Imagine you're trying to figure out what makes plants grow taller. You think sunlight might be important, but it's hard to measure exactly how much sun each plant is getting. So, you decide to use something easier to measure that you think is related to sunlight - the number of hours of daylight.
Here's how instrumental variables estimation works:
1. First, you collect data on the amount of sunlight and the number of hours of daylight for a bunch of plants.
2. Then, you use a statistical technique to see if there is a relationship between sunlight and plant height. You might look at the data and notice that plants that get more sunlight tend to be taller.
3. But, you know that the amount of sunlight a plant gets might be influenced by other factors, like the location of the plant or the type of soil it's in. So, you need to find a way to separate the effect of sunlight from these other factors.
4. That's where the instrumental variable (IV) comes in. In this example, the IV is the number of hours of daylight. You know that daylight is related to sunlight, but it's not influenced by other factors that might affect plant height.
5. So, you use the number of hours of daylight as an instrument to estimate the effect of sunlight on plant height. You can use a statistical formula to calculate this effect, even though you don't have direct measurements of sunlight.
6. This technique is useful when you want to estimate the effect of something that's difficult to measure directly, like sunlight or the effect of a drug on a patient's health. By using an instrumental variable, you can get a more accurate estimate of the true effect.