How is world population calculated? And how accurate is the number? Is it possible that the number can be entirely wrong?
In: Other
Population estimates are “guesses” with error factored in. Let’s say you had to estimate how many hairs were on your body. This is a seemingly impossible task…if you would like to be exact. Hairs are being added and removed by the second so to approach this problem we guess really a range of numbers.
To calculate that range you might pick a fixed unit (e.g. hairs per square inch of your body). You wake up one morning and you count the hairs on randomly selected square inches of your body. You do this several times under different conditions (e.g. time of day or area if your body).
You may find that some of the factors greatly impact the density of hair. These factors need to be collected too (e.g. proximity to your head). The factors can be used to weight you’re overall calculation and predict counts that are missing or you’re less sure about.
You can calculate how much your guesses vary under similar conditions. Let’s say one day you count 75 hairs/square-inch then another day 120 hairs/square-inch. You assume this is due to randomnesses. Now you can calculate an overall average and a range around that average (based on how much your sample estimates varied).
Now given you have an average estimate for your unit of analysis (hairs/square-inch), you meticulously calculate the number of square inches on your body (also an estimate with error). These are aggregated to determine the total number of hairs. However, this isn’t a single number, it’s a range based on how much your numbers varied when you collected data. We tend to communicate it as a single number because it’s easier to interpret.
Initial data collection is really important because if you aren’t identifying the factors that affect your estimates (let’s say you didn’t take different parts of the body into account) or aren’t appropriately sampling in a way that is representative of the population (if you only took measurements for your head as opposed to the different areas of the body)- your overall estimate will be flawed.
I’ve worked with census data (ACS data) and the measurement accuracy is questionable. The amount of personnel, expertise, and money needed is substantial, yet it’s clear the government doesn’t invest as they should.
This is extremely important because: 1) informs political polls and predictions (which have been subpar in terms of prediction); 2) Prediction of crime or domestic terrorism; 3) pretty much any prediction made about people on a macro-level.
I challenge all of the young-bloodz about to head of to college to look into measurement and statistics. We need more passionate minds in the field.
Latest Answers