Dimensions of using digital devices

A recent report published on the level of Hungarian K12 schools' digital equipment and level of connectedness found that Hungarian schools on all three surveyed levels (primary, lower- and upper secondary) only meet half of the EU's mean scores, whereas device ownership of learners roughly follows the EU's total average (EU, 2019). It is also reported that the most frequently used digital devices for learning purposes are smartphones, which are used by a quarter of lower-secondary and half of upper-secondary learners. Although the report is mainly concerned with ownership, it is progressive in a way by reporting learners' confidence in their digital competences following up on a self-assessment tool published by the EU (2015), which also meets the EU's general average.

The EU's (2015) Digital literacy self-assessment grid was designed to cover a number of factors that pay a pivotal role in meaningful digital device use. The five factors are 1) safety, 2) communication and collaboration, 3) information and data processing, 4) problem solving, and 5) digital content creation. Users can categorize their competence levels by following certain descriptors to classify if they are basic, independent or advanced technology users of digital devices. As it has been concluded by Tongori (2012), an inclusive approach towards integrating digital devices into teaching and learning is generally favored as the ubiquitous presence of digital technologies prerequisites individuals to be familiar with digital customs and ethics.

Evidently, teachers are not the only sources of students to learn about new technologies, but it has been argued that they have the means to teach them how to use certain devices and technologies for learning purposes specifically (Pintér, 2019; Tóth-Mózer, 2017). Arguably, the best way of developing teachers' techno-pedagogical skills would be the inclusion of digital methodological training courses in teacher education (Dringó-Horváth & Gonda, 2018), but training practicing teachers should not be neglected either (Öveges & Csizér, 2018). Becker (1999) reported that the teachers most likely to include digital technologies in their everyday teaching practice are “1) younger teachers, 2) teachers who are leaders in their profession, and 3) teachers with constructivist pedagogies” (p. 30). Becker (1999) also stated that, despite the classic and convenient juxtaposition of digital natives and digital immigrants, the age factor might lose its importance over time as a predicting factor of digital device use in teaching. According to Bayne and Ross (2011), age has become a less important factor in competent Internet and digital device use, as using technologies is rather a choice than something linked to age; thus it is suggested that individuals using technologies should be called digital citizens (Bayne & Ross, 2011; Papp-Danka, 2013).

Teachers and technology

The Technological Pedagogical Content Knowledge (TPACK) framework was designed by Mishra and Koehler (2006). The authors proposed seven major knowledge components that together ensure meaningful and professionally grounded technology inclusive instruction in education. These components (Koehler et al., 2014) and their interpretations are as follows (Chai, Koh, Tsai, & Tan, 2011):

1. content knowledge (CK): knowing the contents of the specific subject(s) one teaches such as mathematics, foreign languages, or history

2. pedagogical knowledge (PK): knowing about how to plan a lesson, manage learners and handle pedagogical challenges such as heterogenous learning groups

3. technological knowledge (TK): knowing how to use a computer or other digital devices and the programs or applications that run on them

4. technological content knowledge (TCK): knowing how to find or create content to illustrate what is being taught such as creating a presentation or finding a demonstration video

5. pedagogical content knowledge (PCK): knowing the teaching methods required to teach one's specific subject(s) so that they best facilitate learning, such as how to teach mathematical formulas or vocabulary learning methods

6. technological pedagogical knowledge (TPK): knowing how to facilitate learning or implement lessons using technology such as using learning management software

7. technological pedagogical content knowledge (TPACK): knowing how to implement lessons and facilitate learning by selecting the most appropriate technology for the very purposes of teaching one's specific subject

In a comprehensive literature review summarizing the most important findings of years of TPACK-research, Koehler et al. (2012) concluded that the activation of all TPACK framework components together result in meaningful technology inclusion; that is, it is not enough to have a positive attitude towards technology in general and/or being a good teacher of one's subject. Similar to a general positive attitude towards technology, classic offline methodological knowledge does not translate automatically in a digital environment either.

