ChemDive – a classroom planning tool for infusing Universal Design for Learning

1 Introduction

Increasing diversity in 21st century classrooms is a global reality across all school systems (Gardenswartz & Rowe, 2009; Vock & Gronostaj, 2017). To cope with the resulting challenges, teachers have to undergo professional training. This appears to be particularly relevant for future teachers. However, we currently observe a gap between research in this area and practical implementation. Regarding science teaching, current research has led to the design of effective science teaching lessons – even in highly diverse learning groups – which are however not yet used by (prospective) teachers in practice (McLeskey et al., 2018).

On one hand, research has focused on developing working approaches for teaching chemistry (Becker, 2020; Brüning & Saum, 2009a, 2019b; Johnson & Johnson, 2002; Klahr et al., 2019; Leisen, 2014; Parchmann et al., 2006; Schmidkunz & Lindemann, 1995; Stockard et al., 2018). On the other hand, cross-curricular approaches to inclusive teaching such as Universal Design for Learning (UDL) have been made available (CAST, 2018). The challenge now is to bring these two approaches together in an effective way in order to give optimal support to (prospective) teachers in planning chemistry lessons for inclusive or highly diverse learning groups. For this specific reason the ChemDive (Chemistry for Diversity) planning tool was developed, implemented, and evaluated in teacher education at university level (Holländer et al., 2022a, 2022b).

In the following, the teacher training program in which ChemDive was developed and tested, is introduced. After that, it is explained how ChemDive makes it possible to bring together common concepts of chemistry education and UDL. Finally, the initial results of the evaluation are presented.

2 Teacher training program in chemistry

In Germany, teacher training programs essentially comprise three phases, which are briefly outlined below. In phase I, the prospective chemistry teachers first complete the bachelor’s degree program. This is followed by a master’s degree. The master’s program includes a one-semester internship phase at school. Prior to this, the future teachers are prepared in a seminar that covers aspects of planning and teaching chemistry. These include special education, UDL, and cooperative learning. ChemDive is taught as a planning tool to address all of these elements in diverse learning groups. Within the internship phase, prospective teachers participate in a supplementary seminar aimed at further enhancing their understanding of the specific content. Upon successful completion of the master’s program, teacher trainees engage in a practical school training period lasting 18 months (phase II). Subsequently, the training concludes, and teachers are eligible to enroll in continuing education courses while actively employed (phase III, refer to Figure 1).

Figure 1:

Chemistry teacher-education: Phase I is part of the teacher training program at university, phases II and III take place at school and at further training institutions.

3 Universal design for learning in chemistry lessons

UDL is a framework for designing inclusive instruction and improving overall instructional quality. It is a cross-curricular, organized set of evidence-based strategies for reducing barriers to learning. The goal is to support students in becoming learning experts themselves (CAST, 2018, 2020; Rose & Meyer, 2002). The core idea of UDL is to make learning goals and learning paths more flexible; instructions should be universally accessible to all students regardless of their individual prerequisites and possible impairments.

According to the working concepts mentioned above, chemistry lessons are usually planned in phases which are often presented in a flow chart. The question now is: How can the UDL strategies be systematically considered and successfully integrated in this lesson planning? This step often causes difficulties for (prospective) teachers (Ralabate, 2016) as the strategies, for example, cannot be clearly assigned to individual instructional phases or steps. To improve the accessibility of the lesson, these strategies need to be used in different (and often multiple) teaching phases. This is where ChemDive comes in. The planning tool brings together the usual planning board and the UDL.

4 ChemDive

ChemDive enables the systematic integration of UDL elements in chemistry lessons applying educational functions (Figure 2). The developed educational functions are loosely organized into an opening phase, a main phase, and a closing phase. Thus, they integrate easily and ensure compatibility with other planning concepts. However, the planning tool can also be used on its own.

Figure 2:

ChemDive with educational functions (numbered from 1 to 15).

