UNU-FTP short training course

NOTE

I have added some material at the bottom for the practial biostatistical laboratory.

Lectures

Practical exercises

Biostatistics

Biostatistical excercises - so far session 1, 2, 3 & 5 (for the last part reference will be made to the slides below)
- Lab session 1 Excel tips and tricks
- Lab session 2 Confronting models with data - (solved spreadsheet)
- Lab session 3 GLIM - standardizing catch rate (solved spreadsheet)
- Lab session 4 Introducing the bootstrap concept
- Lab session 5 Simulating fisheries data (“solved spreadsheet”)
  - the slides
Solutions Still to be fully filled out

Note that Lab session 5 is an introduction to simulation of fisheries data. The intent is to take the class step-by-step from an almost empty spreadsheet through to completion. I am still thinking a little what interesting scenarios one could put as a brain-teaser. Although it is a simulation, not a statistical model I think it is a helpful excercise to go through. In the end of the slides that I intend to use during the laboratory excercise there is actually a hint that the simulation could easily be turned into a statistical model. The aim is then to have an excercise in the second week where we would actually estimate the parameters, given some data (working on it).

Note that there is more stuff in the Lab session 5 spreadsheet than needed, including estimates of biomass and total catch in weight. This stuff is there now, but for the classes will only be added in the second week, once we go through the yield per recruit and such stuff.

Stock assessment - practical laboratories

Estimating growth and mortality - all in one go!

The historical methods for estimating growth and mortality from length frequency data is quite cumbersome. Growth rate is normally estimated first and then the mortality is estimated separately. The statistical methods applied are often not quite according to conventional thinking. In some cases no statistic is actually applied, just some numerical algorithm (e.g. VPA). Part of the reason for this is that many of the methods were developed prior to the wide availability of personal computers. Given the wide availability of powerful personal computer more proper statistical methods have become accessable to more users and are now increasingly been used in actual stock assessments (actually in some cases this has resulted in an overkill in the use of statitistics). Here I have attempted to adapt such a thinking in an excel spreadsheet as an educational tool. It is not flawless but should serve the purpose of stimulate thinking :-) The actual statistical model is included in the excel sheet above (Lab session 5), as a separate worksheet called “a model”.