AEA RCT Registry currently lists 12514 studies with locations in 171 countries.
The purpose of this study is to evaluate whether a universal student support intervention that provides personalized video feedback, recommendations, and action-oriented tasks can improve students' academic and college experiences. The intervention is designed to address common challenges experienced by college students, including study strategies, time management, financial stress, social belonging, and other barriers to academic success. We will examine whether students who receive personalized support demonstrate improvements in outcomes such as academic performance.
While feedback is an important part of many real-life situations, how to give feedback remains a hotly debated topic. This paper uses a field experiment to understand whether the ordering of feedback elements matters. University students get positive and negative feedback on exam practice questions, in varying order. The main outcomes are performance and motivation, further outcomes are beliefs and study behavior.
This study evaluates the impact of an AI-powered math tutoring platform and an AI career coaching chatbot on learning outcomes and educational aspirations among 5th-year secondary students in public schools in Lima, Peru. The AI math tutor covers Peru's national math curriculum and is used during regular math class time in school computer labs, supervised by students' regular math teacher. The AI career coach helps students explore post-secondary options, understand economic returns to education, and develop concrete career plans during existing tutoría (guidance) periods. The study uses a cluster-randomized design with approximately 100 schools as the unit of randomization and three study arms: (1) AI tutoring with learning-science modifications plus career coaching, (2) AI tutoring...
Low levels of early childhood literacy remain a major constraint to human capital accumulation in low- and middle-income countries. While parental engagement in early learning activities—such as reading to young children—is known to be crucial, take-up of these practices is often limited by constraints beyond material resources, including parental beliefs, social norms, and intra-household decision-making. This paper studies a large-scale clustered randomized controlled trial implemented in Jordan that aims to increase parent–child reading among households with children aged 1 to 18 months. The intervention combines a book-gifting program delivered through public health facilities with informational components targeting parents’ beliefs about the returns to early reading and prevailing ...
In the Netherlands dissatisfaction with the current scheme of social welfare has led various Dutch cities to run field experiments to test out new ways of delivering income support to its citizens. Among them is the City of Utrecht, where the study goes by the name ‘Weten wat werkt’ (English: What works) and includes more than 700 current recipients of public income support (also social assistance or minimum income guarantee). For the duration of the trial job search requirements and welfare sanctions tied to income support will be eliminated to make room for more autonomy (treatment 1). Other aspects include a more tailor-made approach in supporting welfare recipients (treatment 2) and increasing financial incentives to take up paid work (treatment 3). The study is set up as a randomiz...
We measure the causal impact of comment sections on social media posts promoting racial justice. Using online survey experiments, we vary exposure to the comment section and estimate its impact on beliefs, attitudes, and behavior.
We measure the causal impact of comment sections on social media posts promoting racial justice. Conducted in two phases, it involves an online field experiment with over millions of Facebook users. We first collect and analyze comments to understand how political ideology influences engagement with posts. We then evaluates how exposure to different comment narratives affects subsequent engagement (comments, views), intentions (clicks), and off-platform behaviors (mailing list sign-up, donations).
This laboratory experiment studies how individuals use generative AI tools to complete computer-based, cognitive tasks, and how the productivity effects of AI vary across users with different baseline skill profiles. Participants complete problems in two domains: 'analytical' and 'social'. The experiment has two phases. First, participants work without AI. Second, participants are randomly assigned according to a 2×2 design that crosses AI access (AI vs. no AI) with a brief intervention (practical AI training vs. active control). We measure task performance, beliefs about own performance, AI usage from chat logs, and a range of baseline cognitive and personality measures.
We experimentally test whether individuals' causal belief systems are well described by the workhorse model of the literature on causal cognition and misspecified mental models: Causal Bayesian Networks (CBN). We ask two questions. First, are belief systems rationalizable by the theory that individuals process data through the lens of a CBN or a distribution over CBNs? Second, do individuals' beliefs and choices exhibit the comparative statics predicted by the literature on misspecified learning? Participants interact with computer-generated systems involving binary variables, make incentivized choices, and report beliefs about causal relationships.
Despite widespread access to online job postings, large disparities in employment outcomes persist among similarly qualified candidates, suggesting that informational frictions extend beyond vacancy access. This paper argues that process knowledge about how hiring processes operate is a key and understudied source of advantage. Unlike vacancy information, process knowledge is unevenly distributed through social networks and is strategically costly to share in competitive, rank-order selection environments. Using a lab-in-the-field experiment with university students in Pakistan competing for real jobs, the paper causally examines how access to process knowledge affects information-sharing behaviour.