����ý

Just because information is displayed in a graph doesn't mean that it is robust.

The websites Dog pals and CatMates both display graphs about whether dogs or cats make better pets.

When you look at any graph it is important to ask yourself three questions:

• is this graph detailed?
• is this graph accurate?
• is this graph able to withstand close examination?

Looking carefully at these bar graphs we realise they don't tell us much at all. They show something going upward but we can't tell anything else:

• there is no title - we can't tell what the graph is about
• there are no axes or labels - we don't know what is being measured
• there is no scale - so we don't know any of the numbers involved or how much they change

By comparison, these pie charts give us more detail, including the sample sizes. Knowing how many people were surveyed by each website means we can decide if it was a big enough group to provide useful information. We can also calculate exactly how many people chose cats and how many chose dogs.

But...

Even though this piece of data looks more useful, we still need to consider the validity of their sources. This means assessing whether or not the websites could be biased. A good starting point is to look at the name of the companies.

In this case it is pretty straightforward: one website seems to be for cat-lovers and the other one for dog-lovers. They will very likely be biased!

Names can give you a clue as to who the source would favour (be biased towards). In many cases the websites people subscribe to will have views/opinions they already favour.

Outcome

Our assessment suggests that the data we found on those two websites is not robust enough. We would need to look elsewhere for more trustworthy information.