Introduction to Randomization Design
Previously in the course, we have referenced how experimental design drives the statistical model to be fitted. Recall in Lesson 5, we discussed the two components of the experimental design that accounts for two aspects of a study.
The treatment design component, which was addressed in Lessons 5 and 6, describes the treatment levels of interest, treatment type (fixed vs. random), and also the relationship of treatments with each other (crossed vs. nested).
The randomization design component takes into account the treatment design aspects and also the physical layout of the study setting including other influencing factors such as confounding (or blocking) variables.
In our discussions of treatment designs, we looked at experimental data in which there were multiple observations made at the treatment applications. We referred to these loosely as replicates. In this lesson, we will work formally with these multiple observations and how they are to be collected. This brings us to the right-hand side of the schematic diagram portraying the randomization design component;
How many factors are there?
How many levels of each factor are there?
If there is more than 1 factor, how are they related?
Crossed: each level of factors occurs with all levels of other factors. (Factorial)
Nested: levels of a factor are unique to different levels of another factor.
Are factors fixed or random effects?
Are there any continuous covariates? (ANCOVA)
What is the experimental unit?
An experimental unit is defined as that which receives a treatment, (e.g., plant, person, plot of ground, petri dish, etc.).
Is there more than one experimental unit? (Split Plot)
How are treatment levels assigned to experimental units?
Completely at random? CRD
Restriction or randomization?
two-dimension: Latin Square
Are there sample units within experimental units? How many true replications are there?
Are there repeated measurements made on experimental units?
As can be seen in the diagram above, the treatment design addresses specific characteristics of the experimental factors under study. The randomization design addresses how the treatments are assigned to experimental units. Overall, the experimental design sets the stage in collecting data systematically and also dictates the statistical model to be used and the ANOVA-related calculations.
- Understand the importance of randomization design, the second component of experimental design and how it impacts on our interpretation of results.
- Identify any blocking factors and the randomization design used in a study.
- Use Statistical software to obtain the randomization design that assigns the treatment levels to the experimental units schematically.
- Gain experience in utilizing statistical software to analyze data obtained from a given experimental design.