Description

Description (components of original material, modified material, new material, topic of the material):

The original module uses data to illustrate how pollution affects human health-- in a way that even the most anthropocentric student can identify with. It also introduces hypothesis testing and p-values at a very low-level, so that students who haven't taken a statistics class can complete some basic interpretation.  This provides the student with knowledge of what that term "significant" means in science as opposed to its usage in everyday conversation. The adaptation focuses more specifically on pollution relevant to aquatic systems and drinking water.

The original slide deck was modified to include 1) an explicit description of environmental health, 2) information on (US) environmental pollution and environmental health federal entities, 3) links to online resources on emerging contaminants and PFAS reflecting current knowledge of its health implications.

The dataset was trimmed to focus only on parts A and B; part C on air pollution was removed with the intention of replacing with PFAS exposure data. The instructor was unable to find replacement data and ended up removing part C completely, but future implementations should incorporate at least some water quality data with implications for human health. Possible candidate data could include nitrate in drinking water instead of PFAS (this is a field of active research), or also microbial pathogens or other regulated/unregulated contaminants for which there is a strong body of evidence on mortality or disease incidence.

Context for use (instructional setting, material audience, activity length, other materials or resources):

This module was used in an undergraduate, non-major Water and Society course for a 5- week unit on human water use and human-environment interactions. The slides were presented during lecture and the exercise was assigned as homework rather than as an in-class activity. All materials are provided in the resource content files.

Instructor notes (learning environment, implementation notes):

I delivered the material primarily in lecture format with open discussion, but students would have benefited from structured smaller breakout group discussion to wrestle with the content on emerging contaminants, which many found distressing. Such a discussion could have allowed for more creative exchange and student-led investigation on mitigating exposure, pollution remediation, or environmental justice which are foregrounded in the exercise. 

Students also didn't obtain as deep an understanding of statistical inference as I would have liked and would have benefited from doing the t-test exercise in class, either individually or in small groups. That could be followed up by the discussion of distributions and what overlapping or non-overlapping distributions means, and how the t-statistic and p-value (numbers) are reflective of that visual image. Part B could be modified to include a comparison of data from two states that appear to differ but the difference is NOT statistically significant, to contrast with data from distributions that are. 

Nit picky things: A statistician would argue also that calculating state level mortality statistics by polling state statistics over multiple years is incorrect (pseudoreplication), but for my non-science major student audience, I would argue, this is allowable. One way to revise this activity might be to use county-level data to calculate and compare state statistics.

Finally, instructors using this module’s PFAS information (gathered in Sp2023) should update it as the field of PFAS research is *rapidly* shifting, along with regulation of PFAS in drinking water and science of its bioaccumulation, degradation, and health implications.

*See also the Instructor Notes attached in Extras

Notes

Cite this work