Voogt et al. (2013) argued that teachers' technology integration is based on a number of factors, such as their knowledge of pedagogy, their teaching experience, their beliefs and customs. This suggests that singling out specific elements of TPACK would only result in false claims about what reasoning is behind certain teachers' willingness towards technology inclusion or non-inclusion (Voogt et al., 2013). Reflecting on the complexity of reasoning behind technology use, better and more reliable instruments should be designed (Voogt et al., 2013) to understand about teachers' technology use in its complexity rather than by its single elements stripped from various other factors that influence the choice of technology use. Measurement instruments should be more specific in terms of subjects taught by teachers (Schmidt et al., 2009), local contexts where the integration is taking place (Főző & Racsko, 2020) and the dynamic, rapidly changing nature of what each TPACK element conceptually stands for (Főző & Racsko, 2020; Voogt et al., 2013).

The fact that beliefs overrule training or a seemingly positive attitude towards technology inclusion was observed by many researchers (Chen, 2008; Hennessy et al., 2007; Webb & Cox, 2004). Respondents in these studies were asked to elaborate on their professional reasoning for ICT inclusion in their classrooms. Most of them agreed with the principles of meaningful technology use and were in alignment with TPACK theory (Koehler et al., 2014), stating that ICT inclusion is the extension of a teacher's array of methodology. However, based on the post-interview lesson observations, the same teachers were reported to resort to their beliefs and use more traditional methods in the classroom than they had claimed (Heitink et al., 2016). This is important for two reasons; on the one hand, it confirms that teachers' beliefs are deep-rooted and the lack of experience with theoretically grounded use of technology results in them resorting to more traditional teaching methods. On the other hand, these findings also confirm that there is a positive attitude towards the use of technology in the classroom and teachers' beliefs might be altered through positive experiences and technology-focused training opportunities requiring longitudinal engagement.

A major takeaway of years of TPACK research is that teachers should receive subject-specific techno-pedagogical training. The fact that only 27% of Hungarian upper-secondary schoolteachers received such training as part of their teacher education compared to the 48% EU average (EU, 2019) signals a slow digital transformation of Hungarian teacher education despite repeated calls for action (Dringó-Horváth & Gonda, 2018; EU, 2018; MDOS, 2016; Öveges & Csizér, 2018).

The current study

The aim of this study was to identify and profile groups among Hungarian K12 teachers based on their techno-pedagogical skills to find out more about the nature of teachers' technology use as well as to find viable development areas respective of each group. Because emergency remote education affected all Hungarian K12 schoolteachers, the present study aims to reflect on online education as well. In order to fulfil this aim, a questionnaire was developed, which 1) adapted Schmidt et al.’s (2009) TPACK questionnaire, 2) proposed ten additional constructs on various aspects of emergency online teaching and 3) collected several background variables because to profile teachers belonging to the clusters found to discover similarities and differences between the groups in order to offer some possible ways of how teachers could develop depending on their corresponding clusters. Therefore, this study sought answers for the following research questions:

1. How do Hungarian K12 teachers rate their Technological Pedagogical Content Knowledge (TPACK) components?

2. How many clusters can be observed among Hungarian K12 teachers based on their techno-pedagogical skills level?

3. Are there any statistically significant differences based on age, gender, teaching experience, qualification, place of work or career model among Hungarian K12 teachers in relation to their techno-pedagogical skills groups?

4. What are some development possibilities of Hungarian K12 teachers' techno-pedagogical skills?

5. Are there any statistically significant differences in how Hungarian K12 teachers perceived emergency remote teaching depending on which techno-pedagogical skills group they belong to?

Designing and validating the data collection instrument

Developing the online questionnaire that became the measurement instrument used in this study began in April 2020 by adapting an existing TPACK questionnaire (Schmidt et al., 2009) and conducting three informal interviews with K12 teachers about their experiences and challenges of the COVID-19 triggered emergency remote teaching. The original TPACK questionnaire contains 37 items, and informants are asked to rate them on a 5-point Likert scale depending on to what extent they feel that the statements are true for them. The items are organized into seven constructs that are the elements of the TPACK framework. As requested by the authors of the original measurement instrument, consent was asked and received from the corresponding author to validate the Hungarian translation and adaptation of the instrument.

