This section presents the results of the study, beginning with an examination of teachers’ beliefs about DI and TEDI, and then outlining their reported implementation of these approaches. These findings were published in an article (Kótay-Nagy, 2025); sections from that publication are reproduced here verbatim, supplemented with some additional data and analysis.

In order to see if there are any differences in the participants’ answers regarding their self-efficacy beliefs in DI based on their background variables, an independent-samples t test with gender as grouping variable, and ANOVA with school type, years of teaching experience and region were calculated. The tests yielded no significant differences, which may indicate that, irrespective of the type and location of the school they teach at and their years of teaching experience, there is a relatively similar sense of efficacy among teachers regarding DI.

What appears as interesting is that students’ interests (M = 3.95, SD = 0.91) and learning profiles (M = 3.91, SD = 0.83), which are also among the main pillars of DI, received significantly lower mean values than English language proficiency (t(211) = 6.78, p < .001 and t(211) = 8.44, p < .001, respectively) and general language learning skills (t(211) = 4.80, p < .001 and t(211) = 6.51, p < .001, respectively), which suggests that Hungarian EFL teachers may find students’ readiness levels more important to address in DI than students’ interests and learning profiles.

In an open-ended question, participants were invited to add any other learner differences that they find important to consider in DI. Two categories emerged as especially popular: students’ motivation to learn English (15 mentions) and students’ emotional and mental well-being (15 mentions), which, again, are both linked to the concept of readiness, i.e., they describe certain aspects of students’ “current proximity to specified knowledge, understanding, and skills” (Tomlinson & Imbeau, 2023, p. 33). The word “current” was actually used by eight of the respondents along with other words such as “on a given day” (3 mentions) or “on a given week” (1 mention). This implies that many of the teachers construe the concept of readiness, in line with Tomlinson’s (1999) definition, as situation-specific and ever-evolving rather than fixed.

 

Thirdly, as was discussed in Section 4.1.2.1, in general, teachers consider students’ learning profiles less important to be addressed than their readiness levels, however, when it comes to their self-reported practices of learning profile-based DI, these show a mean value (M = 4.06, SD = 0.57) similar to that of readiness-based DI, with no statistically significant difference between the two. Interest-based DI, on the other hand, received a mean value (M = 3.24, SD = 0.76) that was significantly lower than those of readiness-based DI (t(211) = 16.62, p < .001) and of learning profile-based DI (M = 4.06, SD = 0.57) (t(211) = 17.94, p < .001). This pattern suggests that Hungarian EFL teachers not only find readiness to be the most important learner difference to attend to, but they also actually report to differentiate with this learner difference in focus, with learning profile-based DI being practiced to the same extent, and interest-based DI being significantly less present in their practice.

In order to see if there are any differences in the participants’ answers pertaining to their self-reported practices of DI and TEDI, an independent-samples t test with gender as grouping variable, and ANOVA with school type, years of teaching experience and region were calculated. Comparisons based on gender and years of teaching experience did not reveal any significant differences. However, significant variations were found in terms of the type of institution where participants work. Results suggest that those teachers who teach both in primary and secondary education, use readiness-based DI less frequently (M = 3.62, SD = 0.87) than those who teach either exclusively in primary (M = 4.18, SD = 0.57) or in secondary education (M = 4.05, SD = 0.65) (F(2, 209) = 3.23, p = .041). A tentative explanation for this pattern could be that teachers who work at both levels, such as in grammar schools with 6-year or 8-year programmes, may be teaching in contexts where students tend to be grouped more homogeneously by language proficiency, which might, in turn, lessen the perceived need for readiness-based DI. However, as no supporting evidence has been identified in the literature, this remains a hypothesis that requires further investigation, potentially through qualitative studies that examine DI in such contexts.

Some significant differences between regions were also found. It seems that teachers in central Hungary use readiness-based DI slightly less frequently (M = 3.93, SD = 0.65) compared with their colleagues in the Western (M = 4.2, SD = 0.56) and Eastern (M = 4.1 SD = 0.69) parts of Hungary (F(2, 209) = 3.30, p = .039). This, again, may be explained by the greater heterogeneity of English groups in these regions, which could make the use of readiness-based DI more necessary. However, it is worth noting that the difference is relatively small, which suggests that regardless of some minor regional variation, Hungarian EFL teachers generally report to use readiness-based DI to a similar extent.

