The emergence of household learning surveys, and ensuing debates

• In Event: Beyond Classrooms: Measuring Learning Through Household Surveys

Proposal

The use of learning assessments has a long history in the social sciences. Many attribute the development of such assessments in education to Alfred Binet, one of the founders of modern psychology. His goal in the late 19th century was to ascertain which French schoolchildren were most likely to succeed (author, 2018). Over the decades following, learning assessments became a mainstay in educational institutions worldwide, with large comparative studies capturing most attention, such as OECD’s PISA. Those learning assessments are used for determining systemic educational progress at national and international levels today.
Such international learning assessments have a number of serious limitations (Author, 1990; 1993), such as limited use with local languages that are not part of the top 20-25 languages used comparatively; content that is limited to the tasks most commonly used worldwide but are not in local contexts; constraints on the ability to highlight local skills and contents that can be used to improve learning equity among minority groups in varied countries (Author, 2025).
In LMICs in the 1980s, the first within-country multilingual learning assessments were undertaken in urban and rural contexts in Morocco, with an effort to understand the role of school quality, type of school (public vs. Quranic) parental education, children’s mother-tongue (Arabic or Berber), and other variables on children’s reading and math acquisition from ages 7-14 (Author et al., 1982). Later, multilingual national survey of adult reading and math was undertaken in Zimbabwe for the National Ministry of Education (Author & Srivastava, 1989). This research demonstrated the importance of localized assessments that included the most disadvantaged children.
In the 2000s, major efforts was undertaken to assess learning among young children in a variety of LMICs under the term EGRA (Early Grade Reading Assessments; Gove & Wetterberg, 2011); as well as my several non-profits, such as Pratham (2005) in India. These were regional and national level studies that were highly influential among national and international policymakers, and were able to demonstrate that simply having access to school did not guarantee reading acquisition. This in turn led to approaches to instruction that would support learning in local contexts.
One limitation of the above research is the difficult that policymakers face in describing the low levels of learning. How to compare different scales of learning for different cultural groups of children? Here, it may be useful to return to Binet. Though he and colleague Theodore Simon was not entirely happy (in 1905) with the Binet-Simon IQ score (intelligence quotient), its enormous popularity was partly due to the ease of understanding and conveying a simple concept and simpler statistic.
Today, the same problem of complex findings remains a stumbling block for expert who wish to convey more simply that nature of children’s learning and how to compare it across cultural groups without biases. Recent research, around a Gini Learning Index, has been undertaken, and will be described in this presentation,

Author

• Daniel A Wagner, University of Pennsylvania