The original questionnaire (Schmidt et al., 2009) was translated into Hungarian with some minor modifications. First, the items of the original questionnaire construct, content knowledge (CK) referred to specific content areas such as mathematics or social studies, but for the purposes of this study, they were translated in a way that the items did not include the names of any specific subjects, rather the wording ‘the subject(s) I teach’ was used. For example, the original item I have sufficient knowledge about mathematics was translated as I have sufficient knowledge about the subject(s) I teach. The other modification was that the item pool was mixed up, because in the original questionnaire, items appear one after the other with construct labels, whereas a mixed-up item pool is preferred “to create a sense of variety and to prevent respondents from simply repeating previous answers” (Dörnyei & Csizér, 2012, p. 78).

For the additional constructs, informal interviews were conducted with three volunteering teachers to find common themes about the challenges and difficulties surrounding emergency remote teaching that could later be related to the TPACK scales in March 2020. These themes were turned into questionnaire constructs 8 to 17, summarized in Table 1 (see the Appendix for item lists). All volunteering teachers were women, one of them worked in a secondary grammar school of the capital city, one in an elementary school of the county surrounding the capital city and one in an elementary school in Western Hungary. The development and validation process of the additional questionnaire constructs followed the protocol of Dörnyei and Csizér (2012) as well as some recent Hungarian questionnaire validation studies (Dringó-Horváth & Gonda, 2018; Fekete, 2020a).

Before large-scale data collection was launched, the questionnaire was pre-piloted during two think-aloud sessions with two Hungarian K12 teachers, then the online questionnaire was piloted with 50 participants in the first days of April 2020. To be accepted as valid, the questionnaire constructs had to meet two different criteria; 1) for each construct, Cronbach's alpha had to reach a minimum 0.60 cut-off value, and 2) items of the constructs had to load to the same dimension using principal component analysis (Dörnyei & Csizér, 2012). Table 1 summarizes the reliability analyses of the piloted questionnaire, the emerging problems with certain constructs, the proposed solutions to these problems and the reliability analyses of the final questionnaire constructs.

Participants

A total number of 216 Hungarian K12 teachers participated in this study. The sample consisted of 184 females (85%) and 31 males (15%) and one participant preferred not to disclose their gender. This gender ratio keys in with the 90–10% female-male ratio in Illés and Csizér’s (2018, p. 162) representative sample, which involved 1118 foreign language teachers from Hungarian primary and secondary schools. The age of the participants ranged from 24 to 65 years with a mean of 49.20 years (SD = 8.96). The teaching experience of the informants ranged from 1 to 43 years with a mean of 22.97 years (SD = 10.90).

According to the progressive Hungarian career model of teachers (Eurydice, 2020), 2.3% (N = 5) of the informants were in their compulsory traineeship period, 37% (N = 80) were in Teacher 1 status, 42.1% (N = 91) in Teacher 2, 17.6% (N = 38) were Master Teachers and 0.9% (N = 2) were Researcher Teachers. Teachers must submit a written portfolio and take an evaluating exam to be reclassified into their successive categories, but the last two categories are optional. The categories represent the teachers' qualifications as well as teaching experience.

Geographically, 43% (N = 93) of the informants worked in institutions of the capital city or the county surrounding the capital city, 30% (N = 65) worked in Western and 27% (N = 58) in Eastern Hungary. Hungary's most developed region is its central one, that is, the capital city and the surrounding county, followed by the Western and the Eastern counties (Kiss, 2018). Although this three-way categorization overlooks several regional differences, the tendencies might be relevant from the perspective of this study since the general level of regional development might affect teachers' level of technological knowledge by having access to different means of technology.

To gain a general understanding about the technological knowledge level of the participants, they were asked to rate their perceived level of technological knowledge before and after the curfew measures on a 5-point scale. The mean pre-lockdown average was 3.69 (SD = 0.85), while the post-lockdown mean score was 4.20 (SD = 0.65). A paired sample t-test confirmed that there was a statistically significant difference among how much informants knew about technology before and after the lockdown (t = 12.43; Sig. (2-tailed) < 0.001).

Methods of data collection and analysis

Data was collected through an online Google Forms questionnaire. Because of the online nature of data collection, informants remained completely anonymous even for the researcher. Data collection started in April and ended in May 2020. As a result of nationwide school closures in March 2020, the joining rate to digital and general K12 teaching methodology Facebook groups skyrocketed; therefore, the questionnaire was distributed on these platforms with a short message to teachers. The downloaded Excel data pool was coded and then loaded into Statistical Package for the Social Sciences (SPSS) version 22. SPSS was also used for running reliability analyses as well as descriptive and inferential statistical tests.

