Statistically Speaking Ep 3: Descriptive vs. Inferential Statistics

saraboostani

A lot of benchmarks and "boom" stats that we come across report descriptive statistics — averages, usage rates, etc. They describe what was found in a sample, and they're genuinely useful for communicating what a data set shows (we use them all the time!). But descriptive stats summarize the data you collected, not whether those patterns hold more broadly, or whether you'd find the same thing next quarter or in your next campaign.

In order to make any predictions about what you're likely to find next time, you need inferential statistics. Inferential statistics use probability and hypothesis testing on the data you have to determine whether patterns in your sample can be generalized to a broader population. If you're acting on descriptive statistics alone, you're assuming that what happened in one sample will hold for you. Inferential statistics are what protect you from guessing. We dig into it in episode three!

https://6senseinc.wistia.com/medias/3pmngj36nm?embedType=web_component&seo=true&videoWidth=960

Watch the full episode here

@Kerry