Intervention (Hidden)
To sample our study population, we rely on the project “Corona & Du” (CoDu), which sampled families with children aged 10 to 17 from the Social Security Records of the German Federal Employment Agency. In doing so, the survey also sampled from Unemployment Benefit II Recipient History , and, as a result, a significant share of participants come from families dependent on social welfare (Social Code Book). The intervention will be implemented within the group of fifth and sixth grade students who participated in the baseline survey in fall 2020. The intervention concentrates on a disadvantaged sample, as we expect the effects of e-book readers on literacy to be largest among this group. Indeed, young schoolchildren living mainly in households depending on social welfare are an understudied group to which this panel gives us a unique access.
The CoDu study has already collected a large number of educational information and background variables for each participating child during the pandemic. Variables such as reading skills, educational aspirations, and detailed information on parental background– from both the survey and the registers – are of central importance for our RCT on reading competencies and educational attainment. These variables can both serve as controls and might be used for testing randomization. Moreover, we minimize the measurement error related to recall bias, as the necessary pre-treatment information was collected in a timely manner. Another reason for relying on the CoDu sample is methodological. The availability of rich pre-treatment information allows for more efficient randomization. In a nutshell, we can use this data to match students with similar observable pre-treatment characteristics and then conduct stratified randomization. We stratified our sample based on the following variables: (i) an indicator of receiving social assistance, (ii) sex of the child, (iii) school grade of the child, (iv) migration background of the child, and (v) median split of baseline reading literacy.
Furthermore, the geographic spread of CoDu participants and by implication their schools across different federal states reduces concerns about spillover effects from the treatment to the control group compared to a setting, in which students are sampled only from a single school. Finally, given the respondents’ agreement, the CoDu project allows us to link the survey data to the register data of the Federal Employment Agency – the Unemployment Benefit II Recipient History (LHG) and the Integrated Employment Biographies. Thus, we are able to observe the current labor market situation and welfare dependence of the participating families and to analyze labor market outcomes of the children in the future.
To promote disadvantaged children on their educational pathways and on their transition to the labor market, we examine the extent to which a support measure can lead to improvements in school performance and educational attainment of these children. We focus on children in grades five and six, as these are the years also represent a time of critical educational transition in Germany. In most federal states, parents – accompanied by teacher recommendations – decide whether children move to the lower, middle, or academic track after grade four. In grades five and six, this secondary school choice may have to be adjusted according to the child’s achievement.
One important determinant of academic achievement (in a variety of subjects) and thus of later educational success is children’s reading literacy, as several studies have shown (e.g. Kern and Friedman 2008, Hernandez 2011). While studies document the early emergence and persistence of achievement gaps by socioeconomic background (Herbers et al.2012), the joy of reading seems more important for children's educational success than the socioeconomic status of their family (OECD 2002).
In November 2021, the households assigned to the treatment group receive a letter that announces that they have the opportunity to participate in an e-book reader study, where their child in the fifth or sixth grade will receive an e-book reader with access to more than 1.000 books for one year. The letter describes how the study will proceed and asks parents to opt out if the household does not want to participate. We do not replace the households that opted out. About three weeks later, the households receive the e-book readers, which we configure in advance to be in the child protection modus. The e-book readers already contain access to about 1.000 books recommended for the children’s age group.
Part of our intervention is an about monthly mailing of book recommendations. For this purpose, we collaborated with Stiftung Lesen (a charitable foundation aiming to encourage reading) to recommend age appropriate books and to create short videos for a subset of books. With this we aim to reinforce the intensity of the treatment. These book suggestions are also tailored to the children’s “book access”. We host the videos on a secure platform, where only the treated children can access them. We further aim to gain access to individual data capturing the intensity of reading and the types of books read to investigate whether the reading recommendations increase reading intensity.
We plan to conduct at least two separate follow-up waves of the survey with around half a year between each wave. We plan the first follow-up in the second quarter of 2022 and the second follow-up in the fourth quarter of 2022. For both follow-ups, we will send out an invitation to the online survey to the children via their households.