How do Hungarian K12 teachers rate their technological pedagogical content knowledge (TPACK) components?

To observe how Hungarian K12 teachers rated the questionnaire constructs, Table 2 was created. It was important not only to include the mean values and standard deviations but the minimum and maximum values informants rated each construct, because these values give hints about the individual differences in the sample. It was hypothesized that there are going to be observable differences prompted by standard deviation and the minimum and maximum values.

In average, Hungarian K12 teachers rated their TPACK skills (constructs 1–7) rather high, their content knowledge (M = 4.65; SD = 0.448) being the highest, and technological knowledge (M = 3.98; SD = 0.775) the lowest observed average, yet technological knowledge's higher observed standard deviation, which is true to every other TPACK construct except for content knowledge, suggests many individual differences. The same tendency can be observed among the rest of the constructs. The lowest non-TPACK average is observed for Perceived quality of students' online learning (M = 2.61; SD = 0.742), the highest standard deviation value is observed for Teachers' fatigue (M = 2.72; SD = 1.050), while the highest non-TPACK mean is for Teachers' motivation (M = 4.37; SD = 0.596). These values, although informative, do not tell much about systematic individual differences; therefore, it would be advisable to take these into consideration to allow for drawing more reliable conclusions.

How many clusters can be observed among Hungarian K12 teachers based on their techno-pedagogical skills level?

It was hypothesized that there should be three or four sub-groups among the participants on the basis of how they rated their technological (TK), technological content (TCK), technological pedagogical (TPK) and technological pedagogical content knowledge (TPACK) skills. The hypothesis was supported by observing the minimum and maximum mean values and the frequency of the values participants selected regarding the mentioned constructs (Table 2). The hypothesis was further confirmed by previous research reporting on teachers' very different individual techno-pedagogical skills levels and beliefs on technology inclusion (Chen, 2008; Heitink et al., 2016; Hennessy et al., 2007; MDOS, 2016; Webb & Cox, 2004; Voogt et al., 2013).

To explore if the supposed groups were present in the sample, cluster analysis was used following the protocol of Csizér and Jamieson (2013). First, the cluster forming variables were proposed, these were constructs 1, 5, 6, and 7, that is, those TPACK scales that included technology. A 20% subsample (N = 44) was randomly selected in SPSS for hierarchical clustering. The dendrogram revealed three distinctive clusters that still needed to be confirmed. To check the reliability of the three groups in the sample, now each participant (N = 216) was assigned to their corresponding cluster by SPSS. A further k-means cluster analysis was run on the entire sample that also confirmed the presence of the proposed three groups. As a final reliability check, ANOVA analyses were run on the cluster forming constructs to check whether informants belonging to the three clusters rated the cluster forming constructs differently. The ANOVA analyses confirmed statistically significant differences between the three clusters in the four cluster forming constructs, summarized in Table 3.

Thus, based on the technology-related variables, three groups were established among Hungarian K12 teachers labeled as teachers reported displaying 1) beginner, 2) independent, and 3) advanced level techno-pedagogical knowledge. The proportion of K12 teachers in the groups could also be determined, 20% (N = 43) of the teachers perceived they were beginners, 41% (N = 89) independent, and 39% (N = 84) advanced users of techno-pedagogical skills.

Are there any statistically significant differences based on age, gender, teaching experience, qualification, place of work or career model among Hungarian K12 teachers in relation to their techno-pedagogical skills groups?

To illustrate the differences between the established groups based on age, gender, teaching experience, qualification, place of work or career model, Table 4 was created, which contains the collected descriptive statistics of the selected background variables. As far as the group sizes are concerned, one fifth of the participants belong to Beginners, two fifths to Independent, and two fifths Advanced users – size being the biggest difference observed among all the variables.

