1. Effective Experimental Design (Quantitative)
LEM4001
Effective Experimental Design (Quantitative)

2. The research cycle
Watch:

3. What do you believe?
See: list-of-predatory-publishers-2016/ and ws/boys-paper-really-cool-bats-accepted- journal

4. Methods of data collection (Revision)
Observation
i. Sample or population? Reduce error (measurable) and bias.
ii. Measure or observe the variables in question for the selected elements.
Experimentation
i. Manipulate the values (or levels) of one or more (independent) variables or treatments.
ii. Observe the corresponding change in the values of one or more (dependent) variables or responses.

5. Why ‘experiment’? (Revision)
To determine the cause(s) of variation.
To find conditions under which the optimal response is achieved.
To compare responses.
To develop a model for predicting responses.

6. Example 1

8. Q. What would you do next?

15. Example 2

16. Example 2
TITLE: Ecological sensitivity of subarctic mountain vegetation to anthropogenic disturbance in the Icelandic highlands.
AIMS:
To evaluate the impact of experimental trampling on subarctic native mountain vegetation using methods stipulated by Cole (1993).
To monitor the efficacy of native turf transplants for re- vegetating hiker trails in subarctic mountain environs.
To define the tolerance of native Icelandic flora to simulated trampling.
REFERENCE:
Cole, D.N. (1993) Experimental trampling of vegetation: standard experimental procedures. Biological Conservation 63,

17. Example 2

18. Example 3

19. Example 4

20. Example 5

21. Reflecting on the Past
Reflect on a research / industrial project you have previously undertaken. For it you may have:
Collected some frequency data.
Collected some data to illustrate relationships with one or more independent variables.
Collected some data to test for differences.
Undertook a simple experiment where you controlled the factors or made a set of observations.
Q. Debate how you could re-work your project to produce a more valid, reliable and thorough experimental design that eliminates error or bias with a peer.

22. Sampling Method and Effort
Devise a rough experimental method and propose a realistic sampling effort for some of the following projects: Conservation / Arboriculture Foraging patterns of bats in open mosaic habitats including quarries and landfill sites. Managing Pseudomonas bleeding canker in Horse Chestnut in London Parks. Ecological impact of deer on floristic field layer species richness of ancient semi-natural woodlands in Northamptonshire. Agriculture An investigation into the relationship between lamb growth rate and milk replacer diet in a lowland flock. Agronomic consequences of traffic induced compaction on wheat establishment and growth in different soil types. The impact of plant growth regulator (PGR) and variable rate fertiliser applications on the growth of oil seed rape. How well does your experimental design lend itself to subsequent quantitative analysis?

23. Designing valid experiments
Finally, for your current idea consider the following :
Define a (more precise) hypothesis for your ‘experiment’. This you may have already done, but there may have been changes!
Outline the likely analyses to be conducted and consider your approach with subsequent statistical analyses in mind. Do you need a ‘control’?
Identify potential sources of variation and bias and try to reduce them.
Estimate the required sample size - how many is ‘enough’? What is your experimental design?
Present and defend your approach to a fellow peer.

24. Designing your thesis: conclusive remarks
Q. Can you now use these techniques for your actual thesis?

25. Sources for further information
‘Ready to Research’ portal: Dytham, C. (2011) Choosing and Using Statistics, 3rd Edition. Wiley-Blackwell. Gordon, M.S. & Bartol, S.M. (eds) (2004) Experimental Approaches to Conservation Biology. Berkeley: University of California Press. Ruxton, G.D. & Colegrave, N. (2011) Experimental Design for the Life Sciences. OUP.

26. Homework (Optional)
Read any short paper, paying particular heed to the methods section and experimental design. Evaluate the experimental design by considering the following prompts:
i. Is the experimental design and subsequent method valid? Could the aim / hypotheses of the study by achieved in an alternative way? How could you improve it?
ii. Is there bias in your paper caused by the experimental design? If so, how could the experimental method / design be revised to eliminate this?