Beginning statistics education at early grades as well as focusing on statistical reasoning rather than procedures and computation has been a main movement in statistics education over the last several decades, as variability, distribution, and samples and sampling concepts that are essential for statistical reasoning are considered as difficult and complex ideas that require a considerable amount of time before students develop a deep understanding of them (Ben-Zvi & Garfield, 2004; Watson, 2006). Garfield (2002) proposed concrete activities involving sampling from finite populations in order to students to develop statistical reasoning about samples and sampling.

This study examines fourth graders engaged in three concrete activities involving sampling from finite populations. The first included a survey of popular foods for school meals. The second had them take samples from a box containing white and black marbles to predict how many white and black marbles were in the box. The final activity required them to predict how many times any certain Korean letter would appear in a Korean story book. The results show that the participants can experience and notice different ideas related to samples and sampling in different activities. In the first activity, they acknowledged that samples are useful for obtaining the information about populations. A population survey is difficult and is not overly useful. In the second activity, they recognized that samples cannot be identical to their population but that the information from a group of samples is similar to the information of the population. In the last activity, they devised some ideas about random sampling even though the ideas were immature.

Keywords: Samples and Sampling; fourth graders; concrete activities

Biography: I am a Ph.D. student studying at Seoul National University.