model instead?. Computers & Education 79: 16–27. doi:10.1016/j.compedu.2014.07.004.

20.

Cheng, J., & Koszalka, T. A. (2016). Cognitive flexibility theory and its application to learning resources. Retrieved from http://ridlr.syr.edu/wp-content/uploads/2016/09/CFT_FInal.pdf.

21.

Chuang

H. H.

Weng

C. Y.

Chen

C. H.

(2018) Which students benefit most from a flipped classroom approach to language learning?. British Journal of Educational Technology 49(1): 56–68. doi:10.1111/bjet.12530.

22.

Connell

G. L.

Donovan

D. A.

Chambers

T. G.

(2016) Increasing the use of student-centered pedagogies from moderate to high improves student learning and attitudes about biology. CBE—Life Sciences Education 15(1): 1–15. doi:10.1187/cbe.15-03-0062.

23.

Cooper

Dann

Pausch

(2000) Alice: A 3-D tool for introductory programming concepts. Journal of Computing Sciences in Colleges 15(5): 107–116.

24.

Davies

R. S.

Dean

D. L.

Ball

(2013) Flipping the classroom and instructional technology integration in a college-level information systems spreadsheet course. Educational Technology Research and Development 61(4): 563–580. doi:10.1007/s11423-013-9305-6.

25.

Day

J. A.

Foley

J. D.

(2006) Evaluating a web lecture intervention in a human–computer interaction course. IEEE Transactions on Education 49(4): 420–431. doi:10.1109/TE.2006.879792.

26.

Demir, O. (2015). The investigation of e-learning readiness of students and faculty members: Hacettepe University Faculty of Education example [Unpublished Master Thesis]. Hacettepe University, Ankara.

27.

Demir

Ö.

Yurdugül

(2015) The exploration of models regarding e-learning readiness: Reference model suggestions. International Journal of Progressive Education 11(1): 173–194.

28.

Dennis

J. P.

Vander Wal

J. S.

(2010) The cognitive flexibility inventory: Instrument development and estimates of reliability and validity. Cognitive Therapy and Research 34(3): 241–253. doi:10.1007/s10608-009-9276-4.

29.

Durak, H. (2016). Design and development of an instructional program for teaching programming process to gifted students (Unpublished doctoral dissertation). Gazi University, Ankara.

30.

Elmaleh, J., & Shankararaman, V. (2017, April). Improving Student Learning in an Introductory Programming Course Using Flipped classroom and Competency Framework. 2017 IEEE Global Engineering Education Conference (EDUCON), Athens, Greece.

31.

Enfield

(2013) Looking at the impact of the flipped classroom model of instruction on undergraduate multimedia students at CSUN. TechTrends 57(6): 14–27. doi:10.1007/s11528-013-0698-1.

32.

Fitzgerald

G. E.

(1997) An interactive multimedia program to enhance teacher problem-solving skills based on cognitive flexibility theory: Design and outcomes. Journal of Educational Multimedia and Hypermedia 6(1): 47–76.

33.

Francl

T. J.

(2014) Is flipped learning appropriate. Journal of Research in Innovative Teaching 71: 119–128.

34.

Freeman, S., Eddy, S. L., McDonough, M., Smith, M. K., Okoroafor, N., Jordt, H., & Wenderoth, M. P. (2014). Active learning increases student performance in science, engineering, and mathematics. Proceedings of the National Academy of Sciences, 111(23), 8410–8415.

35.

Giannakos, M. N., Krogstie, J., & Chrisochoides, N. (2014, November). Reviewing the flipped classroom research: Reflections for computer science education. In Proceedings of the computer science education research conference (pp. 23–29). New York, NY: ACM. Retrieved from https://dl.acm.org/citation.cfm?id=2691354.

36.

Gomes, A., & Mendes, A. J. (2007, September). Learning to program-difficulties and solutions. Paper presented at the International Conference on Engineering Education–ICEE, Coimbra, Portugal.

