Numbers are at the heart of most forms of journalism – sports journalism, business-to-business (B2B), data journalism and many others types. You’ll see numbers pop up in press releases and government statements all the time. Public relations firms have become adept at using numbers to mislead journalists and to the hide the truth.
Here, briefly, are the main areas that journalists need to understand.
Percentages:
As journalists we must provide context to a story. One way to do this is to explain things relative to size of a whole. If 35 people were sacked from a university is that a lot of people? What percentage of the university’s workforce is it? Are these 35 staff members 100 per cent (i.e all) of the library workers? If so, is the library closing? These are a few angles you need to check out.
A school may boast that 100 per cent of students passed their exams at grade A-C. Sounds impressive, right? But you need to ask whether all students in the class were entered into the exam.
Averages:
The BBC student finance calculator is a great example of a data journalism project which is discussed in the book. It uses the median earnings in each age group and career group to calculate approximately how much students will pay for their university education.
So what is the median and how does this compare to the mean average? In this example, the median is the earnings figure at which half of people in the group are earning more and half are earning less. The mean average, if you can remember those school maths lessons, is where you add up all the numbers and then divide by the number of numbers.
The average can be seriously affected by values at the extreme ends of the scale. In this example, the average pay for a worker at a hospital may be quite high when you take into account how much hospital consultants and senior managers earn in comparison to nurses and technicians.
Jobs at the bottom of the pile – such as those poorly paid cleaners, porters and catering staff – are often outsourced to private companies and may not even appear in the official figures.
As with all aspects of data journalism it is important to understand who or what is being counted. You must then ask yourself what is being missed out? Sometimes official bodies will deliberately miss out data they are seeking to hide.
Interest rates:
If you have a bank account, credit card, student loan or mortgage chances you’ll probably care a lot about interest rates. The BBC’s student finance calculator includes a range of interest rates based on the Retail Price Index (RPI), a measure of UK inflation.
Loan interest rates are usually expressed as an APR (annual percentage rate) to make interest rates comparable. Adverts for credit cards and loans have to reveal their APR by law, although many people remain confused. Some payday loan companies offer APRs of over 4,000%.
Interest rates get even more confusing when many credit cards have promotional APRs that sound cheap, but may not necessarily apply to the entire balance on the credit card.
We often ignore the impact of what is known as compound interest when taking out loans. This means that any interest accrued also attracts additional interest on top of the interest on the original loan amount. Compound interest can have a massive impact on things like mortgages and other long term loans. It is sometimes known as the annual equivalent rate (AER) and it can be quite complex to calculate.
Many of areas confuse our users, so we need to explain them.
Polls:
You may be aware that seven out of ten cat owners say that their cats prefer a particular brand of pet food. They may be right, but as a journalist you’ll want to check this claim out. You’ll be inundated with press releases based on opinion polls that purport to illustrate what people think about various topics of the day.
We don’t have time to explain all the ways opinion polls can be manipulated to prove a particular point. The most important thing you need to understand is the sample and margin of error. How many people took part in the poll? Were ten people surveyed or perhaps a more reliable sample of 10,000? Also how was the group selected and what questions were asked?
As with all these areas, if you don’t understand what writing about then it’s highly unlikely that your users will. Ask the company responsible for producing the figures to explain it to you clearly. For example respected opinion pollsters, like NOP and YouGov and universities use established research methodologies that they will be happy to explain. However, PR agencies that often conduct polls based on very small samples and with high margins of error, probably won’t welcome you looking too much into the finer details and that, in itself, may provoke a story.
Comments are closed.