Everyone has to face information that they would rather not see sometimes. It might be a speeding ticket or their weight. It might be the result of a survey that gave results that they are not comfortable with. One course of action is to ignore it but that is not always the right or the best action. Sometimes the unpleasant information is a critical warning that must be heeded.
More importantly, ignoring facts can damage credibility. In Court, the requirement is to give "the truth, the whole truth and nothing but the truth". This is also the rule in science where all the results of an experiment must be presented.
When analysing data, these problems arise in the form of values that were not expected or do not fit well with your model of how things should be. The statistician calls these outliers. Can those data values be ignored as "obviously in error"? Or are they actually warnings that contain important information?
The Depression of 1929-30 is an example of economic time series data. It did not fit the pattern of any previous economic data and clearly conflicts with many of the mathematical models currently used by econometricians. But the Depression was real and perhaps taught us more about the world economy than any other event of the last century.