7.1 Concepts, Constructs, Measures, and Variables

Failure to to understand what the data columns in a dataset represent can result in fundamentally flawed analyses. An understanding of concepts, constructs, variables, and measures can help us get clarity of what columns in a dataset represent.

Concepts, Constructs, Measures, and Variables

We use concepts all of the time to help us make sense of our world. A concept is an abstraction or generalization drawn from observing common characteristics of multiple instances. For example, we observe multiple people paying money for the privilege to temporarily occupy a place. We call the concept ‘paying rent’.

Concrete Concepts

We have experience with many concrete observable concepts. For example, objects, behaviors, and transactions are common concrete concepts we collect information about.

• Objects are persons, places and things such as customer, building, and product are common objects.

• Behaviors are things that are done by individuals such as sleeping, studying, and eating.

• Transactions are usually exchanges between multiple individuals or organizations. Common examples include purchase, work, employment, going out on a date, marriage, and so forth.

Height, size, and color are example attributes of objects. Actor names, duration, date, time, location, and amount are common examples of behaviors and transactions.

Constructs

In contrast to concrete concepts, we also use abstract concepts that are not directly observable that help us make sense of our world. Examples constructs include happiness, recreation, love, hate, satisfaction, IQ, morality, etc.

Scientist refer to these abstract concepts as constructs because they have been conceptually constructed to represent a name of something real that cannot be directly observed but may be useful to study or describe.

Since constructs are not directly observable, in order to measure them we often use measures which are deemed to be manifestations of the construct. Measures should be chosen that are reasonable empirical indicators of a construct. For example, resting heart rate is an empirical measure of cardiovascular health.

Sometimes multiple measures exist and are used for the same construct. For example, we could measure happiness by asking people how happy they are on a scale of one to five. Another indicator of happiness is to ask them how often in a day they have positive thoughts. These are just two of many ways that could be used to measure happiness.

Variables

A variable is any characteristics or quantity that can be measured or counted. The term variable is used because because the value may vary across different units in a population, and may change in value over time. A variable may also be called a data item. Example variables include age, sex, expense, country of birth, capital expenditure, class grades, eye color and vehicle type.