To confirm whether the groups portray any statistically significant differences, Chi-square tests were run on the nominal background variables. The only statistically significant difference between the groups (Chi-square = 10.63; df = 4; Sig. = 0.31) was found in the places of work. Significantly less people belong to Cluster 1 (Beginners) from Eastern Hungary than from the other two regions (Adjusted residual = −2.1), and significantly less people belong to Cluster 2 (Independent users) from Central Hungary (Adjusted residual = −2.6), but this might entirely be due to the relatively small sub-sample size. The groups showed no further significant differences, thus teachers in the three groups do not systematically differ in their age, gender, teaching experience, qualification and career model are concerned.

The Chi-square test confirmed that the established clusters based on Hungarian K12 teachers' techno-pedagogical skills are rather homogenous. Most background variables do not seem to be predictors of techno-pedagogical knowledge, and how teachers perceive their techno-pedagogical skills level rather depends on whether they opt for being digital citizens (Bayne & Ross, 2011), further influenced by beliefs about technology use (Voogt et al., 2013) and teacher education or training programs (Dringó-Horváth & Gonda, 2018).

Although the TPACK scales provide more complex grounds for analysis, informants were also asked to rate their perceived level of digital skills before and after emergency remote teaching on a 5-point scale. On the one hand, the TPACK questionnaire items were mixed up to maximize the validity of the measurement instrument; on the other hand, they measured teachers' perceptions without being able to reflect on the hypothesized boost in their skills growth. The reason for asking informants to rate their digital skills levels on two scales one after the other was to check whether they perceived any changes in their digital skills knowledge and to observe the scale of skills growth. As summarized in Table 5, Hungarian K12 teachers belonging to the established techno-pedagogical clusters reported statistically significant digital skills growth confirmed by paired sample t-tests.

In all three clusters, the averages grew and the individual differences (as confirmed by the standard deviation values) lowered. A noteworthy positive tendency is that teachers in the Beginners cluster reported similar digital skills knowledge after distance education that of the Independent users reported before emergency remote teaching. On average, teachers in all three groups perceived that they gained digital skills throughout distance education, and the sudden complete shift required most effort to be invested from teachers belonging to the Beginners group in catching up with online possibilities. Yet, even though they caught up with the pre-remote teaching average of the Independent users, it does not mean that Beginner teachers have become very confident and competent in their digital skills. It can also be said, confirmed by the statistically significant skills gains of all three groups, that a major positive attribute of the exigency of teaching via technology elevated the levels of techno-pedagogical knowledge, creating new standards.

What are some development possibilities of Hungarian K12 teachers' techno-pedagogical skills?

In order to propose possible ways of development respective of the clusters, path analyses are proposed (Székelyi & Barna, 2002). The subsample sizes allowed for running such analyses (Dörnyei & Csizér, 2012, p. 82) as the smallest subgroup was close to the suggested minimum 50 subsample size.

Because the final goal is TPACK development (Koehler et al., 2014; Mishra & Koehler, 2006; Schmidt et al., 2009) – being able to synthesize technological, content, pedagogical, technological content, pedagogical content and technological pedagogical knowledge – the path ends were TPACK in all clusters. Then, backward, the strongest predictors were listed until there were no more statistically significant predictors found. This resulted in three proposed pathway models. Although, in the end, regression analyses were used to create the models, Pearson correlations were also run on the clusters to confirm the paths, which is a widely used and recommended additional reliability measure (Székelyi & Barna, 2002).

Figure 1 illustrates the path model for TPACK skills growth in Cluster 1. The strongest predictor of TPACK development is Technological Pedagogical Knowledge (TPK), the knowledge of how teachers can facilitate and support the learning processes of their students by using technological tools, such as e-mailing or learning management systems (e.g., Edmodo, Google Classroom, Microsoft Teams). For teachers reporting Beginner levels of techno-pedagogical skills, it mainly takes learning more about such specific technologies that enable keeping in touch with students asynchronously to show TPACK development, while technological content knowledge and pedagogical content knowledge components are not part of the model at all. Furthermore, for Beginners to develop TPK knowledge, the strongest predictor is technological knowledge (TK), that is, learning about how to operate computers and computer programs, such as e-mailing or learning management systems. This model suggests that Beginners need help on the very operational levels of technology, and technological knowledge is focused on educational technologies only.

