Week 6
Day 1
In Class
- Lecture: Stratified Random Sampling
- Lecture Code: Stratified.R
Assigned
Please bring a laptop next class
Due 1 week: Canvas Quiz 3
Day 2
In Class
- Please bring a laptop to class
Exercise: Optimizing Sampling Allocation within Strata
A colleague asks you for help in planning their next study. They are interested in estimating the total number of a fish species throughout a large region (assume detection probability is 1). They can survey a total of 100 stream segments. However, they know there is spatial variation in the numbers of this fish and want to use stratification to help minimize the variance. To help your colleague, you ask for some data and your colleague offers up count / population size data of a related species.
Create a RMarkdown script/HTML to optimally allocate \(n = 100\) fixed samples among strata; you have to decide on the strata!
Estimate the allocation of samples per strata and the proportion of samples per strata.
Investigate whether the proportional optimal allocation of samples per strata changes if you increase or decrease the total number of samples.
The optimal n per strata (for population mean or total population)
\[ n_h = \frac{nN_h\sigma_h}{\sum_{k=1}^L N_k \sigma_k} \]
Assigned
- Due Next Class: Read Thompson Chapter 12
- Due in 1 Week: Submit your RMarkdown script and html from the above inclass exercise on Canvas.
- Due in 1 Week: Chapter 11, exercise 1.