Descriptive Statistics: What Your Data Looks Like

THE MENTAL MODEL: Descriptive statistics is about summarizing the data you have in your hands. It doesn't try to guess or infer properties about a larger population you haven't measured. Think of it as taking a photo of a crowd (descriptive) versus using that photo to guess the average height of everyone in the entire city (inferential). It describes the sample, period.

HOW IT WORKS: You apply mathematical functions to a dataset to generate summary statistics. These functions measure central tendency (like mean, median), variability or spread (like standard deviation, range), and frequency (like counts, percentages). For example, given a list of API response times, you can calculate the average time (central tendency) and the range from fastest to slowest (variability).

WHEN TO USE IT: Use descriptive statistics as the first step of any data analysis to understand its basic characteristics. It's essential for creating dashboard KPIs like 'average session length,' reporting on sample demographics in a study, or spotting outliers in system logs by looking at max/min values. Even when inferential statistics are the main goal, descriptive statistics are always presented first to ground the analysis.

WHEN NOT TO USE IT: Do not use descriptive statistics alone to make predictions or generalize findings from your sample to a larger population. Calculating the average purchase amount for 100 customers doesn't, by itself, prove what the average will be for all 1,000,000 customers. For that, you need inferential statistics, which uses probability theory to quantify the uncertainty in your generalization.

ONE CANONICAL EXAMPLE: A research paper on a new software training program reports on its 500 participants. A table shows descriptive statistics: the sample size (N=500), the average age (34.2 years), and the proportion of participants from different departments (40% engineering, 60% sales). This table gives readers a clear, concise summary of who was in the study, but it doesn't claim that all employees at the company have this exact profile.