View Full Size

Path analysis of TPACK in cluster 1

Citation: Journal of Adult Learning, Knowledge and Innovation 5, 2; 10.1556/2059.2022.00056

• Download Figure
• Download figure as PowerPoint slide

Figure 2 proposes a path model for Independent users that would ultimately result in their TPACK development. The biggest difference between the models of Beginners and Independent users is the absence of TK as a significant predicting factor in the latter model, suggesting that Cluster 2 teachers are confident users of computers and computer programs, and they do not have basic computer operational hindrances. The strongest predicting factor of TPACK is TPK, similar to Cluster 1, but with a lower Beta value. The most forceful predicting factor of TPK is TCK, that is, beyond finding ways of keeping in touch with learners (TPK), Cluster 2 teachers try to find or create relevant digital content (e.g., finding video demonstrations, creating presentations) to facilitate the knowledge construction of learners. This created or researched content (TCK) predicts how Cluster 2 teachers select the means and methods of delivering it to their learners (TPK).

View Full Size

Path analysis of TPACK in cluster 2

Citation: Journal of Adult Learning, Knowledge and Innovation 5, 2; 10.1556/2059.2022.00056

• Download Figure
• Download figure as PowerPoint slide

A possible way of development for Cluster 2 teachers could be to experiment more with online classes and teaching approaches that deviate more from traditional, offline classroom practices. That is, as opposed to implementing lessons online that resemble traditional classroom approaches such as frontal teaching in front of the camera or checking homework by asking students to provide the correct answers one by one, they could experiment with more progressive possibilities such as flipped classroom techniques and project work. Naturally, the subject they teach comes heavily into play in this respect, and Cluster 2 teachers should always rely on their generally good level of subject-specific knowledge and the criticality towards new approaches that originate from their subject-specific knowledge in determining which progressive practices to experiment with non-arbitrarily.

Figure 3 shows the path model for Advanced users. The biggest difference of this model is that TPACK is predicted by three different knowledge domains. Similar to Independent users, TPK is a predictor of TPACK, but TCK emerges as another predicting factor, and not as a predicting factor or TPK. This suggests that Cluster 3 teachers have an array of techno-pedagogical knowledge and knowledge about educational technology at their disposal compared to what is seen on Cluster 2 teachers' model, where the created or found digital content predicted the means of facilitating students' learning. It is interesting to see that in the case of Advanced users, pedagogical content knowledge (PCK), i.e., how teachers are able to teach their specific subject(s) with the most appropriate methods, is also part of TPACK predicting factors, although it is a non-technological scale.

View Full Size

Path analysis of TPACK in cluster 3

Citation: Journal of Adult Learning, Knowledge and Innovation 5, 2; 10.1556/2059.2022.00056

• Download Figure
• Download figure as PowerPoint slide

According to the path model, pedagogical content knowledge development also results in TPACK development, prompting that for Cluster 3 teachers, being able to teach their subject through appropriate digital technologies is considered to be part of their general, subject-specific teaching methodology. Cluster 3 teachers' further development could target trying out a vast array of digital teaching possibilities because their PCK, TCK and TPK knowledge as TPACK development predictors ensure sufficient pedagogical reasoning and learner-centeredness for the implementation of various progressive techno-pedagogical instructional methods.

Are there any statistically significant differences in how Hungarian K12 teachers perceived emergency remote teaching depending on which techno-pedagogical skills group they belong to?

Questionnaire constructs 8 to 17 were formed based on interviews with three Hungarian K12 teachers to identify common themes in what their perceptions were with digital education. The reason for including these scales was to relate them to the established clusters to learn more about the dimensions of emergency remote teaching in relation to techno-pedagogical skills. For this, ANOVAs and post-hoc Duncan analyses were run on questionnaire constructs 8 to 17 summarized in Table 6.

Based on the findings, the constructs can be grouped into four categories. The first category can be established based on the constructs showing no statistically significant differences among clusters (1 = 2 = 3), including constructs Teachers' fatigue, Digital skills improvement, Disadvantages of online teaching. This suggests that Hungarian K12 teachers experienced emergency remote teaching similarly irrespective of their techno-pedagogical groups in these respects. Although in the construct-forming interviews all three teachers expressed that they experience fatigue, Hungarian K12 teachers were only moderately tired specifically because of the online environment (M = 2.72; SD = 1.05), but the relatively high standard deviation value suggests many individual differences.

