What is the point of a statistcal tool in research? How does it affect the gathered data?

181 views

What is the point of a statistcal tool in research? How does it affect the gathered data?

In: 0

3 Answers

Anonymous 0 Comments

Thanks everyone! The replies here were really helpful

Anonymous 0 Comments

Research involves two primary things.Start with a good research question, and design a study to answer that question. In that process you will (1) choose and use statistical tests to analyze the relevance of the data,, and (2) you will make methodological choices about how to gather and analyze the data.

The “best” studies look at one variable, and make all the other variables that can possibly affect an outcome the same. As the important variable is changed, do the outcomes differ? If so, it may well be that the *cause* of the changing outcomes is the variable that is changing.

The goal is to find Truth in the Universe, or TITU. Properly applied statistical tests give us an estimate of the uncertainty around the conclusion we’ve reached, and for the vast majority of us the mechanism by which that witchcraft occurs is a black box. BUT if a number of statisticians independently agree that this or that statistical test is the proper one for this data set, we’re probably good.

Hold on though, back to the study design. For anything beyond bench research, the methods chosen to collect and analyze data have inherent flaws and biases, so these must also be considered when estimating the error between our experimental conclusion and TITU. The ability (and desire) to skeptically apply critical reasoning to published studies isn’t all we might want it to be, which is why science is rarely settled, and learning the TITU isn’t as easy as we would wish.

Anonymous 0 Comments

The point of statistics is to quantify uncertainty; putting a number on how likely it is that the difference or association you’re seeing in your data is truly there, or just a result of the inherent randomness of sampling and measuring.

Good research involves thinking about statistics *before* gathering data, to make sure you set things up in a way that will ultimately let you draw meaningful conclusions.