"Throughout my college career and the internships I've completed, if there's one thing that I've learned it's that only good data will give you good results. Naturally, high-quality data is data that has had all errors and inconsistencies removed from it. One way I clean data is to segregate it by certain characteristics so I can then look at similar data grouped together. This is also helpful because it puts the data into smaller groups, which makes it easier to spot errors. I also tend to spend time analyzing the statistics of each data column, and many times I create scripts that automatically cleanse common errors from my dataset. I know how vital it is that the data be free of errors, and before I analyze any data I always cleanse it first."