The second category consists of the constructs where teachers with Advanced techno-pedagogical skills rated the constructs significantly differently than Beginners and Independent users (1, 2 < 3). This category includes Giving online feedback and Online learning management. Giving online feedback is the only construct that was rated higher by Beginners than Independent users, but the exact methods and practices teachers used to give online feedback were not included in the study. All in all, the more advanced techno-pedagogical skills teachers reported, the more confident they were in the possibilities of providing online feedback on student work as well the more they feel confident in online learning management such as contacting learners quickly, letting them know clearly what tasks they should do and keeping track of online tasks submissions.

The third category consist of constructs Teachers' motivation in which constructs there is a statistically significant difference between Beginners and Advanced users, but Independent users can be included in both groups (1, 2 < 2, 3). In all cases, the mean values of clusters are always higher in relation to the level of techno-pedagogical skills. Teachers' motivation was rather high in all three clusters, although the motivation of learners was much lower than the teachers', and the perceived quality of students' online learning was also rather low. It is by no means surprising that the more advanced teachers' are with technology, the more motivated they are, and the more motivated they are, the more motivated they perceive their learners are with digital learning, yet the levels of teachers' motivation and the perceived levels of learner motivation and learning quality do not approximate one another.

Finally, the fourth and last category includes construct Offline vs. online teacherly self, where there is a statistically significant difference between all clusters (1 < 2 < 3). In this construct, teachers were asked to rate statements that compared offline teaching to online teaching. It was found that the more advanced techno-pedagogical clusters teachers belong to, the more they favor offline, in-class education as they feel teaching in the online space to be less effective, despite their high techno-pedagogical skills. It is important to note that although teachers in all clusters seem to prefer face-to-face instruction, this does not mean that they prefer not developing or using their techno-pedagogical skills. Although teachers in all clusters prefer face-to-face instruction to online instruction, this finding might be linked to the high levels of TPACK knowledge also confirmed by the TPACK pathway analysis (Fig. 3) of Cluster 3 teachers, because using digital technologies seems to be part of their general teaching methodology, thus being an expansion of instruction possibilities rather than a substitution. In this respect, the fact that the higher techno-pedagogical skills self-reported by teachers, the more they are aware of the fact that technology should be integrated into everyday teaching practices as opposed to substitute traditional education.

Summary of main findings

This non-representative study found that Hungarian K12 teachers can be classified into three distinctive groups as far as their techno-pedagogical skills are concerned on the basis of the technological scales (TK, TCK, TPK, TPACK) of Schmidt et al.'s (2009) TPACK questionnaire. In the study (N = 216), approximately 20% of the informants were Beginners, 41% were Independent, and 39% were Advanced users of techno-pedagogical tools and methods. Chi-square crosstab tests confirmed that the only statistically significant difference between the groups concerned where teachers worked, as there were less teachers among the Beginners from Eastern Hungary, and there were less Independent users from Central Hungary; however, this might be due to a small sub-sample size. Other than this, no statistically significant difference was found between Hungarian K12 teachers in the different groups based on their age, gender, teaching experience, qualification, and career model.

Teachers in all groups reported gains in their techno-pedagogical skills, and this gain proved to be statistically significant. It seems that Covid-19-triggered emergency remote teaching set up new techno-pedagogical standards for teachers, but there is still much room for improvement in teacher education and training programs (Dringó-Horváth & Gonda, 2018; Öveges & Csizér, 2018), because Beginners, for example, only now learnt how to use devices for keeping in touch with their students. Practicing teachers should be offered relevant, preferably subject-specific technological-methodological training courses, and such courses should be offered in teacher education programs, because as it was seen, techno-pedagogical levels were irrespective of age, qualification, and teaching experience. Furthermore, in the present study, it was found that the more advanced group teachers belonged to, the higher they rated their non-technological TPACK knowledge components: content knowledge (CK) and pedagogical knowledge (PK).

