Often, anecdotal evidence is dismissed as insignificant. Just because one thing happened to one person does not mean we can generalize it to the general population.
Anecdotal evidence can be unreliable because we often hear it from a distance: secondhand or thirdhand or even more distant than that, the stories get changed around as people draw the wrong conclusions. You heard from a friend who heard from another friend that this happened to their cousin–this is not a good way to draw general conclusions.
Larger data sets with statistical analysis can be so much better in so many aspects. It’s a better way to draw general conclusions about the population. If you want to establish causation, it’s good to have a randomized trial.
But there are also situations where anecdotes can be a lot better than statistical data.
Psychology is the study of the mind, and much of what we know about the mind is anecdote. Anecdotes can be incredibly important to understand people.
Anecdotes can lead to understanding the personal significance of problems. Anecdotes are important because individuals are important.
If I’m trying to change my behavior, I want anecdotes.
If I’m trying to figure out how to help individuals with specific problems, I want anecdotes.
Just because it is more probable that one outcome will occur doesn’t meant the other outcome won’t happen. We often need anecdotes to put the statistics in the right perspective, as our brains don’t deal with statistical data very well.
Anecdotal data is still valid, but it can be more or less useful depending on the situation. So sometimes we need data and sometimes we need anecdotes.
If I’m doing a medical study about the effectiveness of a drug, I need a randomized trial with lots of data. But if I’m trying to figure out how to fight social problems such as racism or addiction or illegal immigration, I need anecdotes to better understand why these things are happening.