Top Page | Upper Page | Contents | About This Site | JAPANESE

Quantity Way of Making Hypothesis

By the development of Information technology, data in computers are enormous. In these days, "Finding facts from data. Then consider the hypothesis (reason) of the fact." is the powerful approach.

This approach is not new. But the power is increasing.

In this approach, "Finding the fact from data" is important part. This part needs the ability of the data-analyst.

Statistics and Multi-Variable Analysis are not enough because these tools are made for the type of data in old days. The data-analyst is needed the ability of Data Mining and Data Literacy.

Of course, if data is not enough, quantity way is not good. So Quality Way of Making Hypothesis is also important to solve problems.

Difference of Reasons to Use the Function of Models between Prediction Analysis and Cause and Effect Analysis

For example, I use the correlation coefficient to study the strength of relationship among variables. I do not use the, to study the validity and robustness of the model. It is the difference of reasons to use functions of models between the prediction analysis and the cause and effect analysis.

It is same that statistical models are not enough to formulate complicated phenomena. But it is useful to get hints to find the cause and effect relationship.

NEXT Making Hypothesis by Correlation