Statistical knowledge helps in the use of appropriate data collection methods, correct analyses and effective presentation of results. Why do researchers need to use statistical tools? One of them is that you will have the raw data right at your fingertips and you won't have to wait until the data mining process is complete before you can begin using it to make sense of it. There are many reasons why you would use a data mining tool in this way. This type of tool allows you to get statistical information directly from the real-time behavior of the user. Lastly, it should be able to withstand different forms of statistical testing and analysis.ĭata mining, on the other hand, has become one of the most popular ways of finding statistically relevant data sets. Third, it should be reliable, particularly if the results come from a large sample. Second, it should be simple and easy to implement. You should use a specific statistical test for your research only if it meets all of the following criteria: First of all, the test should generate accurate estimates of the parameters that you are looking at. In this example, you need to run multiple regression (a typical data analysis) on the price discrimination variable and take care of any lags in the data. Instead, you have to try to fit the model to the real data, i.e., find out what the optimum price would be for a company to discriminate against (in this case, price discrimination). The strengths of a particular method are usually related to the area of study, while the weaknesses generally pertain to the extent of freedom given to the user.įor example, if you want to conduct economic research on the effect of price discrimination, then you should not just use any economic model that comes along because it might not generalize across companies.
All of these statistical methods have strengths and weaknesses.
TYPES OF STATISTICAL TOOLS FOR DATA ANALYSIS SERIES
Some of the popular statistical tools used by researchers are Surveys, Regression, Analysis of Time Series Data and Conditional Correlation Analysis (CCA). The most popular statistical methods for researchersįirst, you need to use a statistical tool for analysis. The use of statistical methods extracts information from research data and provides various ways to assess the robustness of research results. Statistical analysis in research consists of mathematical formulas, models, and techniques used to analyze research data. What is statistical analysis in research? There are two main basic statistical methods in data analysis: descriptive statistics, which summarizes data from a sample using indicators such as mean or standard deviation, and inference statistics, which draws conclusions from data, subject to random variation (e.g., observation errors, sampling variability). It would be completely wrong to "cherry pick" your variables to support your own case without analysis. After all, if you look at a data set and say "This data set really is perfect", wouldn't you say that it is definitely the right thing to do, since it provides so many facts and figures that support your argument? In short, yes - but only up to a certain point. For those who do not work in research, the question, "Why should I use a specific statistical tool for my research?" may seem a bit strange.