37.

Gundurao

H. K.

Manjunath

N. S.

Nachappa

M. N.

(2010) Computer technology and computer programming, Mumbai, India: Global Media.

38.

Hamdan, N., McKnight, P., McKnight, K., & Arfstrom, K. M. (2013). The flipped learning model: A white paper based on the literature review titled a review of flipped learning. Flipped Learning Network, Pearson, George Mason University, Virginia, USA.

39.

Hao

(2016a) Exploring undergraduates' perspectives and flipped learning readiness in their flipped classrooms. Computers in Human Behavior 59: 82–92. doi:10.1016/j.chb.2016.01.032.

40.

Hao

(2016b) Middle school students' flipped learning readiness in foreign language classrooms: Exploring its relationship with personal characteristics and individual circumstances. Computers in Human Behavior 59: 295–303. doi:10.1016/j.chb.2016.01.031.

41.

Heppner

P. P.

Petersen

C. H.

(1982) The development and implications of a personal problem-solving inventory. Journal of Counseling Psychology 29(1): 66–75. doi:10.1037/0022-0167.29.1.66.

42.

Herala, A., Vanhala, E., Knutas, A., & Ikonen, J. (2015, November). Teaching programming with flipped classroom method: A study from two programming courses. In P. Kinnunen, & J. Sheard (Eds.) Proceedings of the 15th Koli Calling Conference on Computing Education Research (pp. 165–166). New York, NY: ACM. Retrieved from https://dl.acm.org/citation.cfm?id=2828983.

43.

Horton, D. & Craig, M. (2015, February). Drop, fail, pass, continue: Persistence in CS1 and beyond in traditional and inverted delivery. In A. Al-Zubidy , J. C. Carver, S., Heckman, & M. Sherriff (Eds.) Proceedings of the 46th ACM Technical Symposium on Computer Science Education (pp. 235–240). New York, NY: ACM. Retrieved from https://dl.acm.org/citation.cfm?id=2677273.

44.

Hung

M. L.

Chou

Chen

C. H.

Own

Z. Y.

(2010) Learner readiness for online learning: Scale development and student perceptions. Computers & Education 55(3): 1080–1090. doi:10.1016/j.compedu.2010.05.004.

45.

Hwang

G. J.

Lai

C. L.

Wang

S. Y.

(2015) Seamless flipped learning: A mobile technology-enhanced flipped classroom with effective learning strategies. Journal of Computers in Education 2(4): 449–473. doi:10.1007/s40692-015-0043-0.

46.

Jacobson

M. J.

Spiro

R. J.

(1994) A framework for the contextual analysis of technology-based learning environments. Journal of Computing in Higher Education 5(2): 3–32. doi:10.1007/BF02948569.

47.

Johnson, L., & Renner, J. (2012). Effect of the flipped classroom model on secondary computer applications course: Student and teacher perceptions, questions and student achievement (Unpublished doctoral dissertation). University of Louisville, Louisville, KY.

48.

Jonassen

D. H.

(1997) Instructional design models for well-structured and III-structured problem-solving learning outcomes. Educational Technology Research and Development 45(1): 65–94. doi:10.1007/BF02299613.

49.

Karaca

Ocak

M. A.

(2017) Effects of flipped learning on university students' academic achievement in algorithms and programming education. International Online Journal of Educational Sciences 9(2): 527–543. doi:10.15345/iojes.2017.02.017.

50.

Karakus, M., Uludag, S., Guler, E., Turner, S. W., & Ugur, A. (2012, June). Teaching computing and programming fundamentals via App Inventor for Android. Information Technology Based Higher Education and Training (ITHET), 2012 International Conference, İstanbul, Turkey.

51.

Karasar

(2005) Scientific research method, Ankara, Turkey: Nobel Publishing.

52.

