Basic HTML Version
© 2014 Pacific Crest
237
5.4
Transforming Data
P
urpose
The role this topic plays in quantitative reasoning
The use of data is pervasive and extends to nearly every aspect of life. Data is collected from many
sources and used in many different ways. Learning to transform data will strengthen the value and
meaning of your data by building upon the first three experiences in this chapter: Generating data,
Obtaining data from others, and Organizing data. At its most basic, data transformation is changing the
nature of the data so that it can be reused for additional purposes.
Because of the way that data is captured and stored, its value is often minimal unless it is transformed.
Data transformation is a powerful part of the quantitative reasoning process that can make data that
seems unlikely to produce any meaningful results into a much more useful resource. Different types of
transformations can produce new data structures, new forms of data, and can even change the quantity
of data in a data set.
The more that you understand data, the more useful it will be. A comprehensive overview of a data set
would include:
●
Understanding the context of the data: What you observed (the target of that observation), the
number and range of observations, and the time and place of these observations
●
Knowing the units and quality of the measurements of the observations (each observation has a
set of characteristics; the measurements of a characteristic are stored in what is called a variable)
●
Noting any unusual features of the process of collecting the data that may have affected the data
The ability to transform data provides an opportunity to take a current data set and produce both different
structures as well as new variables in order to make the data more valuable. There are five major facets
to transforming data:
1. Changing the structure of the data set by collapsing or expanding the number of observations
2. Expanding the number of variables by transforming the current variables
3. Changing the unit of measure (conversion)
4. Expanding the number of variables by using the current variables to derive new variables
5. Projecting the data
L
earning Goals
What you should learn while completing this activity
1. Appreciate the opportunities and potential for data transformations
2. Become familiar with a tool set of standard transformation techniques that are commonly used to
produce greater meaning for a data set
3. Identify the types of limitations and conditional concerns that can come from the transformation of
original data
4. Document the justification and tracking of the transformation process