 

The fact that planning-related challenges were rated as the most concerning issues may be linked to the nature of DI where “the bulk of the work is in the upfront planning” (Blaz, 2016, p. 161), which makes additional planning time an inherent characteristic of the approach. However, the analysis of the responses given to the open-ended question on the challenges of DI show that these concerns are further intensified by certain contextual issues. The most frequently mentioned problems were the presence of large class sizes (7 mentions), often including 20 or more students with varying levels of English proficiency, and the substantial workload (6 mentions), involving 24 or more lessons per week. As the participants pointed out, in such circumstances “differentiation appears great in theory, but remains to be an elusive dream in practice” (participant A), as it is “impossible to systematically plan with DI in mind no matter how much one would like to” (participant B). Teachers who do differentiate in such circumstances find it “extremely difficult” (participant C), experience “lots of failures” (participant D), and “almost end up with burn-out” (participant E) or use DI on an “ad hoc basis, in fire-extinguishing mode” (participant F).

Secondly, while the difficulty of acquainting students with DI was rated as only moderately concerning (M = 3.18, SD = 1.13), it emerged as one of the most prevalent issues in the responses to the open-ended question (8 mentions). As some participants put it, many students are accustomed to traditional frontal teaching methods and lack the collaborative skills necessary for differentiated learning. Such mismatches can result in discipline issues and feelings of exclusion or inferiority among students, which may increase teachers’ reluctance to implement DI.

Furthermore, and quite interestingly, out of the challenges listed for participants to rate, methodological uncertainties received the lowest mean value (M = 3.00, SD = 1.19), being significantly lower than the mean values of both the aforementioned two planning-related challenges (t(211) = 10.32, p < .001 and t(211) = 10.89, p < .001, respectively) and lesson delivery-related challenges, namely multitasking (M = 3.23, SD = 1.24) (t(211) = 2.45, p = .015) and time management (M = 3.18, SD = 1.19) (t(211) = 1.99, p = .048), as well as than acquainting students with DI (t(211) = 2.33, p = .021). This implies that while teachers do encounter some difficulties with methodology, the most important challenges of implementation are related to the increased planning demands and the inherent characteristics of the “workshop-style environment” (Blaz, 2016, p. 160) of differentiated classes.

This interpretation is further reinforced by the fact that the enabler that teachers rated as most important was access to differentiated materials (M = 4.58, SD = 0.84), a factor that could considerably reduce planning time (t(211) = 2.93, p = .004 when compared with ICT tools in the classroom; t(211) = 5.81, p < .001 when compared with exchange of DI-related tips among staff members; t(211) = 5.76, p < .001 when compared with support from school administrators; t(211) = 6.52, p < .001 when compared with exchange of student information among staff members; t(211) = 7.86, p < .001 when compared with in-service training on DI; t(211) = 9.73, p < .001 when compared with teaching staff’s positive attitude towards DI; t(211) = 11.38, p < .001 when compared with regular communication with parents). Another important enabler, with a mean of 4.41 (SD = 0.87), was ICT tools, which can also assist teachers in planning and managing DI (Benjamin, 2014; Smith & Throne, 2007, Stanford et al., 2010). Similarly to access to differentiated materials, ICT tools also received significantly higher mean values compared with the other enablers (t(211) = 1.98, p = .049 when compared with exchange of DI-related tips among staff members; t(211) = 2.61, p = .010 when compared with support from school administrators; t(211) = 3.30, p = .001 when compared with exchange of student information among staff members; t(211) = 4.87, p < .001 when compared with in-service training on DI; t(211) = 6.90, p < .001 when compared with teaching staff’s positive attitude towards DI; t(211) = 8.65, p < .001 when compared with regular communication with parents). Besides, in responses to the open-ended question on enablers, teachers emphasised the importance of educational policy measures to alleviate their workload (5 mentions) and to create smaller, more homogeneous EFL groups (6 mentions), which are consistent with the contextual challenges highlighted earlier, and would also help mitigate increased planning demands.

