Unit of Analysis, Concepts, Constructs and Variables

  • Unit of Analysis: It is the target of the study as part of the social science research. It can be a person, group, organization, country, technology, objects etc.
  • Concepts: Properties or characteristics of objects, events or people which can be generalized.
  • Constructs: They are abstract concepts that are useful to explain a phenomenon. They can be unidimensional where the constructs have one concept (or) multidimensional where the constructs have multiple concepts. They should have operational definitions which makes it clear as to how they will be measured.
  • Variables: In general, it represents something that is variable. In scientific research, a variable is a measurable representation of an abstract construct. For example, Intelligence is a construct and IQ score is a variable. There are different types of variables. Each of these variables can be either continuous (those that can take continuous values) or discrete (those that can take non-continuous values).
    • Independent Variable: They explain the variation in other variables like dependent variables and mediating variables.
    • Dependent Variable: They are variables that depend on independent variables and mediating variables.
    • Mediating Variable: They are variables which explain the dependent variables or other mediating variables. Their variation is explained by other mediating variables or independent variables. Mediating variable is one which is an intermediate step in the relationship between a dependent variable and an independent variable.
    • Moderating Variable: They are variables that are influencing the relationship between an independent variable and a dependent variable.
    • Control Variable: These are variables which that must be controlled (kept constant for a given experiment) in a scientific study as they may have some impact on the dependent variable though they will not be explaining the variation on the dependent variable.
    • Confounding Variable: These are variables that which influence both the independent and dependent variable. Such variables must be correlated with the independent variable (without necessarily having a causal relationship) and must be having a causal relationship with the dependent variable.

↑ Research

References