Type a new keyword(s) and press Enter to search

PItfalls of Data Analysis

 

            
             The paper entitled "Pitfalls of Data Analysis" written by Clay Helberg M. discusses some common pitfalls of the use of statistics and offers ways to ensure that statistics are clear and accurate (Helberg 1). In the introduction Helberg ponders the reasoning behind the public's conclusions regarding the lack of reliable statistical information. The paper then goes on to consider three broad classes of statistical pitfalls.
             1. Sources of bias.
             2. Errors in methodology.
             3. Interpretation of results (Helberg 1).
             The first pitfalls discussed are sources of bias which involve the conditions/circumstances that affect the external validity of the results. The sources of bias included in the paper are representative sampling and statistical assumptions. Representative sampling is abused when the target population of the study is not a true representative sample (Helberg 2). This can escalate into a big problem when studying people because it is so difficult to have access to the entire population. The researcher can then only hope that the group picked isn't different somehow from the target population. .
             The next example of bias illustrated by Helberg is the use of statistical assumptions. The author shows how to avoid problems with assumptions such as ignoring an assumption and linked observations which can create bias in the study. .
             It is suggested that in order to avoid these pitfalls the statistician needs to understand the assumptions of the statistical procedures and be sure they are satisfied. .
             The next abuses of statistics covered in the paper are errors in methodology which can lead to inaccurate or invalid results. Helberg states that improper use of statistical power leads to errors. Power refers to the ability of the statistical test to detect true differences of a particular size (Helberg 4). If there is too little power the effect the experiment is trying to uncover can be overlooked. This can be caused by using a very small sample size or by creating a test that is not sensitive enough to find differences.


Essays Related to PItfalls of Data Analysis