Using the example of the functions “Grab attention” and “Establish the basis & provide exemplary problem-solving strategies” (numbers 1 and 5 in Figure 2), the ideas behind the elements of ChemDive are illustrated.

The column indicates the phase in which the respective functions need to be considered. “Grab attention” should be implemented immediately in the opening phase of each lesson. Typical teaching elements are: Welcoming students, clarifying issues, resolving social conflicts, checking formalities. This implements UDL checkpoint 7.3 “Minimize threats and distraction”. In consequence, conflicts, problems, or obstacles for the current lesson should be reduced.

The function “Establish the basis and provide exemplary problem-solving strategies” can be employed at the conclusion of the opening phase. This function guarantees that the necessary knowledge to effectively tackle the problem or task is accessible to students in a variety of ways. Teaching components may include: Showcasing exemplary action steps, demonstrating strategic and planning steps, offering examples and counter-examples of learning products. The purpose of this educational function is to establish a groundwork for the subsequent work phases, during which students are expected to work independently from the teacher. Thus, UDL checkpoints 5.3 “Build fluencies with graduated levels of support (Scaffolding) for practice and performance”, 6.2 “Support planning and strategy development”, and 6.3 “Facilitate managing information and resources” can be realized.

5 Research questions and design

The preparatory seminar first dealt with how to handle curricula and other specifications for teaching chemistry. Then it was about the analysis of the learning group and the diversity of the learners: What are the different learning prerequisites of the students in the classroom? Special needs such as in emotional-social development or learning were also discussed. UDL was then taught as an interdisciplinary teaching approach that considers the diversity of learners. A video showing a lesson was analyzed for accessibility and the lesson design was enriched with possible implementations for individual UDL elements. Cooperative learning was not only practiced in all units, but also reflected in one slot regarding its suitability for highly diverse learning groups. Finally, the ChemDive planning tool was taught and demonstrated in a detailed example. Different educational functions were exemplarily implemented and reflected. In the planning exercises, for example, the frequently missing possibility of different ways of presenting results (educational function 10) could be taken up and supplemented by concrete implementation possibilities. With regard to the educational function “creating attention”, it was important to sensitize to the many disruptions that can occur in the classroom and, in terms of classroom management, to deal with them as far as possible before the lesson. The prospective teachers benefited from the exchange of their different school experiences. For the educational function of “Establish the basis and provide exemplary problem-solving strategies”, both the content aspects of the main phase and the learning requirements must be anticipated. This was practiced and reflected on another given teaching project (cf. Figure 1). In conclusion, the seminar participants applied all their skills about lesson planning in a simulation exercise. For this purpose, a lesson was planned in groups, acted out in a role play and reflected on together. During the internship semester itself, the prospective teachers utilized ChemDive for planning their chemistry lessons in schools. Within this period, they attended a complementary seminar, which took place for a total of 3 days, spread over the internship semester. Here, the prospective teachers exchanged experiences from school practice and discussed selected educational functions of ChemDive (cf. also Figure 3).

Figure 3:

Research design.

To evaluate the effectiveness of the seminar with focus on accessible chemistry instruction and on ChemDive itself, the following research questions were addressed:

To answer the questions Q1, Q2 and Q3, a pre-post design during the preparatory seminar was used (cf. Figure 3). In order to capture the planning skills of the future teachers, an identical task was set before and after the intervention: A previously published lesson plan (cf. Häusler & Pavenzinger, 1993) was to be modified with the aim of adapting it to a highly diverse learning group, ensuring maximum accessibility for all students. The accomplishment of this task (documented in a planning board and explained in an interview) was analyzed and evaluated using a coding manual (see also results).

The intervention consisted of a 12-session preparatory seminar that included lessons from UDL and ChemDive. During the internship semester, the contents of the preparatory seminar were applied and in addition, the prospective teachers were further supported in an complementary seminar and the contents on ChemDive were deepened. At the end of the internship semester, an additional questionnaire (Jasper, 2021) on the use of ChemDive in lesson planning in school practice was employed to answer question Q4 (cf. Figure 3).