Kelleher

Pausch

(2005) Lowering the barriers to programming: A taxonomy of programming environments and languages for novice programmers. ACM Computing Surveys (CSUR) 37(2): 83–137. doi:10.1145/1089733.1089734.

53.

Kim

M. K.

Kim

S. M.

Khera

Getman

(2014) The experience of three flipped classrooms in an urban university: An exploration of design principles. The Internet and Higher Education 22: 37–50. doi:10.1016/j.iheduc.2014.04.003.

54.

Klegeris

Hurren

(2011) Impact of problem-based learning in a large classroom setting: Student perception and problem-solving skills. Advances in Physiology Education 35(4): 408–415. doi:10.1152/advan.00046.2011.

55.

Kong

S. C.

(2014) Developing information literacy and critical thinking skills through domain knowledge learning in digital classrooms: An experience of practicing flipped classroom strategy. Computers & Education 78: 160–173. doi:10.1016/j.compedu.2014.05.009.

56.

Kostaris, C., Sergis, S., Sampson, D. G., Giannakos, M. N., & Pelliccione, L. (2017). Investigating the potential of the flipped classroom model in K-12 ICT teaching and learning: An action research study. Educational Technology & Society, 20(1), 261–273. Retrieved from https://www-jstor-org.web.bisu.edu.cn/stable/jeductechsoci.20.1.261.

57.

Lazakidou

Retalis

(2010) Using computer supported collaborative learning strategies for helping students acquire self-regulated problem-solving skills in mathematics. Computers & Education 54(1): 3–13. doi:10.1016/j.compedu.2009.02.020.

58.

Lepp, M., & Tonisson, E. (2015, January). Integrating flipped classroom approach and work in pairs into workshops in programming course. The Steering Committee of The World Congress in Computer Science, Computer Engineering and Applied Computing (WorldComp), Las Vegas, Nevada, USA.

59.

Martin

M. M.

Anderson

C. M.

(1998) The cognitive flexibility scale: Three validity studies. Communication Reports 11(1): 1–9. doi:10.1080/08934219809367680.

60.

Martin

M. M.

Anderson

C. M.

Thweatt

K. S.

(1998) Individuals' perceptions of their communication behaviors: A validity study of the relationship between the Cognitive Flexibility Scale and the Communication Flexibility Scale with aggressive communication traits. Journal of Social Behavior and Personality 13(3): 531–540.

61.

Martin

M. M.

Rubin

R. B.

(1995) A new measure of cognitive flexibility. Psychological Reports 76(2): 623–626. doi:10.2466/pr0.1995.76.2.623.

62.

McLaughlin

J. E.

Griffin

L. M.

Esserman

D. A.

Davidson

C. A.

Glatt

D. M.

Roth

M. T.

Mumper

R. J.

(2013) Pharmacy student engagement, performance, and perception in a flipped satellite classroom. American Journal of Pharmaceutical Education 77(9): 1–8. doi:10.5688/ajpe779196.

63.

Mueller, R. O., & Hancock, G. R. (2008). Best practices in structural equation modeling. In J. W. Osborne, (Ed.), Best practices in quantitative methods (pp. 488–508). Thousand Oaks, CA: Sage.

64.

Nunnally, J. (1978). Psychometric methods. New York: McGraw-Hill.

65.

Palloff

R. M.

Pratt

(1999) Building learning communities in cyberspace, San Francisco, CA: Jossey-Bass.

66.

Papert, S. (1980). Mindstorms: Children, computers, and powerful ideas. New York, NY: Basic Books.

67.

Pawelczak, D. (2017, June). Comparison of traditional lecture and flipped classroom for teaching programming. 3rd International Conference on Higher Education Advances. Editorial Universitat Politècnica de València, Spain. 391–398. doi:10.4995/HEAD17.2017.5226.

68.

Peled

Blau

Grinberg

(2015) Does 1: 1 computing in a junior high-school change the pedagogical perspectives of teachers and their educational discourse. Interdisciplinary Journal of e-Skills and Lifelong Learning-IJELL 11: 257–271.

