ABSTRACT
Before conducting any data analysis, it is always a good idea to look at the data first. No matter how well you feel you know the data, there is always the possibility that it contains some unexpected characteristic. One very good reason for doing this data examination before any analysis is that the results of looking at the data may cause you to change the dataset. If you look and see an unusual case in the dataset (perhaps it has very high or low values for one or more variables), then you may want to check it out in some way. Perhaps there has been a problem in data entry and the case in question has had its data input incorrectly. For instance, in a study involving children, one would be surprised to find a 99-year-old. However, closer inspection may reveal that the child was in fact 9 years old and the 99 recorded is due to a slip of the finger when inputting the data. Alternatively, you may discover that you have cases in your dataset that should not be there. If you are conducting a study involving children and discover a genuine 99-year-old in the dataset (that is, the 99 is correct and not a data input error), then you
will probably want to remove him or her because they do not belong to the population you want to study.
