A Sample of 3,000
To experience firsthand a simulation of randomly sampling 3,000 participants from a population. Imagine that for the US population, people spent on average 15 hours a week on social media. What's more, 68% of the population spent 12 to 18 hours a week on social media. How well could a sample of 3,000 participants represent the entire population?
Directions
Below is the Begin button, a histogram of the population (in terms of their Social Media Usage), and a histogram of the sample. Once you click begin, participants (represented by a yellow dot) will be randomly selected one at a time from the population. After each participant is selected, the bottom histogram will be updated. The simulation stops when all 3,000 participants have been randomly selected. At that point, compare the sample mean with the population mean. As a representative sample, we would expect the sample and population mean to be very similar. Any difference between them would be considered an error, and referred to as sampling error. You may restart the simulation by clicking the "Begin" button.
Controls
Starts a fresh run, clearing previous dots and the sample histogram. Keyboard: Space also begins.
Population
Population distribution: mean = 15 hours/week, standard deviation = 3. X-axis ticks every 3 from 0 to 27. Y-axis is proportion up to 0.25 with ticks every 0.05.
Sample
Sample histogram updates after each draw (bin size = 1 hour). Y-axis is proportion and rescales to 0.05 above the tallest bin (rounded up to the next 0.05), starting at 1.00 for the first draw. X-axis ticks every 3 from 0 to 27. Bars are blue with thin black borders; a black vertical line marks the sample mean.
Sample Mean Updates
Sample size | Sample mean |
---|