# How to do your research project

## Analysing data

• explore using statistical tests to analyse data

At this stage of the research process, you should have organised, described and displayed your data. You can now use your data to provide evidence to support your research hypothesis - to answer your research question.

Inferential statistics are used to analyse data, and you can use these statistics to find out...

• if your sample is representative of a population
• if there are differences between groups
• if there are changes over time
• if there is a relationship between variables

Things you should consider could include...

• how your data is distributed - is it normally distributed or not? (If you don't know what a normal distribution looks like, find out)
• is your data discrete (e.g. gender or choice of transport), or continuous (e.g. weight or height)
• are the values from your sample in line with the values for the wider population?
• are there any outliers or anomalous results?
• look for associations - expected or unexpected
• does the data support your hypothesis, or will you need to reject your hypothesis?

The tests you use to analyse your data will be constrained by the type of data you have and by your knowledge of statistics. There are however, some relatively simple tests that you'll probably be expected to use. These include the confidence interval, standard error, t-tests and chi-squared tests.

You have probably guessed that statistics is a very complex area - there's a good reason why most statistics text books are at least 500 pages long. However, don't panic: you'll not be expected to do anything too complicated (unless you're a maths or statistics student). You simply need to show that you've made sense of your results and tried to use your data to answer your research question.

##### Activity: descriptive statistics: the basics - 10 minutes

If you're a statistics novice, this activity will remind you of some of the common statistical tests you can use to analyse your data. It's your job to find out which you'll need to use and how to use them.

The tests mentioned in the activity are simply some of the most commonly used, there are many more. You may be aware of the level of statistical analysis you'll be expected to carry out in your dissertation, but, if in doubt, ask your supervisor.

You may find Engage in Research: Step-by-step statistics helpful when completing this activity.

1 .Some of the most common statistical tests used to analyse data are...

• confidence interval
• standard error
• one-sample t-test
• unpaired t-test
• paired t-test
• chi-squared test

2. For each of these tests, find out...

• what types of data sets they can be used to analyse (for example, is the data normally distributed? Is the data discrete or continuous?)
• what they are used for
• how you carry out the test

3. Consider which, if any, of these tests you will use to analyse your data. Remember, if you are in any doubt, discuss it with your supervisor; he or she will be pleased you've made the effort to try to find out for yourself about data analysis.