In this study, we investigated the effects of a PDP based on the model of contingent teaching (consisting of the four steps: (1) diagnostic strategies, (2) checking the diagnosis, (3) intervention strategies, and (4) checking students’ learning) on teachers’ scaffolding behaviour. Whereas previous scaffolding studies mostly used highly structured tasks in one-to-one or lab-situations, we studied scaffolding in the classroom context using open-ended tasks. Furthermore, hardly any experimental studies on scaffolding in classroom situations exist; we conducted such a study with a relatively large sample size. The model of contingent teaching was used in a PDP to teach teachers how to scaffold.
The model of contingent teaching appeared to be a powerful tool in learning how to act contingently upon students’ understanding. The teachers who participated in the PDP learned to perform high quality steps of contingent teaching. They elicited student answers with a higher quality mode (demonstrations of understanding instead of claims of understanding) than the teachers who did not participate in the PDP when using diagnostic strategies (step 1) and when checking students’ learning (step 4).
Furthermore, the support they gave was more contingent, both in terms of adapting the level of control to students’ understanding and in terms of the extent of uptake of students’ contributions. An overview of the findings can be found in Table 6 and the findings per step of contingent teaching will be discussed into more detail in the following.
Table 6
Overview of the Findings of this Study
Quantity Quality-focus (more
Using the step of diagnostic strategies (step 1) was promoted by the PDP in this study mainly in terms of quality. Teachers in the scaffolding condition elicited more extensive answers (i.e., demonstrations) from students when diagnosing their actual
understanding. Eliciting demonstrations in students’ answers provides teachers with more information about the students’ understanding and facilitates contingent support.
Teachers who participated in the PDP also used more diagnostic strategies and increased their focus on conceptual understanding in this step compared to teachers who did not participate; however, these effects were not significant. The step of diagnosing students’ understanding is considered very difficult by teachers (e.g., Wittwer & Renkl, 2008) and is considered to be one of the most important teaching skills (e.g., Seidel &
Shavelson, 2007). It is a contribution of this study that it demonstrated that the quality of this skill can be increased and that a programme was developed for teaching it.
The step of checking the diagnosis (step 2) was used more often by teachers who participated in the PDP than teachers who did not participate. However, overall, the occurrence of this step still stayed relatively low. This outcome is comparable to the outcome of the study of Knežiç (2011), who also found that the use of this step – although it increased – stayed relatively low. Future research should investigate reasons for the relatively low occurrence of this step. Investigating teachers’ attitudes towards this step might prove useful as attitudes appear to be highly related to teachers’ behaviour (Avalos, 2011). Previous research (Van de Pol et al., 2012) showed that teachers who had negative attitudes towards this step – they felt this step was redundant because they already performed many diagnostic strategies – used this step to a much smaller extent than teachers who a positive attitude – they felt that with this step they showed interest and it made their teaching more efficient. As opposed to previous research, we did not only investigate the quantity of this step, but also the quality in terms of focus (factual or conceptual). Although the teachers who participated in the PDP did increase the quality of this step (they increased their focus on conceptual understanding) compared to the teachers who did not participate, this effect was not significant. This is probably related to the fact that the same effect for diagnostic strategies (step 1) was neither significant; if a teacher diagnoses students’ factual understanding, the check of the diagnosis (step 2) will consequently also be on students’ factual understanding, not on their conceptual understanding.
The most prominent finding of this study is that when providing support (i.e., step 3), we found that the teachers who worked with the model of contingent teaching, increased the contingency of their support significantly more compared to teachers who did not work with the model. In other words; the quality of this step increased significantly. Adapting support and building upon the students’ understanding has been proven to be difficult for teachers in previous research (e.g., Ruiz-Primo & Furtak, 2007).
The teachers who worked with the model, however, learned to adapt the support given to the understanding of the student (i.e., contingency-control) and to what the students said (i.e., contingency-uptake). The model of contingent teaching and especially
diagnosing and checking while electing demonstrations stimulated this contingency.
