With the standardized testing that is now required throughout the United States, educational databases are growing almost daily. Children are being assessed and the data analyzed not only by the federal government and school districts, but also by major universities. Northern Illinois University has launched a website dedicated to the problems in educational research and data mining. In its data management module, three concerns about the problems with data mining stand out as major educational limitations: reliability and validity, statistical significance and analysis of data.
Reliability and validity are the two most important facets of educational statistical analysis. Data must be reliable, which means that results can be duplicated under precise conditions. Additionally, data must be valid, meaning that it actually indicates what is reported in the analysis results. Reliability and validity are being scrutinized more closely in data mining because of the sheer amount of data and the number of researchers who are analyzing that data. With a greater number of researchers, there is an increase in the likelihood of the data being recorded and classified incorrectly.
An additional problem is determining the "statistical significance" of the data being generated. It must be ascertained whether analyzed data is actually being used correctly to diagnose and correct a problem. For example, when dealing with "clinical" testing and data analysis, educators must be very cognizant of using the data correctly so that a child is not mislabeled or the wrong educational plan set up to correct a problem.
One of most problematic educational data limitations is the educator's ability to analyze the different types and amounts of data. Staff members need to be better trained in the analysis and interpretation of collected data. Data integrity can be compromised when researchers incorrectly use interpretation techniques that may have been designed for qualitative analysis, but are instead applied to a quantitative study. More training is needed for educational researchers.