In: Statistics and Probability
You are a wine importer for Singapore and are seeking new sources of wine to offer more diversity to the local wine market. You are considering Greece because its Assyrtiko grape, grown mainly on the island of Santorini, is the current darling among informed Western wine consumers. You are considering between two suppliers of wine from Greece (Supplier A and Supplier B). You conducted a taste test of wine from the two suppliers in Singapore. Respondents had to taste the wine and report a score, with 1 being poor taste and 10 being excellent taste. The data collected shows the distribution of the scores to be normal. The standard deviation of score of wine from Supplier A is 2.0 and that of Supplier B is 3.5. 25% of respondents who sampled Supplier A’s wine reported a score above 8.0. 20% of the respondents who sampled Supplier B’s wine reported a score below 5.5.
a. If you were choosing a wine supplier for the local market based on measures of central tendencies/location, which wine supplier would you choose?
b. Comment generally on the measures of dispersion of the respondents’ perception of wine from Greece (in relation to the 10-point scale)? (Note: There is no need for calculations.)
a) A biased approach which people usually take in this case is that Suppliers A's wine where 25% reported a score above 8.0 will be the ideal choice, but the overall spread of the rating is low as its standard deviation is less compared to the supplier B, hence the rating will have a strong bias in deciding the supplier, so it is important to see the standard deviation and compare,so based on facts and data we can conclude that Supplier B's wine should be chosen instead of A because og higher standard deviation of the scores
b) Dispersion of A and B have a drastic difference as Dispersion of A will be highly skewed than that of B, because of the 25% proportion of people will have a rating of 8 and the the standard deviation is just 2 meaning that within 68% of data Of Supplier A lies between 2 score points(as mentioned in the question) and the case id different for the spread of B as the standard deviation is varied and have a good chance of performing than A because of less bias and better estimate