69.

Pintrich

P. R.

(2000) An achievement goal theory perspective on issues in motivation terminology, theory, and research. Contemporary Educational Psychology 25(1): 92–104. doi:10.1006/ceps.1999.1017.

70.

Poole

D. M.

(2000) Student participation in a discussion-oriented online course: A case study. Journal of Research on Computing in Education 33(2): 162–177.

71.

Portelance

D. J.

Strawhacker

A. L.

Bers

M. U.

(2016) Constructing the ScratchJr programming language in the early childhood classroom. International Journal of Technology and Design Education 26(4): 489–504. doi:10.1007/s10798-015-9325-0.

72.

Puarungroj, W. (2015, December). Inverting a computer programming class with the flipped classroom. Paper presented at the Proceedings of the International Conference on eLearning for Knowledge-Based Society, Bangkok Siam Technology College, Bangkok, Thailand.

73.

Rizvi

Humphries

Major

Jones

Lauzun

(2011) A CS0 course using Scratch. Journal of Computing Sciences in Colleges 26(3): 19–27.

74.

Roach

(2014) Student perceptions toward flipped learning: New methods to increase interaction and active learning in economics. International Review of Economics Education 17: 74–84. doi:10.1016/j.iree.2014.08.003.

75.

Robins

Rountree

(2003) Learning and teaching programming: A review and discussion. Computer Science Education 13(2): 137–172. doi:10.1076/csed.13.2.137.14200.

76.

Roper

A. R.

(2007) How students develop online learning skills. Educause Quarterly 1: 62–64.

77.

Ryan

R. M.

Deci

E. L.

(2000) Intrinsic and extrinsic motivations: Classic definitions and new directions. Contemporary Educational Psychology 25(1): 54–67. doi:10.1006/ceps.1999.1020.

78.

Saadé

R. G.

Kira

(2007) Exploring dimensions to online learning. Computers in Human Behavior 23(4): 1721–1739. doi:10.1016/j.chb.2005.10.002.

79.

Sáez-López

J. M.

Román-González

Vázquez-Cano

(2016) Visual programming languages integrated across the curriculum in elementary school: A two year case study using “Scratch” in five schools. Computers & Education 97: 129–141. doi:10.1016/j.compedu.2016.03.003.

80.

Sahin

N. H.

Heppner

P. P.

(1993) Psychometric properties of the problem solving inventory in a group of Turkish university students. Cognitive Therapy and Research 17(4): 379–396. doi:10.1007/BF01177661.

81.

Savery

J. R.

Duffy

T. M.

(1995) Problem based learning: An instructional model and its constructivist framework. Educational Technology 35(5): 31–38. Retrieved from https://www-jstor-org-443.web.bisu.edu.cn/stable/44428296.

82.

Schermelleh-Engel, K., Moosbrugger, H., & Müller, H. (2003). Evaluating the fit of structural equation models: Tests of significance and descriptive goodness-of-fit measures. Methods of psychological research online, 8(2), 23–74.

83.

Solmaz, E. (2014). In programming language instruction the effect of Alice to course achievement, the skills of critical thinking and problem solving and metacognitive awareness level (Unpublished doctorate dissertation). Gazi University, Ankara.

84.

Spiro

R. J.

Collins

B. P.

Thota

J. J.

Feltovich

P. J.

(2003) Cognitive flexibility theory: Hypermedia for complex learning, adaptive knowledge application, and experience acceleration. Educational Technology 43(5): 5–10.

85.

Spiro, R. J., Coulson, R. L., Feltovich, P. J., & Anderson, D. K. (1988). Cognitive flexibility theory: Advanced knowledge acquisition in ill-structured domains (Tech Report No. 441). Champaign, IL: Center for the Study of Reading, University of Illinois.

86.