With TPACK growth as the ultimate goal (Koehler et al., 2014; Mishra & Koehler, 2006; Schmidt et al., 2009), path analyses were conducted to find development possibilities of the techno-pedagogical knowledge of Hungarian K12 teachers specific to their current techno-pedagogical levels. Beginners need to develop on the very basic levels of using technology to be able to supervise and correspond with their learners, and are rather far from complex, pedagogically reasoned implementations of online classes. Independent users seem to be the ones who try to put their traditional pedagogical practices into the online space by creating or finding relevant digital content to facilitate learning processes that are not necessarily implemented in the forms of synchronous online classes. Teachers in this group need to develop in the facilitation and implementation of online classes.

Advanced users are able to synthesize their subject pedagogical knowledge and teaching online classes. In this respect, for them, TPACK development could be trying out even more online pedagogical methods and practices with a sufficient level of criticality while always keeping their learners' needs in focus to test the effectiveness of certain digital teaching possibilities. Independent and Advanced users, furthermore, could also help their colleagues in their techno-pedagogical skills development, because development is most effective if it is related to the specific teaching-learning context, including educational level and specific school subjects. It is also important to add that, with the exception of Beginners, learning more about technology and digital devices would not result in techno-pedagogical gains (Dringó-Horváth & Gonda, 2018; Pintér, 2019).

Finally, as far as the remote teaching experiences and perceptions of Hungarian K12 teachers in the different groups are concerned, the most notable difference was that Advanced techno-pedagogical teachers rated items of the construct, Offline vs. online teacherly self the highest, meaning that they are the ones who most want to return to the classrooms. This phenomenon can be explained by what was observed on their TPACK path models, that is, Advanced users use techno-pedagogical methods and tools as part of their everyday, in-class teaching practices. In all three groups, teachers' motivation for online teaching was much higher than the perceived motivation of learners, and teachers reported that they observed growing fatigue of their learners. The lowest mean values were attached to the quality of students' learning throughout emergency remote education.

Limitations

Limitations of this study concern sampling, sample size and the online nature of data collection. These limitations are mainly due to Covid-19, as during spring curfew measures, there was no other way of collecting data than distributing an online questionnaire in various teaching methodology groups on Facebook. Although the number of members joining such Facebook groups grew rapidly because of Covid-19, it is possible that rather those teachers filled in the questionnaire who had more positive beliefs about technology and techno-pedagogy, or the ones who had had more experiences with it. However, the disposition of teachers (gender, age, experience, etc.) in each established techno-pedagogical group suggests that the sample, although non-representative and not generalizable, reflects Hungarian K12 teachers well.

Additionally, confirmatory factor analysis was not carried out on the scales adapted (translated into Hungarian) from the original TPACK questionnaire (Schmidt et al., 2009), nevertheless data analysis was preceded by a pilot phase, and construct reliability was confirmed with principal component analyses and Cronbach's alpha values. It is also important to mention that cluster analysis, lacking a statistical model, is exploratory rather than confirmatory (Csizér & Jamieson, 2013), thus the results represent and reflect the current sample only and are by no means generalizable. Nonetheless it is hoped that the findings are transferable to similar contexts.

Future directions

It would be worthwhile replicating the study using the same data collection instrument with more participants. A larger sample size could give grounds for comparing the findings of the two studies and would allow for more precise categorization of teachers based on their techno-pedagogical levels. A larger sample size would also allow for more detailed data analysis, for example by comparing specific school levels and regions of Hungary.

In the future, as a possible follow-up and expansion of this hypothesis-testing approach, interviews with selected teachers from all three clusters could be conducted to find out more about their specific ways of conducting online classes, their positive and negative experiences, the contents of their lessons, the ways they perceive their learners were engaged and the possibilities of testing their learners through digital devices.

It would also be important to offer subject-specific digital teaching methodology courses as part of teacher training and teacher education and following them up would be a viable continuation of this study. Based on the present findings, these courses could be personalized to a certain extent, that is, bearing in mind the already existing level of techno-pedagogical knowledge of teachers.

Finally, to help teachers more closely and reliably determining their current techno-pedagogical levels and to give them specific ideas and ways of how to develop their knowledge, the TPACK scales of the present data collection instrument could be used as a self-check survey that would determine which group the participants belong to. The teachers filling in the survey would also be data providers for observing the continuously changing normal and cut-off values that could help determine which cluster they belong to. Such an online tool would be profitable for both teachers and researchers.