P-Value [ˈpiː-ˈvæljuː]: Probability of your data points, if your hypothesis weren’t true. The p-value is the typical measure of choice to test a hypothesis and is widely used in all empirical sciences.
Say you want to test whether the number of Journocode posts someone has read influences their coding skills. You first formulate the hypothesis you want to proove: Readers of Journocode are better at coding. The opposite, that they code just as good as others, is called the null hypothesis. You then go and gather your data. In this case, you might ask 100 people how many Journocode posts they have read and rate their coding skills on a scale from 1 to 5. You might find that, on average, Journocode readers’ coding skills rate one point higher than others’ (YAY!). Finally, you can ask: How likely would these (or even more extreme) results be, if the null hypothesis was true? What is the probability that the higher average was nothing but random? You accept the hypothesis if this probability – called the p-value – is smaller than a set significance level, typically 0.05 or 0.01. In this case, we call the hypothesis significant.[button url=”http://journocode.com/data-journalism-dictionary/” new_tab=”” button_style=”btn-info” button_size=”btn-default”] Back to Dictionary[/button]