Research School for Socio-Economic and
Natural Sciences of the Environment
Research School for Socio-Economic and
Natural Sciences of the Environment
Agenda

Design and Analysis of Experiments

Date: 20 May 2019 - 21 May 2019
Location: UFZ, Leipzig, Germany

Performing (ecological) experiments always requires statistical thinking, not only when analyzing experimental data but already when planning and designing an experiment.
Without a proper statistical analysis it is not possible to publish papers in high-ranking journals. Data analysis, on the other hand, critically depends on the experimental design (i.e. the logical structure of the experiment).

Contents

This course aims at understanding the basics of hypothesis testing, the structure of ecological data, as well as various experimental designs and their appropriate statistical analysis. Statistical methods and experimental designs are exemplified by research typical for the UFZ and experimental facilities of the UFZ will serve as examples.

The course will cover the basics of the hypothetico-deductive method in science, requirements of ecological data and the basic principles of designing an experiment according to specific research questions. In addition, we will relate different types, or families of models to each other, thereby clarifying the sometimes confusing terminology.

In the second day, we will define what experiments are (and distinguish them from measurements and preparations) and explain the role they play in testing scientific hypotheses. Using ecological experiments currently conducted at the UFZ, application of the most common experimental designs (ranging from completely randomized to split-plot designs) as well as their appropriate statistical analysis will be introduced. A last issue will refer to the analysis of experimental data that are measured repeatedly (in time or space) on the same experimental units.

The course will answer questions such as: “Which basic principles do I have to consider when designing an experiment?”, “How can I design an experiment involving more than one factor?”, “How can I handle undesired environmental variation, e.g. within my greenhouse chamber”, and help avoid questions like “Why are the data of my laborious experiment useless?”

Didactic Aims

  • Understand the hypothetico-deductive method
  • Understand principles of null hypothesis testing
  • Describe ecological data (e.g. structure and distribution)
  • Avoid mistakes in planning experiments and gathering data (dependence of samples, lack of randomization, pseudo-replication)
  • Understand the characteristics, including advantages and disadvantages, of various experimental designs
  • Understand the consequences of different designs for statistical analysis of experimental data
  • Be mindful of pseudo-replication, in particular when conducting repeated measurements on the same experimental units

Didactic Elements

Interactive lectures, practical exercises, discussions

Target Group

  • Doctoral Researchers
  • Postdocs & Senior Scientists
  • Technical Staff
  • Science Managers

> More information about this course

> More UFZ courses

Participation fees do not apply to cooperation partners, i.e. PhD candidates of SENSE partner WIMEK.
The current conditions of participation and use apply.
Please note: In order to register, a personal account has to be created first.





Search course

Follow this link to search for interesting courses!