This is the repository for the ICIS 2018 symposium on "From pre-registration to ethical data peeking: Practical solutions for improving infant research" Feel free to browse slides, code, and get in touch if you have any questions. A preconference on “best practices” at ICIS2016 brought together infancy researchers from a variety of backgrounds to identify the challenges the field faces in improving research integrity, which arise due to due to noisy data, small sample sizes, and lab-to-lab variation in protocols. This symposium will continue building these bridges at ICIS2018, by identifying practical solutions that work in the context of infant research. Paper 1 presents a summary of 14 meta-analyses, providing an unprecedented picture of typical effect sizes across a range of infant experimental paradigms. Such data are crucial for performing accurate power analyses. An implication of the results is that infancy researchers need to substantially increase typical sample sizes, a proposition that can raise serious practical concerns. Paper 2 will address such concerns by providing ways to achieve this goal while retaining some flexibility during data collection. The authors propose ethical ways to decide whether to continue data collection after peeking at the data. One concern may be that this flexibility increases the false positive rate. This is addressed in Paper 3, which discusses pre-registration, and addresses barriers that may hold infancy researchers back from using this in their own work. Finally, figures are a central way to communicate results, but suboptimal visualizations can lead to less informative or even misleading representations of the underlying data. Paper 4 presents new ideas for informative visualization of data from infant experiments. Our symposium will continue the discussion on best practices while acknowledging the realities that infancy researchers face, in an effort to maximize both research payoff and integrity.