I was brought up short by the remark: "We we are finding some interesting correlations, such as a link between divorce rates and cheese eating"
As someone who applied for a job advertised by a nascent market research company in the 1950s, I was interested to learn from the article about Ben Page (The Actuary, November 2015, bit.ly/1WRnq5b) the developments in that industry over the intervening 65 years.
I was brought up short by his remark: "We have so much data now, and we are finding some interesting correlations, such as a link between divorce rates and cheese eating." A statement such as this should not go unquestioned. Although I have no statistical evidence in support, I would say that there is also positive correlation between cheese eating and left-handed people.
Assuming answers to obvious questions such as 'positive or negative correlation?' and 'which divorce rates?' - here I am tempted to repeat the old actuarial chestnut "broken down by age and sex" - I infer that he is telling us that they found that divorced people in general (regardless of factors such as age, sex, time since divorce) eat more cheese than an identically constructed control group of non-divorced people.
This interests me personally because, being recently widowed and living on my own, I do in fact eat far more cheese than I did during the many years of my marriage. So my immediate first question would be: "Is there a similar cæsura in the habits of his group of cheese eaters?"
In the form in which Ben Page states it, the correlation would be of little interest to a cheese purveyor and even less so to anyone else - except, perhaps, a divorce lawyer looking for clients. But, as he tells us that "We have so much data now", what would be of 'interest' to the cheese purveyor would be to know more about the statistical features of his group of cheese eaters and his control group, and even more so if given a range of control groups each exhibiting the various factors mentioned above.
One can envisage various marketing strategies developing from the results - for example, relating to time since divorce:
"Recently divorced? Special offer - Buy two and get one free!"
I was brought up short by his remark: "We have so much data now, and we are finding some interesting correlations, such as a link between divorce rates and cheese eating." A statement such as this should not go unquestioned. Although I have no statistical evidence in support, I would say that there is also positive correlation between cheese eating and left-handed people.
Assuming answers to obvious questions such as 'positive or negative correlation?' and 'which divorce rates?' - here I am tempted to repeat the old actuarial chestnut "broken down by age and sex" - I infer that he is telling us that they found that divorced people in general (regardless of factors such as age, sex, time since divorce) eat more cheese than an identically constructed control group of non-divorced people.
This interests me personally because, being recently widowed and living on my own, I do in fact eat far more cheese than I did during the many years of my marriage. So my immediate first question would be: "Is there a similar cæsura in the habits of his group of cheese eaters?"
In the form in which Ben Page states it, the correlation would be of little interest to a cheese purveyor and even less so to anyone else - except, perhaps, a divorce lawyer looking for clients. But, as he tells us that "We have so much data now", what would be of 'interest' to the cheese purveyor would be to know more about the statistical features of his group of cheese eaters and his control group, and even more so if given a range of control groups each exhibiting the various factors mentioned above.
One can envisage various marketing strategies developing from the results - for example, relating to time since divorce:
"Recently divorced? Special offer - Buy two and get one free!"
Filed in