Hypothesis Testing and Categorical Data

Statistics education reform efforts emphasize the importance of informal inference in the learning of statistics. Research suggests statistics teachers experience similar difficulties understanding statistical inference concepts as students and how teacher knowledge can impact student learning. This study investigates how teachers reinvented an informal hypothesis test for categorical data through the framework of guided reinvention. We describe how notions of variability help bridge the development from informal to formal understandings of empirical sampling distributions and procedures for constructing statistics and critical values for conducting hypothesis tests. A product of this paper is a hypothetical learning trajectory that statistics educators could utilize as both a framework for research and as an instructional tool to improve the teaching of hypothesis testing. First published May 2015 at Statistics Education Research Journal Archives

Download Full-text

Essential statistics and epidemiology

10.1093/med/9780198745471.003.0022 ◽

2016 ◽

Author(s):

Paul Cleary ◽

Sam Ghebrehewet ◽

David Baxter

Keyword(s):

Hypothesis Testing ◽

Categorical Data ◽

Quantitative Data ◽

Geographical Information Systems ◽

Geographical Information ◽

Central Tendency ◽

Descriptive Epidemiology ◽

Frequency Distributions ◽

Disease Clusters

This chapter provides a grounding in basic statistics, descriptive epidemiology, analytical epidemiology, and hypothesis testing appropriate for health protection practitioners. The analysis of categorical data using frequency distributions, and charts, and the interpretation of epidemic curves is described. The description of quantitative data including central tendency, standard deviation, and interquartile range is concisely explained. The role of geographical information systems and different disease map types is used to demonstrate how disease clusters may be detected. Determining possible association between specific risk factors and outcome is described in the section on analytical epidemiology, using the risk ratio and the odds ratio. The use of these in different study/investigation types is explained. The importance of confounding, matching, and standardization in study design is described. The final part of the chapter covers hypothesis testing to distinguish between real differences and chance variation, and the use of confidence intervals.

Download Full-text