Using the step of checking students’ learning (step 4) was promoted by the PDP mainly in terms of its quality. Teachers who participated in the PDP elicited demonstrations (instead of just claims) of understanding from students more often when checking whether students’ understanding had improved through the support given, compared to teachers who did not participate in the PDP. Teachers who participated in the PDP also used this check more frequently and focused more on students’ conceptual understanding (as opposed to factual understanding) compared to teachers who did not participate. However, these effects were not significant. Previous research showed that when checking students’ learning, teachers mostly elicit short and simple answers (Kao et al., 2011). Therefore, having found a way to increase the quality of this step in terms of mode elicited is an improvement. The fact that teachers who worked with the model of contingent teaching did not increase their focus on conceptual understanding compared to teachers who did not work with the model, might be explained by the fact that when the whole conversation was on factual understanding, it seems illogical to suddenly shift to checking students’ conceptual understanding. So because the effect was not significant for the first step (diagnostic strategies), it makes sense that this effect was also nonsignificant for the subsequent steps of contingent teaching.
Overall, an increase in quality for teachers who worked with the model of contingent teaching was mainly found in the fact that the teachers elicited demonstrations more often when using the steps of contingent teaching; not so much through a shift in focus towards conceptual understanding. It could be the case that exploring and checking students’ understanding on a conceptual level and simultaneously thinking of how to use the gathered information in the support that is to follow was just too much for the teachers in the scaffolding condition. Feldon (2007) for example suggests that teachers might suffer from cognitive overload when both having to deal with complex subject matter and with exploring another person’s understanding of that subject-matter at the same time. Focusing on conceptual knowledge can be considered as more complex than factual knowledge which might explain our finding. Future research could explore ways to reduce the load for teachers by for example using visual representations of the students’ understanding built together with the students. However, on the other hand, staying at a factual level may have been the right level for the students. Maybe exactly because the teachers in the scaffolding condition used diagnostic strategies and elicited students’ demonstrations before giving support they might have found out that moving to a conceptual level was not appropriate yet and therefore might have not moved the focus to a conceptual level. Particularly for teaching substantive concepts, scaffolding and especially diagnosing and eliciting demonstrations seems a valuable approach as scaffolding reveals and anticipates the students’ understanding of rather
abstract and theoretical concepts.
Limitations
The duration of the intervention was relatively limited, namely 8 weeks. The teachers in the scaffolding condition had four lessons to practice their scaffolding skills.
Slavin (2008) for example advised an intervention to last at least 12 weeks as changing one’s teaching practice is known to be difficult and time-consuming (Loucks-Horsley
& Matsumoto, 1999). A longer programme gives the teachers more opportunities for practicing several steps and reflecting on their lessons. On the other hand, the PDP was already quite labour-intensive. The first author filmed five lessons of each teacher that participated in the programme and in addition, reflected on each of those lessons together with the teacher. As reviews on characteristics of effective PDP’s indicate (e.g., Loucks-Horsley & Matsumoto, 1999; Timperley, 2007; Van Veen et al., 2012), working collaboratively in a community (e.g., with colleagues) can be very effective.
Future research should therefore focus on the possibilities of integrating this aspect in the PDP. In that way, the work-load for the teacher educator can be reduced while the length of the programme could be increased.
Experimental studies naturally suffer from unexpected events that happen in practice. Because we more or less disturbed an existing routine in the classrooms of the scaffolding condition, the teachers who participated in the study had much to cope with. Keeping good discipline in the classroom, for example, is probably a necessary condition to be able to perform scaffolding. However, we got the impression that keeping good discipline is more difficult when changing one’s teaching practice because the teachers need to focus on the new (in this case scaffolding) behaviour. Furthermore, while some teachers in the scaffolding condition seemed to have really incorporated the scaffolding style of teaching, some other teachers still seemed to struggle with it towards the end of the intervention. However, overall, the teachers who worked with the model of contingent teaching did increase the quality of the steps of contingent teaching, including their degree of contingency which indicates a true absorption of the idea of scaffolding; not the mere performance of a trick.
Conclusion
With this study, we have shown that the model of contingent teaching is an effective tool for promoting teachers’ scaffolding behaviour. Especially the promotion of the quality of diagnostic strategies and of teachers’ contingent behaviour is noteworthy as this is known to be a highly difficult, but also crucial teaching skill (e.g., Ausubel,
1968; Ruiz-Primo & Furtak, 2007).
A contribution of this study to both research (e.g., scaffolding effectiveness research) and practice (e.g., teacher professional development), is that it concretised the scaffolding process into discrete steps of contingent teaching and showed that working with these steps, can result in more than just using the steps: For the teachers involved in this study it resulted in true contingent behaviour.