You were asked: Write down two or three factors that telephone ownership indicates which might link it to cancer of the colon.

High telephone ownership in a country indicates a western lifestyle. We can call it a proxy for such a lifestyle. One feature of a western lifestyle is a highly refined, meat-based diet. It is probably this diet that is a direct cause of cancer of the colon. In other words, telephone ownership and cancer of the colon are both linked to the western lifestyle. This seems a more sensible way of explaining the correlation than trying to suggest that owning a telephone directly causes cancer of the colon?

Defintion: Proxy
This is a relatively simple substitute for something that is too complex to measure in its entirety (e.g. lifestyle). The point of the telephone-cancer example is that statistical correlations by themselves prove nothing. We need other evidence to establish fully whether poverty causes disease, or whether it is the other way round. Are the two linked to a third factor? Perhaps the correlation is purely chance.

So where does this leave us with our poverty-disease correlation?

My own opinion is that poverty is a basic cause of disease. Malnutrition, lack of education, lack of clean water and lack of access to health facilities are all also positively correlated with disease incidence. It is because these indicators are themselves so strongly linked to poverty that many people call them proxy indicators.

Study skills review: reading numbers in tables and graphs

Numbers by themselves are meaningless. To turn them into useful information they have to be compared with other numbers, to make out patterns and to draw links between numbers that represent different things.

That is what you have been doing for most of this taster.
Sometimes the examination has been fairly superficial, as in the earlier section when you drew a link between education and disease by simply examining a few countries at the top and bottom of the adult illiteracy and under-five mortality ranges and concluding that if one is high, the other is likely to be high too.

Looking at high and low numbers in the columns of a table is a good start to reading them. It gives a feel for the data and suggests the patterns that probably exist. Something similar happens when you listen to a piece of music for the first time. You might know you like it, but you will have to listen to it several times before you discern properly its melody and words. But it is also important to note the numbers that don't easily fit into the broad pattern. We are dealing with the real world where nothing fits perfectly and these exceptions can often tell us a great deal.

In Activity nine you went beyond the superficial examination and, by drawing a graph, established in some detail the pattern that exists between poverty and disease. There are many kinds of graph and chart (the one you drew is called a scatter diagram) but they all share in common the aim of making obvious the patterns that exist in tables of numbers. Whenever you decide to draw a graph, ask yourself first whether or not it will help to illustrate the patterns. If not, don't bother.

Tables and graphs, in the end, are no more than models of reality, where the real world is represented in number form and re-jigged so that the numbers supply useful information that can further our understanding.

In section six, we'll consider ways that health standards can be improved.

  < previous   Page 4 of 4

Comments