Experimental Design and Research Flashcards
What is a control?
A control is the part of the experiment that does not receive any treatment or which is the normal practice.
It is a standard so that variations between treatments can be compared.
Eg. No fertiliser, normal feed
What are the main elements of experimental design?
- Control
- Randomisation
- Replication
- Standardisation
What is randomisation?
Randomisation is using a method of selection to ensure that all members of a population have an equal chance of being involved in a particular treatment.
This ensures that no bias is included during sampling and treatment allocation.
Eg. Place chickens in alternate boxes with eyes shut.
What is replication?
Replication attempts to overcome problems associated with the occurrence of natural variation in living things.
It increases the reliability of the results. You could also repeat the experiment a number of times.
Eg. Cost, land space, housing, feeding
What is standardisation?
Standardisation is where the experimenter tries to make all conditions in an experiment as equal in each treatment as possible.
Eg. Plant experiment - light, temperature, water, fertiliser
Animal experiment - feed, environment, water, management
Calculate and explain the mean.
- the average value
Calculate and explain the standard deviation.
- square root of the variance (measure of how closely values clutter around the mean)
Calculate and explain the standard error.
- ‘Significance’ of a result is the probability at a chosen level of significance that the result will hold true.
- 5% is the normal significance level chosen for agricultural experiments so the results should be obtainable 95% of the time.
- If the difference between arithmetic/numerical means is greater than twice the standard error, the results are significantly different.
Eg. Record keeping Daily weights Deaths Feed eaten Cost of feed Final weights Dressing percentage
Tables
- record data - easily understood
- sort data - trends
- mathematical transformations
- title, column headings, units
Measures of spread or dispersion: range
Range is the difference between the highest and lowest of the variables.
Measures of spread or dispersion: variance
Variance is the measure of how closely values cluster around the mean.
Measure of spread or dispersion: coefficient variation
Coefficient variation is regarded as an absolute measure of dispersion. The smaller the coefficient variation, the greater the reliance that can be placed on the result.
Explain normal distribution.
As a sample from a population becomes larger, agreement between sample values and population values increases and so the sample mean approaches the value of population mean.
- Control (x)
- Standardisation (x)
- Randomisation (x)
- Replication :)
Curve is:
- symmetric
- asymptotic to horizontal axis
MEAN, MEDIAN, MODE are coincidental.
Describe methods of agriculture record-keeping for broilers.
- Weekly weight
- Deaths
- Previous trial records
- Feed amounts and costs
- Temperature
What are the measures of performance?
- Gross margins
- Yield
- Profitability
- Daily weight gain
- FCR (Food Conversion Ratio)
- Live weight vs. dressed weight
- Mortality rate
- Least cost feed