Apple cider is delicious on its own but have you ever wondered how the cider’s amino acid profile is related to the apple species from which it was made? This client’s research was not only fascinating, it was delicious!
She was interested in quantifying which species of apples differed in amino acid composition, both in absolute terms (“how much of the amino acid was present”) as well as in relative terms (“what proportion of amino acids present were of this type”). The client arrived having already plotted the data so she wanted to know if the differences she was observing in her plots were large enough to be called a different.
My associate and I deliberated to determine the proper technique to answer her question. We considered a few models but couldn’t quite identify one that would specifically answer the research question. We decided to use an ANOVA (a statistical technique to test for differences among means) which would account for the average level of each amino acid as well as the average level attributable to each apple cultivar. Importantly for our purposes, the ANOVA technique will estimate how much of the variation in the data is attributable to the interaction between species and amino acid: precisely the question we are trying to answer! We applied this technique for each pair of apples taking care to account for multiple comparisons via a Bonferroni correction. The end result of this procedure was a list of apple pairs with concentration differences and the specific amino acids causing that difference.
Our client was delighted to see that most of the differences she observed in her plots were statistically significant.