Statistics is obviously all about data, either data that exists in one variable (such
a column of people’s ages) or multiple variables (such as columns of various financial
measures for a sample of companies).
However, as discussed in Chapter 2, data itself is not what you are aiming for. The
ultimate aim of statistics is usually to interpret that data through statistics.
This process includes the following steps:
-
Based on a research need, get data that reflects constructs and models you wish to study.
-
Generate statistics that summarize the data. Previously we didn’t explain the process. Having seen some of these statistics in
Chapters 7-10, we now see in
Figure 11.1 Extended diagram of the statistics process that the statistics generation process is in itself a sub-process, as follows:
-
Look for patterns in the data. See if the data has certain shapes or relationships either that you expect in advance
(you wanted to test a theory) or that you find along the way (data mining).
-
If you find a pattern,
create a summary model of the pattern. If you find patterns in the data, you will summarize the patterns into mathematical
”models,” which are composed of a few numbers that explain the way the pattern works.
-
Interpret the model. Use the summary model found in Step b above to explain how the world connected with
your data works and, if needed, predict how future similar data will operate.
The following sections expand on these steps and ideas within the sub-process of generating
statistics from data.