Activity: Gators

source : Prieto-Marquez, Gignac, and Joshi (2007)

–>

Paleontologists are interested in identifying dinosaur sex through their fossil remains. In an effort to explore this, Prieto-Marquez, Gignac, and Joshi (2007) compared various traits believed to differentiate sexes in alligators (Alligator mississippiensis), which are close relatives to dinosaurs. One of the characters was pelvic canal width, which they suspect differed between alligator sexes, but may also differ with body length (a common metric used for alligator body length is the length from the alligator snout to its vent).

Objective: assess if pelvic canal widths differ between male and female alligators, while controlling for the alligator’s snout vent length. Use these results to discuss whether pelvic canal width, or similar diagnostic measures, may be useful for determining alligator sex, and whether similar diagnostics may apply for dinosaurs.

Steps:

i. set up your script!

  1. import the data from https://raw.githubusercontent.com/laurenkolinger/MES503data/main/week12/gator.csv

  1. look at the first few rows of the data (head) and describe it

“in this dataset, each row represents…”


  1. Answer some basic questions about the ANCOVA, referring to the above context.

What are the independent, dependent, and covariate variables?

What are the statistical null and alternative hypotheses for the main effects of the ANCOVA ?


  1. Make a boxplot of the main effects of the data

does it look like there is a difference in PelvicCanalWidth between Sexes? which sex has greater PelvicCanalWidth?


  1. Make a scatterplot of the data with the covariate and regression line.

does it look like there is an effect of SnoutVentLength on PelvicCanalWidth? Describe the trend for either sex, if any.

Does it look like the covariate (snout vent length) affects each of the levels of the main effect the same? NOTE: remember, this is one of our assumptions of the ANCOVA


  1. Test assumptions

Assumption 1: Independent/Random Samples

Assumption 2: Equal Variance

Assumption 3: Normal Distribution

Assumption 4: Homogeneity of Regression slopes


  1. Explore model fits.

run a regular one-way ANOVA, and extract the R^2 and RSE values

run the ANCOVA model, and extract the R^2 and RSE values

Which model is a better fit of the data? Describe why and report specific variables to back up your claim.


  1. Look at the results and discuss.

answer the following questions:

Prieto-Marquez, Albert, Paul M. Gignac, and Shantanu Joshi. 2007. “Neontological Evaluation of Pelvic Skeletal Attributes Purported to Reflect Sex in Extinct Non-Avian Archosaurs.” Journal of Vertebrate Paleontology 27 (3): 603–9. http://www.jstor.org/stable/30126361.