Martine van der Pluijm

166 Chapter 6 decontextualized language and other language) improved in Activity 2 but not in Activity 1. This finding implies that better delivery by teachers leads to gains in important aspects of parent-child interaction in Activity 2, but not in Activity 1 even with a good delivery quality. We first explain the positive effects that we found for both activities. Interestingly, we found gains in several aspects of the quality of interaction irrespective of the quality of delivery of teachers in classrooms. Both activities were designed to stimulate the joint attention of dyads and to require minimum teacher preparation. However, the activities were designed for phased implementation. Activity 1 was aimed at lowering the threshold for parents to become involved in parent-child activities in the classroom. During this activity, parent and child were encouraged to talk about other family members. In classrooms, this would enable teachers to join the dyads, listen to their conversation, and exchange backgrounds. Activity 2 was designed for use in classrooms when parents were familiar with parent-child activities. This activity was aimed at stimulating rich interaction between parent and child. The dyads played hide and seek (What’s gone?) with wooden fruit or with Rory’s Story cubes, taking turns, eliciting language, and having fun. Our results showed that both activities improve child initiative (i.e., increasing child involvement) and emotional support by parents (e.g., parental encouragement and support of autonomy) with little effort from teachers. An explanation for this finding that both our talk and play activities led to more interaction is that even in the condition of minimal coaching the context in classrooms may have triggered parental interactive behavior. After all, all teachers provided guidance, involving parents in weekly parent-child activities to talk together. Experience in these activities may have led to the gains found for child involvement, autonomy, and encouragement. In addition, parent behavior may have been influenced by examples of other parents in the classroom. The literature on adult learning shows how parents in children’s classrooms can be role models for each other and how these models can impact parent behavior (Fantuzzo, Stevenson, Kabir, & Perry, 2007; Prins & Van Horn, 2012). Dyads seem to have benefited from this type of learning that requires limited teacher coaching. Our analysis shows that this was the case for aspects of the quality of interaction, which may be relatively easy to learn by being repeatedly involved in both of our parent-child activities. A remaining question is why quality of delivery affected parental cognitive responsive behavior, quantity, and quality of language in Activity 2 but not in Activity 1. As discussed in Chapter 5, the different nature of these activities might have played a role. Activity 1 may be less sensitive to instruction and coaching directed at turn-taking and having fun compared to Activity 2. Activity 2 becomes more joyful when parents encourage their child to think and talk, instead of directing the child to give correct answers. For example, parents can pretend not to know the right answer when it is their turn, encouraging children to help them by giving more details and using more words (i.e., scaffolding). Lower-educated parents may not be familiar with these kinds of strategies (See also Chapter 3). For this reason, explanations and modeling