Conference of the International Group for the Psychology of Mathematics Education, Khon Kaen, Tayland, 19 - 22 Temmuz 2021, cilt.1, ss.121
The instructional goals in statistics education shift from computational skills to conceptual
understanding of basic statistical ideas (Franklin et al., 2007). Accordingly, it is expected that
people reason with statistical ideas and make sense of statistical information named as
statistical reasoning (Garfield, 2002). Distribution is one of the important statistical ideas to
get as it is an overarching concept that includes the interrelated concepts of shape, center, and
spread. The understanding of how the data is distributed is essential to notice variability
among measures so that it rises to analyze the data by selecting appropriate statistics (e.g.,
mean, median, and mode) and graphical representations.
This paper is the part of design-based research that aims to develop the reasoning about
distribution. In this paper, we examined 14 preservice middle school mathematics teachers’
reasoning about distribution, which is the starting point of the design experiment. The
participants were selected through purposive sampling method. A test of 10 questions
including measures of center, measures of spread, data displays, shape, comparing
distributions, and bivariate distribution was applied to evaluate their existing statistical
knowledge about distribution. Based on the answers given for the test, semi-structured
interviews were conducted to understand their reasoning about distribution in depth. We
analyzed the data from the test descriptively and used thematic analysis for interview
transcripts.
The data obtained showed that the preservice teachers could compare distributions with
unequal sample sizes with being aware of multiplicative reasoning. Furthermore, they tend to
use mean as average regardless of the shape or spread of the distribution. While they could
reason about the shape of the distribution, spread or center, separately, they could not
establish a relationship among them. For bivariate distribution, they could interpret the
relations of variables with each other. The results revealed the need to improve the preservice
teachers’ reasoning about distribution by connecting shape, center and spread.