Science Educator Professional Development: Big Data and Inquiry Learning

Background

In 2000, approximately 25% of the world’s information was stored in a digital format. Today, more than 98% of all stored information is digital (Cukier & Mayer-Schoenberger, 2013). This recent surge in data accumulation heralds the changes in the collection, organization, storage, and analysis of information. Scientists are now integrated into this new technical environment. In the past, data collection was costly and time-consuming. For this reason, random sampling methods were developed and used to infer something (with a margin of error) about the total population or phenomenon under investigation. This methodology works well in answering many questions, such as what food is the most likely to be eaten at a certain middle school cafeteria. By randomly selecting a group of 75 students from a population of 400 students, one could predict middle-grade food preferences fairly accurately. However, the strength of this statistical method falls apart when researchers ask about the favorite food of African American girls at that school. This random sample is too small to make any meaningful prediction in that regard. However, if a researcher collected data from all 400 students, the question could be answered easily. When researchers are limited by collecting only a percentage of the population (sample), they must know, before data collection begins, exactly what data to collect and what question(s) are to be answered. By having all the data (or at least significantly more data regarding any population or phenomenon) on-hand, there is no longer a need to know the questions beforehand. Thus, the task becomes how to harness and extrapolate meaning from enormous pools of information. Analysis focuses on finding patterns and relationships across data types and sources, rather than carefully curating pristine samples from information-constrained datasets. Our technological revolution has brought about this important paradigm shift.