Our following analyses include a total of N = 37 prospective teachers from 4 different cycles from winter semester 2020/21 to winter semester 2022/23 (cf. Figure 4). To be included in the preparatory seminar sample (top row in Figure 4), participants had to be present at both measurement points. All future teachers in the complementary seminar sample (bottom row in Figure 3) completed a survey.

Figure 4:

Intervention schedule (due to the Covid-19 pandemic we had to conduct the seminars in winter semester 2020/21 and summer semester 2021 online, before we could switch to the face-to-face format.).

6 Results

To answer research questions Q1a and Q2a, we developed a specific manual to rate the modified lesson plans (28 items, Likert scale from 0 (not implemented) to 3 (fully implemented), ICC = 0.746, Gutt, 2021, based on Schlüter & Melle, 2018). This manual allowed us to determine the extent to which the UDL strategies were implemented by the prospective teachers. For each lesson plan, all 28 items were scored, resulting in a maximum total score of 84. At the end of the coding process, the total scores were normalized with respect to the maximum score to make the results on UDL comparable to those on ChemDive, for which a different coding manual with a respective Likert scale from 0 (not implemented) to 2 (fully implemented) was used (see Figure 5a and b and see below for details).

Figure 5:

Results normalized with respect to the respective maximum score regarding the accessibility of the prospective teachers’ lesson plans in terms of (a) UDL and (b) ChemDive before (“pre”) and after (“post”) the seminar. The standard deviation from the mean is depicted by the error bars.

Using a paired-samples t-test, we compared the prospective teachers’ use of UDL strategies before (“pre”) and after (“post”) the preparatory seminar (see Figure 5a). The result was that the seminar participants (N = 37) were significantly more likely to implement the UDL strategies after the seminar than before with a large effect (M_pre = 0.30, M_post = 0.50, t (36) = 9.533, p < 0.001, Cohen’s d = 1.286). The effect size of Cohen’s d = 1.256 indicates that the overall increase in competency associated with the inclusion of UDL elements in lesson planning is substantial.

In Figures 5a and 6 can be seen that some of the seminar participants already had considerable prior knowledge about UDL before the seminar started, which also explains the comparatively high mean value of M_pre = 0.30. Furthermore, after the seminar (“post”) a relatively large spread around the median can be seen (Figure 6). This can be explained, among other issues, by the fact that the task in the evaluation is to be considered a transfer task and thus was subject to influencing factors outside the teaching of the seminar content. Control variables such as cognitive abilities or self-efficacy, which could eliminate these effects, were not recorded.

Figure 6:

Boxplot of results normalized with respect to the respective maximum score regarding the accessibility of the prospective teachers’ lesson plans in terms of (a) UDL and (b) ChemDive before (“pre”) and after (“post”) the seminar. Circles mark statistical outliers.

To further elaborate on research question Q2a, which pertains to the development of the prospective teachers’ planning competence during the seminar with respect to individual UDL guidelines (CAST, 2018), the total score was divided accordingly (see Table 1 and Figure 7).

When looking at the scores for the individual guidelines, varying levels with relatively wide spread of competence can be seen; nevertheless, a substantial and significant enhancement in each competence is noticeable across all of them (see Table 1 and Figure 7).

