Methods for visual and statistical comparison of the distribution of
codes from content analyses
Jöran Petersson, Judy Sayers & Paul Andrews
The FoNS-project (Foundational Number Sense) in Mathematics education at Stockholm University aims to compare different countries with respect to opportunities for acquiring FoNS in first school year. We do this by analysing textbooks, teachers’ classroom activities and parent engagement with respect to mathematics. This means that the FoNS-project focuses less on the pupils themselves, but more on the mathematical education environment the pupils meet. One part of the FoNS-project is to compare textbooks through content analysis. The generated data are a sequence of codes from each textbook. In the generated data, the authors found that besides comparing frequencies, there is also much to be found in comparing the distribution of the codes for different textbooks. For this purpose we developed a novel method that allows a visual comparison of how the content in each textbook is distributed and we present this method and compare it with other methods found in a literature review. Another challenge is to make statistical inference tests on the generated data. Typically, the generated data set is large and a drawback is that with large data sets even small differences can result in statistically significant but in practice irrelevant differences. The authors will also compare various statistical methods for dealing with this challenge for both the case of
comparing frequency and for comparing distribution of the data.
Source for this document
Petersson, J. (2019). Methods for visual and statistical comparison of the distribution of codes from content analyses. NOFA7 Abstracts, Stockholm University, 13-15 May, 2019 (p. 170). <
https://www.su.se/polopoly_fs/1.437496.1557495063!/menu/standard/file/NOFA7 abstracts 190510.pdf>