Stelzer

Gladding

Mestre

J. P.

Brookes

D. T.

(2009) Comparing the efficacy of multimedia modules with traditional textbooks for learning introductory physics content. American Journal of Physics 77(2): 184–190. doi:10.1119/1.3028204.

87.

Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston, MA: Allyn & Bacon.

88.

Talbert, R. (2015). Three critical conversations started and sustained by flipped learning. Retrieved from www.facultyfocus.com/articles/blended-flipped-learning/three-criticalconversations-stated-sustained-flipped-learning.

89.

Thorndike

E. L.

(1932) The fundamentals of learning, New York, NY: Bureau of Publications.

90.

Timarová

Salaets

(2011) Learning styles, motivation and cognitive flexibility in interpreter training: Self-selection and aptitude. Interpreting 13(1): 31–52. doi:10.1075/intp.13.1.03tim.

91.

Topalli

Cagiltay

N. E.

(2018) Improving programming skills in engineering education through problem-based game projects with Scratch. Computers & Education 120: 64–74. doi:10.1016/j.compedu.2018.01.011.

92.

Tugun

Uzunboylu

Ozdamli

(2017) Coding education in a flipped classroom. TEM Journal 6(3): 599–606.

93.

Vaca-Cárdenas, L. A., Bertacchini, F., Tavernise, A., Gabriele, L., Valenti, A., Olmedo, D.E, … E. Bilotta (2015, September). Coding with Scratch: The design of an educational setting for Elementary pre-service teachers. Paper presented at the Interactive Collaborative Learning (ICL), International Conference on, IEEE, Florence, Italy.

94.

Wiginton, B. L. (2013). Flipped instruction: An investigation into the effect of learning environment on student self-efficacy, learning style, and academic achievement in an algebra I classroom (Unpublished doctoral dissertation). The University of Alabama, Tuscaloosa.

95.

Wolber, D. (2011). App inventor and real-world motivation. 42nd ACM Technical Symposium on Computer Science Education (SIGCSE ‘11). New York, USA.

96.

Yildiz-Durak

(2017) Turkish adaptation of the flipped learning readiness scale for middle school students. Bartin University Journal of Faculty of Education 6(3): 1056–1068. doi:10.14686/buefad.328826.

97.

Yildiz-Durak

(2018a) Digital story design activities used for teaching programming effect on learning of programming concepts, programming self-efficacy, and participation and analysis of student experiences. Journal of Computer Assisted Learning 34(6): 740–752. doi:10.1111/jcal.12281.

98.

Yildiz-Durak

(2018b) Flipped learning readiness in teaching programming in middle schools: Modelling its relation to various variables. Journal of Computer Assisted Learning 34(6): 939–959. doi:10.1111/jcal.12302.

99.

Yildiz-Durak, H. (2018c). The effects of using different tools in programming teaching of secondary school students on engagement, computational thinking and reflective thinking skills for problem solving. Technology, Knowledge and Learning. doi:10.1007/s10758-018-9391-y.

100.

Yildiz-Durak, H., and Güyer, T. (2018). Design and development of an instructional program for teaching programming processes to gifted students using scratch. In J. Cannaday (Ed.) Curriculum development for gifted education programs (pp. 61–99). Hershey: IGI Global.

101.

Yildiz-Durak, H., & Şahin, Z. (2018, May 2–4). Examining the opinions of computer science students on using app inventor in programming instruction. Paper presented at the ICITS 2018-International Computer & Instructional Technologies Symposium, Izmir.

102.

Zimmerman

B. J.

(2000) Self-efficacy: An essential motive to learn. Contemporary Educational Psychology 25(1): 82–91. doi:10.1006/ceps.1999.1016.

103.

Zohar, A., & Dori, Y. J. (2012). Introduction. In A. Zohar & Y.J. Dori (Eds.) Metacognition in science education (pp. 1–19). Dordrecht, Netherlands: Springer.