In order to answer research questions Q1b and Q2b, i. e. the extent of the prospective teachers’ implementation of ChemDive and the seminar’s effect thereon, we used a separate coding manual (108 items, Likert scale from 0 (not implemented) to 2 (fully implemented), Gutt, 2021, Cohen’s κ = 0.45, α = 0.926). It must be pointed out at this point that the implementation of all 108 points in one lesson is neither expected nor useful. The coding manual is based on all specified lesson elements that can be used to implement the educational functions. Therefore, values of 1 for the normalized results are not to be expected. Again, a significantly higher level of planning competence could be observed regarding the implementation of the educational functions of ChemDive (see Figure 5b). Again, lesson plans created by the prospective teachers were compared before and after attending the seminar using a paired samples t-test. It is evident that the prospective teachers implement educational functions significantly more frequently after the seminar than before with a considerable effect (N = 37, M_pre = 0.10, M_post = 0.28, t (36) = 10.172, p < 0.001, Cohen’s d = 1.448). As with UDL, the ability to use ChemDive in lesson planning is shown to increase meaningfully. The effect size is even slightly higher with Cohen’s d = 1.448, which can be explained by the lower prior knowledge compared to UDL (M_pre = 0.10). Missing prior knowledge gained by other courses can also explain the lower spread of the results before the seminar. The scatter after the seminar is very comparable to the results for UDL. This can be explained analogously with the demanding test task with e.g. high demands on the higher-order cognitive abilities.

To further examine research question 2b, which focuses on the development of prospective teachers’ planning competence towards individual educational functions during the seminar, we analyzed the score per each individual function. As was already the case with the UDL guidelines, our analysis indicated differing levels of competence across the different functions; once again however, a significant improvement in competence was observed across all functions (cf. Figure 8).

Figure 7:

Results regarding the implementation of individual UDL Guidelines in seminar participants’ lesson plans. The standard deviation from the mean is depicted by the error bars.

Figure 8:

Influence of the seminar on the execution of the individual ChemDive educational functions in the prospective teachers’ lesson plans. The standard deviation from the mean is depicted by the error bars.

Even more than with the UDL, the results scatter strongly (see error bars in Figure 8). The file analysis by means of boxplot (cf. Figure 9) shows that the standard deviation is almost exclusively caused by a considerable number of statistical outliers. It is essential to exercise caution and adopt a restrained approach when interpreting the results presented in the breakdown. These findings should be considered as semiquantitative rather than providing definitive quantitative conclusions. To strengthen the validity and generalizability of the study, it is advisable to increase the sample and incorporate further variables, as previously described.

Figure 9:

Boxplot of individual ChemDive educational functions in the prospective teachers’ lesson plans. Results Circles mark statistical outliers, asterisks indicate extreme outliers. EF stands for educational function, followed by the number of the function and the measurement time (pre or post).

Furthermore, the comparison of the pre- and post-survey results shows that the prospective teachers not only enhance their knowledge of ChemDive substantially, but are also able to apply it actively.

After a positive development was observed during the seminar, the investigation was carried out to determine whether there was a correlation between the implementation of both UDL guidelines and ChemDive educational functions (research question Q3). The correlation of the means for the overall UDL guidelines and ChemDive educational functions implemented in the pre- and post-lesson planning was calculated (see Table 2). The strong positive correlations at both measurement points show that UDL is well-implemented when lessons are planned with ChemDive. This is a gratifying result, as ChemDive was developed to facilitate the implementation of UDL in subject-specific instruction.

The final research question Q4 relates to the questionnaire that was used during the internship semester (N = 20 from the summer semester 2021 and winter semester 2021/22, N = 8 from summer semester 2022, cf. bottom row in Figure 4). The prospective teachers were supposed to indicate whether they had implemented the individual functions of ChemDive. If they answered yes, they were required to give examples. These answers were then evaluated by an expert (author and another well-trained scientist). Insights for adapting the seminars were derived from the results of the questionnaires administered during the summer semester 2021 and the winter semester 2021/22 (see Figure 10), which are briefly outlined below. The effects of these modifications can be observed through the presentation of the results from the summer semester 2022 (see Figure 11). However, it should be noted that the sample size is reduced due to this approach, thus rendering the results semiquantitative in nature.

