 Hi, I'm Lucy and in this video we're going to look at sampling and bias in statistics. Statistics is its own branch of maths that deals with data. It deals with the collection, organisation, presentation, analysis and interpretation of data. Data comes up everywhere. You will even see statistics when reading the newspaper. So it's very important to understand what statistics is and understand the limitations of it and how to best use all of the different tools that are available to us. In our statistics videos we will cover these different things. In this video we're going to look at sampling and bias. Every organisation collects data on their consumers. For example, Google and Facebook are constantly collecting huge amounts of your personal data. They do this so that they can carefully select which advert you'll be most interested in and then they only show these to you. Without data they wouldn't be able to do this. Data collection is usually done by sampling a population. A sample is taken because usually the population is far too large to collect data from everyone. It is too time consuming and expensive. In statistics a population might not be human. It consists of everything or everyone that's being studied. The population can be people, animals, items, events or pretty much anything really. Taking a sample can be tricky though. The sample has to represent the whole population and can't be biased. Choosing a sample that is as representative as possible is a skill. Before elections polls are taken to try and predict which way the election is going to go. If the poll only phones up or knocks on people's doors during the working day to collect that sample it won't be a fair representation. People that are home during the day are not representative of the whole population. They're just parts of the population. There are different methods for collecting a sample to try and remove as many biases as possible. In random sampling each member of the population has an equal chance of being chosen. Often random number generators are used for this. This is the preferred way of sampling but it's often difficult to do as it requires a complete list of every member of the population. Systematic sampling is easier than random sampling. Everyone's given a number then a random number generator is used to choose where to start which is to here and then take for example every third person after that. Convenient sampling is the easiest to do but it's the least representative. You simply survey the people you bump into. As I'm sure you can imagine this has lots of bias and isn't representative. Cluster sampling divides the population into groups. Then the clusters are randomly selected and all elements within that cluster are sampled. Like if we'd split the USA into time zones and then randomly selected these two clusters. Stratified sampling divides the population into groups called strata. A random sample is then taken from each of these strata. Whichever method of sampling you opt for there will always be sampling errors because however much we try a sample just isn't as good as a whole population. There are ways of estimating how close our statistics are to the true population. You may have come across this in future if you stick with maths or if you use statistics either in your degree or future working life. For now you just need to be aware that there are different methods to help us choose a random sample and that you need to carefully design your study to remove as much bias as possible. Thank you.