EXPERIMENTATION CONCEPTS

REPLICATION

1. What is replication?

- This is the repetition of a test or complete so that the variability associated with the test phenomenon can be estimated. - The repetition of all the treatment combinations to be compared in an experiment. - The observation of two or more experimental units under the same conditions.

2. What determines a unit of replication in an experimental trial?

- Levels of variation - Treatment units - Experimental units

3. In practice we find that the number of explicit replicates used is often between 2 and 4. What considerations might result into the need of more replications?

- hidden replications in a factorial treatment structure - Insurance of a long term project that is in a risky environment will require more replicates - Increasing the of e.g. an experiment may be done on 3 different soil types instead of one type to increase its validity. However, this will depend on objectives. - Increasing precision- knowledge of the precision you need and the variation in the experimental material will make it possible to estimate the number of replicates needed which may be more than four.

4. What is hidden replication?

- Extra replications in a factorial treatment structure. - Extra multiplications of treatments in

1. What is blocking as applied to experimental trial?

- Blocking is the arranging of environmental units into homogeneous/similar groups with the aim of increasing the precision of the treatments comparisons by controlling variation among them.

2. Describe the randomised complete block design.

- An experiment where subjects are divided into blocks that are equal in size, the size of the blocks is the same as the size of the treatments and one block is equal to one replication of each treatment before they are assigned to a treatment group at random. 3. What criteria should be used in the choice of the size and number of blocks?

- The number of factors - Past experience of a particular experiment - Familiarity with the materials to be used

PLOTS/EXPERIMENTAL UNITS

1. What factors determine the choices of plot size and shape in field ?

- Available area - Practical considerations such as machines to be used - Treatment factors such as irrigation

RANDOMIZATION

1. What is the importance of experimentation?

- It prevents/removes bias

- A proper randomization has ensures that, in analysis of and on the null hypothesis the treatments have no effects and the squares for the treatments and error have equal expectation. - It removes other sources of extraneous variation which are not controllable. - Randomization together with replication forms the basis of any valid statistical experiment/test.

2. Give examples where practical considerations could well outweigh the statistical requirement for randomisation.

- In siting control plots in an agroforestry experiment, trenches surrounding the plots may be needed to ensure non-interference from trees in neighbouring plots. If the control plots are placed in a corner of the experimental area, then trenches are needed only on two sides of the control plot, rather than on all four sides, thus limiting the effort needed in conducting the experiment.

- Form of treatment does not allow randomization. For example if the treatments were storage times or times for fermenting a particular substance, randomizing the time element may not be possible.

Data management and analysis

1. How does the knowledge of the (CV) assist in experimentation?

- Helps to determine the number of units that are required - Helps to assess the reliability of the experiment