Section 5.3 Types of Data Science Questions
We will look at a few different types of questions that you might want to answer from data. This flowchart gives some questions you can ask to figure out what type of question your analysis focuses on. Each type of question has different goals.
There are four classes of questions that we will focus on:
-
Descriptive: The goal of descriptive data science questions is to understand the components of a dataset, describe what they are, and explain that description to others who might want to understand the data. This is the simplest type of data analysis.
-
Exploratory: The goal of exploratory data science questions is to find unknown relationships between the different variables you have measured in your dataset. Exploratory analysis is open ended and designed to find expected or unexpected relationships between different measurements. We have already seen how plotting the data can be very helpful to get a general understanding about how variables relate to one another.
-
Inferential: The goal of inferential data science questions is to is to use a small sample of data to say something about what would happen if we collected more data. Inferential questions come up because we want to understand the relationships between different variables but it is too expensive or difficult to collect data on every person or object.
-
Predictive: The goal of predictive data science question is to use data from a large collection to predict values for new individuals. This might be predicting what will happen in the future or predicting characteristics that are difficult to measure. Predictive data science is sometimes called machine learning.
One primary thing we need to be aware of is that just because two variables are correlated with each other, doesn’t mean that changing one causes a change in the other.
One way that people illustrate this idea is to look at data where two variables show a relationship, but are clearly not related to each other. For example, in a specific time range, the number of people who drown while falling into a pool is related to the number of films that Nicholas Cage appears in. These two variables are clearly unrelated to each other, but the data seems to show a relationship. We’ll discuss more later.
