Abstract

Understanding and interpreting box plots remains a persistent challenge for students across various educational levels, due to conceptual misunderstandings related to their key components and difficulties in connecting these graphs to real-world contexts. Through a systematic literature review conducted in accordance with the PRISMA protocol, 21 studies were analyzed that focus on identifying these difficulties and evaluating instructional strategies aimed at improving their teaching. The results reveal frequent errors related to the interpretation of the median, quartiles, whiskers, and outliers, as well as misunderstandings stemming from a superficial reading of the graph, particularly regarding data density and distribution shape. These errors reflect persistent misconceptions about the representation of variability and data distribution, hindering a deep understanding of the box plot. The documented pedagogical strategies are organized around two main axes: addressing conceptual errors and connecting with real-world contexts through technological tools and contextualized activities. Among the most relevant proposals are the use of refutational texts, the combination of multiple graphical representations, and problem-based learning grounded in real-life situations. Beyond merely correcting misconceptions, these strategies support a holistic and coherent reading of box plots by linking their various components with the underlying statistical concepts. Teacher training and the strengthening of statistical language emerge as key factors for the effective teaching of box plots. This systematic review contributes to the field of statistics education by providing a theoretical and methodological foundation for the design of future educational interventions that incorporate evidence-based practices, aiming to strengthen statistical literacy across different educational levels.