2 Good Experimental Design Minimises Suffering (And Produces Better Science)
Even aside from the question of using animal subjects, we should want all scientific research to be well conducted and well reported. Good experimental design is all part of doing good science, as is clear reporting of our experimental design and the results delivered by the experiment. What you learn here in truth applies across all scientific investigation, but these considerations are heightened when animal subjects are involved, because of the ethical element of the research practice.
Experimental design and statistics work together to deliver sound, meaningful research that moves our collective understanding forward. Considering both of these before conducting research will improve your science, and is arguably essential for good science.
2.1 Experimental Design
Once a research hypothesis has been devised, Experimental Design is the process by which the practical means of answering that question is constructed. The design should aim to exclude extraneous or confounding influences on the experiment such that the causal factors are isolated and measurable, and any difference in outcome as a result of changing those factors (the โsignalsโ) can also be measured cleanly.
2.2 Statistics
All measurement is associated with irreducible error (i.e. โnoiseโ). Experimental subjects come with their own inherent variability, and even the best experimental design often fails to exclude all possible external causes of variation.
The kind of generating process (e.g. count data vs averages of length or concentrations) affects how measured values can behave, and the way we use small groups to represent larger populations also influences our measurements.
Statistics is the branch of applied science that allows us to make probabilistic inferences about our certainty in the โsignalโ - measurements, comparisons and experimental outcomes - even in the face of natural variations in processes and โnoise,โ and the way we choose small groups to represent populations.
In general, you should design your experiment specifically for:
- your population or subject group (e.g. sex, age, prior history, etc.)
- your intervention (e.g. drug treatment)
- your contrast or comparison between groups (e.g. lung capacity, drug concentration, etc.)
- your outcome (i.e. is there a measurable or clinically relevant effect)
Can you rely on examples from the published literature, or advice from experienced scientists, instead of understanding experimental design and statistics for yourself?
In other words, does all published science exemplify good practice in experimental design and statistics?