Abstract:
A batch of 300 carrots of an open pollinated variety was analysed for sensory characteristics such as colour, juiciness, fruitiness, sweet taste, bitter taste, crispness and total texture.
The physical variables weight, length and diameter were measured for each carrot and the shape(cylindricality) was computed from these measurements.
The variation in the sensory and physical variables was studied by a Principal Component Analysis (PCA). About 60% of the total variation within the batch was described by two significant PCA-factors: Factor 1, mainly a "sensory" factor accounted for about 42% of the variation, while about 18% was described by a second, mainly "physical" factor.
An analysis of covariation (Partial Least Squares-regression) revealed one significant factor.
Only 3% of the sensory variation could be predicted by physical variables with weight and diameter contributing the most.
A sample size of n=50 is a minimum to cover the variation in physical and sensory characteristics of this batch of 300 carrots.
Stratification as a method for sampling is also examplified.
|