I have not found an accepted way to write asymmetrical confidence intervals. Two possibilities are rendered in the image below. I personally prefer the second one, which uses the standard ± sign already in use for symmetrical confidence intervals, simply splitting the confidence interval boundaries into two stacked numbers.
Another possibility is to write directly the upper and lower limits:
Gapminder is a wonderful, public tool that lets you visualize current data and historical trends in many demographic, economic and social indicators. Do you want to know how life expectancy in the USA has developed compared to, say, Cuba and Qatar? Just go here and click Play (then come back here, please). Three rather different development paths toward the same life expectancy. Fantastic.
The image to the right is the first thing you see on gapminder. It shows
the relationship between income (horizontal axis) and life expectancy
(vertical axis) in countries throughout the world. Each circle is a
country and its size is proportional to the country’s inhabitants (the
big red circle is China, the big blue one India).
There is clearly a positive relationship.
In a TED lecture, Gapminder founder Hans Rosling uses this image to argue that focusing on economic growth is the best thing to do to improve life expectancy in developing countries (the low-income, short-life blue dots in the image represent Africa). I think this is true, but the statement should be qualified. The catch is that the horizontal dimension does not show raw income, but orders of magnitude of income – notice how the distance between, say, $200 and $400 is the same as between $2000 and $4000. This is called a logarithmic scale and is used for better display of certain patterns.
If you display income dollar-for-dollar, as in the image on the right, the pattern looks a bit different. Now it shows clearly that a very small income, by Western standards, can be associated with a rather long life: people in big red China achieve a life expectancy of 73 years with under $5000 per year. And so do many other countries, like Vietnam (where under $2500 per year buy you 74 years of life) or Syria (74 years for $2200). Beyond this point – about 75 years and $5000 per year – additional income buys relatively little, as the fat yellow dot (the USA) says: 78 years for $42000 per year.
Thus there seems to be a lot to gain from economic development in very poor countries (Africa is blue), and relatively little to gain from further increase in income in moderately rich to very rich countries.