Intra-institutional cooperation, such as the exchange of DI-related tips (M = 4.28, SD = 0.91) and student information (M = 4.17, SD = 0.93) among staff members and support from school administrators (M = 4.24, SD = 0.96) also received high mean values. This indicates that teachers value the supportive power of what has been referred to as cohort groups (Theisen, 2002), study groups (Tomlinson & Allan, 2000, p. 81) or professional learning communities (De Neve & Devos, 2016) at schools, as well as school administrators’ efforts in nurturing such cooperation.

Finally, it is worth highlighting that while methodological uncertainties were seen as less problematic by the participants, they did rate in-service training on DI as an important enabler of DI (M = 4.00, SD = 1.08), which indicates a strong need for further professional development in differentiation techniques. Regular communication with parents (M = 3.64, SD = 1.13) also received a higher than moderate value, which shows that most teachers find this factor important in facilitating DI practices.

 

Before presenting the results of each of these regression analyses, I first provide an overview of the descriptive statistics of the scales that measured the presence of the various enablers of DI (see Table 25). For data reduction purposes, I did not include each of these enablers individually into the regression analyses but organised them into broader categories when measuring their perceived presence. The categories were based on the enabler-related themes identified in Study 1: professional preparedness (participation in in-service training, feeling methodologically prepared, availability of differentiated materials), cooperation with stakeholders (exchange of DI-related tips and student information among staff, staff’s positive attitude towards DI, communication with parents), and ICT tools (presence of ICT tools, teacher’s use of ICT tools, students’ use of ICT tools). This categorization, validated through reliability analysis in both the pilot and main study via Cronbach’s Alpha and PCA, confirmed that these items form unidimensional constructs and thus could be treated together (for a detailed description of this process, refer to Sections 3.2.2.2 and 3.2.2.4, respectively).

The outcomes of the regression analyses are summarised in Table 26. Several patterns merit attention here. Firstly, it seems that self-efficacy beliefs in DI do not only have a significant positive influence on all four dimensions of DI implementation, but also act as the most influential predictor variable in each of these models (readiness-based DI: β = 0.28, p < .001, interest-based DI: β = 0.30, p < .001, learning profile-based DI: β = 0.34, p < .001, differentiation of the learning environment: β = 0.28, p < .001).

It is also interesting to note that, to a smaller extent, beliefs in the importance of TEDI act as a predictor of readiness-based DI (β = 0.19, p < .05), indicating that an acknowledgement of the role of ICT in enhancing DI practices positively influences teachers’ decisions to address their students’ readiness levels. In a similar vein, the use of ICT tools seems to act as an enabler of learning profile-based differentiation (β = 0.16, p < .05) and the differentiation of the learning environment (β = 0.20, p < .05). This suggests that having opportunities to use technology has a positive influence on teachers’ decisions to cater for their students’ learning profiles and to create a learning environment conducive to differentiated learning.

What may seem interesting is that out of the four DI dimensions, beliefs in the importance of DI were found to be influencing only one aspect, the differentiation of the learning environment, but it was not identified as a predictor variable of any other DI dimensions. This indicates that teachers’ confidence in their ability to implement DI plays a more important role in their decision to use DI than their opinions about the approach itself.

It is worth taking a look at the variables predicting TEDI practices, too. Results show that teachers’ use of TEDI is influenced by their beliefs in the importance of TEDI (β = 0.48, p < .001), as well as by their professional preparedness (β = 0.23, p < .001), by the use of ICT tools (β = 0.23, p < .001) and by their beliefs in the importance of DI (β = 0.11, p < .05). It seems, therefore, that both positive perceptions of DI and TEDI and the use of digital infrastructure are important impetuses for teachers to use ICT for DI purposes.

Before discussing these results in conjunction with those from Study 1, it is important to issue a cautionary note. As indicated in Table 26, except for the regression analysis concerning TEDI practices (R2 = 0.57), the models exhibit relatively low explanatory power. This indicates that there must be other significant predictor variables, not examined in this study, that influence teachers’ decisions to adopt DI, and further exploratory research is needed to uncover and understand these factors.