Lecture 02-cwo

# Lecture 02-cwo - Roadmap • Experimental Design •...

This preview shows pages 1–10. Sign up to view the full content.

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Roadmap : • Experimental Design? • Replication • Random assignment • Controls • Applied Example: • Marine reserves • Statistics Definitions: • Experiment: A study in which the investigator manipulates the factor of interest (independent variable), while keeping other factors “constant”, and observes the effect of the variable of interest (dependent variable). • Only method for demonstrating causation Definitions: • Statistical population: all possible units that we could study. • Sample: the subset of the population from which we actually take data. • Replicate: A sampling unit. Challenge : • Four groups (e.g., treatments), from which we’ll draw samples • They may or may not differ in their true mean • The first person to tell me which group(s) has the largest mean and which has the smallest can have \$5. It costs \$1 to play. • If you can tell me the true mean for each Green Blue Yellow Red 10 20 30 What do you think now? Non-random Why was this hard? Why do we replicate? 22.5 21.5 13.0 22.5 True Means Key Elements : • Replication: having >1 study unit. • Random sampling: units that are sampled are representative of the population. • Random assignment: units are assigned at random to treatments. • Control: A treatment that provides a Why? • Replication: Nature is variable • Random sampling: Results apply to the population • Random assignment: Treatment effects should be unbiased Independence & Pseudoreplicatio n • True replicates must be “independent” (no shared error). • “Pseudoreplication”: lack of independence among replicates • Examples:- heights of students at UF vs. FSU- effects of temperature (incubators) Random assignme nt • Take all units and randomly allocate them to treatments....
View Full Document

## This note was uploaded on 07/13/2011 for the course PCB 4034C taught by Professor Osenberg during the Spring '11 term at University of Florida.

### Page1 / 44

Lecture 02-cwo - Roadmap • Experimental Design •...

This preview shows document pages 1 - 10. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online