How Teacher Expectations and Parent Involvement Predict Children's Summer Reading Behaviors

• In Event: 73.047 - The Not-So-Lazy Days of Summer: Interventions to Improve Achievement and Attainment

Abstract

The purpose of our study is to examine whether and to what extent teacher expectations and parent involvement predict the amount of summer book reading among treatment group children who were part of a large experimental study. In summer 2013, we conducted a randomized experiment to evaluate the effects of a scaffolded summer reading intervention, in which students were mailed matched books and reading activities over the summer, in 59 high poverty and mid-high poverty elementary schools in North Carolina. As part of the intervention, we collected baseline data in which teachers were asked to rate each child in his/her classroom on two measures which are the focus of this study. First, teachers provided subjective ratings of each child’s reading level (i.e., the quartile of the reading level of each child on a national norm referenced test). Second, teachers were asked how often each child’s parent/guardian participated in school conferences, volunteered in school, attended school open houses, and attended art/music events at schools. Teachers in 464 homeroom classrooms completed the two measures.

At the end of the school year, children were randomly assigned to control or treatment groups. In the treatment group, teachers implemented 6 comprehension lessons preparing students for the summer reading activities. Parents were invited to attend a family literacy event where parents learned how to monitor their children’s use of the reading routines with matched books that were mailed home during the summer months.

We will address two questions in this paper. First, how do teacher expectations predict the amount of summer book reading? Second, how do teacher ratings of parent involvement at school predict the amount of summer book reading? To address the first question, we created a teacher expectation measure by taking the difference between the teacher rating and the child’s actual performance on a national norm-referenced test (Iowa Test of Basic Skills). For example, if the teacher rated the child as above the 4th quartile nationally (i.e., 75-100th percentile) and the child scored between the first and second quartiles (i.e., 25-50th percentile), the teacher expectation score would be 2 (4 –2). We computed an expectation score for each child in the study and created a mean expectation score for each of the 464 homeroom classrooms. Overall among treatment group children who received books the summer, mean teacher expectation scores predicted the amount of book reading, as measured by the reading trifolds that children completed. These results were robust across a number of model specifications (see Table 1).

To address the second question, we created a mean parent involvement score for each classroom. Similarly, we found consistent evidence that mean parent involvement levels predicted more book reading amount children in the treatment group.

We caution that the results here are not causal because children were not randomly assigned to classrooms that varied in their teacher expectation or parent involvement levels. However, the robust correlations suggest that teacher expectations and parent involvement during the school year predicts children’s reading behavior during the summer.

Authors

• James S. Kim, Harvard University

• David Quinn, Harvard University