Analyzing Images to learn Mathematics and Statistics
Dendroclimatologists can reconstruct climate records far further into the past than our written records extend, by examining tree rings. Students are asked to "reverse-engineer" this process by considering growth rates of a tree species in response to various climatic factors.
In this module, students analyze images of tree rings already collected from the field - eastern hemlock trees from Northeastern US forests. Then, students will be asked to think about the relationship between annual growth in trees and different aspects of climate. They will make one or more novel hypotheses and test those hypotheses with the data they generate from the tree ring analysis, and publicly available long-term climate records. The data and hypotheses most students will generate will be amenable to analysis by linear regression. The image analysis part of the module can easily be supplemented with local fieldwork, where possible, enriching the experience and offering really interesting opportunities to talk about data sources, sharing, and reuse.
Potential Learning Objectives:
Be able to conduct and interpret simple linear regression analyses, including interpretation of p-values, R2, and the equation for the line of best fit.
Be able to implement and understand the necessity of standardizing data by a covariate.
Gain appreciation for the utility of image analysis in biology, and gain some experience with simple data collection from images (linear measurements).
Be able to describe the basic procedures and purposes of dendrochronology, including how cores are obtained and prepared for observation, what measurements are routinely taken, how data are treated, and identify some scientific questions that can be addressed.
Be able to manipulate and re-organize messy data into a format that is amenable for analyses.
Discuss the impact of measurement error on scientific studies and analyses.