In groups of four, students will decide on a topic to present to the class involving the reporting of statistics and work together on a ten minute group oral presentation to be given during the sixth week of the semester. They can use any medium to present--PowerPoint, overheads, or handouts--but should try to have everyone in the group talk for at least two minutes. The groups should present some of the issues involved in writing about statistics in connection to the chosen topic, and should cover some of the following:
In evaluating the research sources, students should consider the kind of the data and numbers the source is using (OR NOT using!). How reliable is the source? Can the study be duplicated? Is there a reasonable sample size so that the study results would apply to a larger group? Which groups might disagree with this source's interpretation of the data, and why? How do those groups use statistics to support their own views?a description of the audience who would be reading the anticipated article and the length of the article
the importance of the statistics for the audience's understanding of the topic
the level of detail and accuracy needed in order to adequately inform a chosen audience
ways of presenting the statistical information clearly to that audience
problems that could arise in the mishandling of statistics
ways in which the statistics could be manipulated to convey information in a biased way
how to deal with the contrasting approaches and statistics used by your research sources (ie: a pro- and an anti-abortion group's use of statistics)
Sample topics:
the reporting of a rise in the number of HIV cases in New Jersey
coverage of an article that implies a link between a rise in the deer tick population and a rise in West Nile Virus based on misleading use of numbers
studies reported in advertising that seem to give false hope to people who might take a certain medication
a government report on a certain policy initiative that purports to increase public health
a study linking breast cancer with a certain cause through "circumstantial" evidence