Thanks for giving the context.
If there are 15 women who work in the same place who all have the same cancer diagnosis, you might be tempted to infer that there is a causal link. You might jump to the conclusion that the working conditions were causing the high incidence of cancer, especially if it is higher than the national incidence. But you have to bear in mind that 'chance is lumpy' - clusters like this occur naturally, for no reason. There may be no link whatsoever between the place where these women work and the fact that a number of them happen to have the same cancer diagnosis. In fact, it would be odd for these clusters not to occur - just as a dice is not going to fall an equal number of times on each of its six faces.
Here's my understanding of the term:
Lumpy is the opposite of smooth. Imagine the difference between making a nice smooth sauce and one full of lumps. According to statistician Robert Abelson, who coined the term, chance creates 'lumps': variables, which ought to be randomly distributed, have a tendency to gather together inexplicably.
For example, let's say that the proportion of right-handed to left-handed people in a country's population is one in ten. Logically, you would expect every town of 10,000 inhabitants to have 9,000 right-handed people and 1,000 left-handed people. But in reality, it doesn't work that way. Statistics will show that Town A has only 400 left-handed people, while Town B has nearly 2000. There's no logical reason why this should be the case - it's just how it happens. You might not worry about this. But what if Town A has 400 people with chronic heart disease and Town B has 2000? Then you might start to think that there is a reason, and try to look for causes. Abelson's point was that often there <em>is</em>no cause. Chance creates random 'lumps'.
I am laughing, because I'm aware that maths is not my strong suit, and know that I have made an idiot of myself more than once on these pages by getting basic arithmetic wrong. I checked and double-checked my ones and nines before submitting my answer, and I thank you for the reassurance regarding the lack of errors. Phew.
Just curious, why should the use of the word 'lump' make you presume that the term originated outside your continent?
It's well known that truly random patterns of dots appear to have clumps or clusters... due to some characteristic of human perception.
See for example https://en.wikipedia.org/wiki/Clustering_illusion
Su.Ki.
For the past 28 years, up to my retirement one year ago, I worked as a private occupational health and safety consultant. Occupational diseases was one of the areas of concern due to the risk of overexposure to toxic chemicals inthe workplace. So, my studies included epidemiology which included the study of rates of occurence of adverse health effects and their statistical analysis. I had heard the term cluster, but had never heard the term 'lump', hence I thought it may have been a term used in and peculiar to Europe, Australia and/or the UK.
I'm so relieved to know that you knew that I was just teasing you about your math skills :)