Figure 10:

Prospective teachers-reported implementation of each educational function and the actual implementation according to the expert evaluation. The numerical values and the graphical representations in the form of bars reflect the frequency of prospective teachers who successfully implemented the respective educational function. (Analysis of the questionnaire responses from the complementary seminars conducted in summer 2021 and winter 2021/22, N = 20).

Figure 11:

Prospective teachers-reported implementation of each educational function and the actual implementation according to the expert evaluation. The numerical values and the graphical representations in the form of bars reflect the frequency of prospective teachers who successfully implemented the respective educational function. (Analysis of the questionnaire responses from the complementary seminars conducted in summer 2022, N = 8).

For 9 out of 15 educational functions (as depicted in Figure 10, highlighted in blue), there is a correspondence between the self-assessed implementation reported by prospective teachers and their actual implementation. Additionally, there are certain functions where discrepancies between self-assessment and actual implementation exist (as shown in Figure 10, highlighted in orange). This is e.g. remarkable for the function “Grab attention”. The future teachers frequently justified the implementation of this function by selecting interesting and motivating lesson introductions. However, this does not align with the specified teaching elements such as welcoming students, establishing presence and concentration, and creating safety in the classroom through routines, which is why it was deemed non-implemented by the experts. Additionally, functions that generally were less in use, could be identified. The seminars content was thus be re-sharpened concerning these functions. In Figure 11, it can be observed that the repeated discussion of educational functions “Grab attention” and “Establish the basis and provide exemplary problem-solving strategies” resulted in not only increased consideration but also improved implementation of these functions. As a further step, supplementary explanatory videos that provide a clear explanation of the educational functions “Grab attention” and “Establish the basis and provide exemplary problem-solving strategies” have been produced. The videos were integrated into the seminar sessions since winter semester 2022/23 and concurrently made accessible on the digital learning platform throughout the internship semester in summer 2023. However, there are still some educational functions that have not been satisfactorily implemented, particularly in the main phase. Further refinement is needed in this area. This can be achieved, for example, by providing more examples of potential supportive elements such as digital tip cards.

7 Conclusions

This paper addresses the lesson planning tool ChemDive, a tool for improving future chemistry teachers’ ability to plan accessible chemistry lessons by systematically incorporating elements of the Universal Design for Learning. In a preparatory seminar at university, participants were taught the principles behind the tool and how to use it. For instance, various learning prerequisites and limitations among students are addressed, which result in barriers to learning and understanding chemistry content. UDL is introduced as a means of designing inclusive instruction. The application and implementation of UDL using ChemDive are practiced through various instructional planning formats. Regular reflection discussions are an integral part of the seminar. Future teachers report the most substantial learning gains from the final seminar exercise, where they plan a lesson in groups, simulate a role play in the seminar, and then reflect on it collectively. The results presented in this article illustrate selected findings from the evaluation: The prospective teachers’ skills regarding both UDL guidelines and ChemDive educational functions improved significantly during the seminar, with almost all guidelines and functions being implemented significantly more often. In addition, the UDL guidelines were best implemented when ChemDive functions were also considered. In summary, ChemDive proved to be a promising tool for planning chemistry lessons that are accessible in terms of UDL. The results also indicate that the preparatory seminar provides a solid basis for a first contact with the tool. However, even though we observed that the prospective teachers were able to improve in their use of ChemDive as a result of attending the seminar, the still low mean values for the implementation of some educational functions (see Figure 8) imply that there is still much room for improvement. Since lesson planning is complex in itself (Vogelsang et al., 2022) and will certainly become even more complex with the additional consideration of accessibility (Ralabate, 2016) teacher training must not end at this point. On the contrary: It would be, on one hand, optimal if students could work with similar planning models at other points in their university education; it is however much more important that they continue to work with ChemDive later on, following an approach of lifelong learning. It is therefore urgently necessary to carry the knowledge and approaches gained into the second and third phases of teacher training (see Figure 1) and to design appropriate advanced training courses. By doing so, we can also reach teachers who are already working in schools at the moment.