A more complete discussion of the goals and fields of statistics is available in through the Essential Statistics lecture notes
What are the goals of statistics / data analysis
Toolbox to help us finding out
- If observations are solid evidence, or could arise by coincidence
- What parameters or hypotheses are most likely, and what the uncertainty is
As with any complicated tool, you have to know how to use it. If you press the wrong buttons, you’ll get the wrong result.
What are the problems?
Correlation is not causality, validity, reproducibility, researchers degrees of freedom, choice of the appropriate statistics
Branches of statistics / data science
-
Descriptive statistics is the branch of statistical methods that basically display data, but do not directly draw conclusions from the data.
-
The counterpart of descriptive statistics would be inferential statistics, which is often further classified in parametric and non-parametric methods.
-
A further branch of statistics that has become important recently are methods that concentrate on predictions, in particular the so-called machine-learning methods.
-
Finally, somewhat perpendicular to the others, there is the field of experimental design, which is concerned with how to collect data
See more discussion in the Essential Statistics lecture notes
An intro via youtube
