Some sources of error in data collection.

There are many ways data scientists can make error in sampling the population.
Using non random samples will usually not give a good presentation of the population. Taking a survey in a specific neighborhood will not result in a good confidence level dealing with data in the city.
If the investigator is bias, he will try to skew the data.
If the data collected is old, most probably it will not lead us to the right interpretation and therefore a wrong decision.
If the population is huge we need more than one experiment or one survey to analyze the data especially in the public sector.

Leave a Reply