Summary

Making measurements and conducting experiments both require that the investigator make choices about what and how to measure and about what treatments to apply and what controls and replication to use. The development and use of measurements, and the design of ecological experiments, are not routine tasks with single correct solutions.

Four principles of measurement for new concepts are presented:

1. There must be a postulate under investigation so that the purpose of measurement is clear and precision required can be defined.

2. Each new measurement must be specified in a data statement so that the accuracy required can be assessed.

3. More than one measurement of a new concept should be investigated because different measurements may inform in different ways.

4. The variability of a measurement should be investigated explicitly.

Two principal types of experimentation are distinguished; response-level experimentation and analytical experimentation. Response-level experiments are usually designed to investigate the magnitude of an effect when the type of response is already known. Most of the statistical techniques of experimental design and analysis have been developed for response-level experiments.

Analytical experimentation – the focus in this chapter – investigates how ecological systems function. To make an experiment, choices have to be made to investigate particular postulates and develop specific measurements and design appropriate control procedures, treatment controls and replication.

Introduction

At some point research moves from observation and general conjectures to detailed analyses. Then measurements stimulate the development of postulates and determine whether new postulates will be accepted. Scientists devote large amounts of time and money to developing measurement systems.