Helping students make sense of the 'real world' data mess

摘要

School science is often very different from "real world" science. One important difference, and possibly the main one, is that in school science the relationships between variables have often been sanitized - essentially "cleaned up" - so that there is very little (and often no) variation in the data from the relationship students are being taught about. This happens because school science often (a) deals with "settled" or "consensus" science where scientists have essentially agreed on what the relationships between the variables are; (b) involves bivariate (i.e., two-variable) relationships where all other variables that might be in a "real world" relationship have been identified and are removed from the example dataset/activity (e.g., minimizing friction so it becomes negligible to understand movement); and, (c) involve known "controlled variables" that can be easily controlled to reduce their influence on the outcomes. This type of removal of variation is problematic as it can make it more difficult for students to learn about, understand, and make sense of "real world" data relationships outside of school settings. It is therefore important for them to engage in school investigations that help them learn to make sense of data with considerable variation in the measured variables. This paper describes and discusses such an activity.