Definition of Variables
Variables are an essential concept in the field of research. Variables must be measured regardless of the subject matter of the study. Research would be impossible without variables.
A variable is any entity that exhibits variations. In other words, a variable is any characteristic or concept that is prone to change. Students' exam scores, for instance, are a variable since their scores will vary, i.e., they will receive different marks. Again, the ages of students in a school will be a variable, as individuals will have different ages.
If a concept is not a variable, it is a constant; in this instance, it does not undergo any changes. For instance, the gender of students at a school for females is a constant, as all students will be female.
Defining Variables
Every variable must be theoretically and operationally defined. The theoretical definition of a variable is the literature-based definition of the variable. For example, Kotler (2000) defined customer satisfaction as: “a person’s feelings of pleasure or disappointment resulting from comparing a product’s perceived performance (or outcome) concerning his or her expectations”. This is the textbook definition of customer satisfaction, which is attained by citing the concept's source.
However, this definition may not be very helpful in our efforts to measure customer satisfaction. Herein lies the role of the operational definition. The operational definition describes how the variable is measured or how it is observed in action. Therefore, for customer satisfaction, we will consider signs that indicate a particular consumer is satisfied. We may also choose to base this on literature. A delighted client may, for instance, leave a tip, smile, return for the same product or service, or refer the business to a friend.
In the design of your questionnaire to assess customer satisfaction, you might ask a respondent:
After a service at MTN Ghana, which of the following do you do (tick all that apply):
a) I smile
b) I give a tip
c) I return for the same service
d) I recommend the service to a friend
These allow the researcher to quickly determine whether or not a consumer was satisfied with a service. To put it another way, the researcher was able to measure customer service.
We say a variable has been operationalized when we indicate how it can be measured precisely.
Attributes or Values of a Variable
A variable's attributes or values are the qualities or characteristics of that variable. In other words, it refers to how the variable is operationalized, to borrow a term from the preceding section. The attributes of the variable gender, for example, are male and female. The table below shows some variables and their attributes or values:
In a close-ended questionnaire, the attributes of the variable are presented as options. Consider the following scenario:
Q1. What is your religion?
a) Christian
b) Muslim
c) Hindu
d) Traditionalist
e) Krishna
Q2. What is your gender?
a) Male
b) Female
Q3. Do you smoke?
a) Yes
b) No
It's worth noting that the variable relates to the question in each case, whilst the attributes correspond to the options. The attributes are entered into the values column in the variable view tab of data analysis software like SPSS.
In SPSS, variables are input via the variable view's variable tab. If it is a questionnaire, each question represents a variable. Consequently, if your survey contains 15 questions, you will have 15 variables. For each question, a variable name is created. For example for a question like:
How often do you wash your car? We can coin a variable like Carwashing_frenquency.
Classes of Variables
Variables can be categorized as follows:
a) Quantitative or Qualitative variables
b) Independent and dependent variables
c) Moderating variables
d) Intervening or mediating or intermediary variables
e) Extraneous variables
In a research endeavour, one or more of these categories of variables may be present.
a) Quantitative or Qualitative Variables
These are also known as continuous (quantitative) or categorical (qualitative) variables. These variables represent the quantitative or qualitative characteristics of the variations in the measurement of a variable.
Quantitative variables are those whose measurements result in numeric values. For example, for a variable such as test performance, the variable will be measured in numbers, such as 50 per cent, 40 per cent, 30 per cent, 89 per cent, etc. And the disparities will also be quantitative; for example, if two students scored 40 per cent and 30 per cent respectively on a test, the difference in their performance will be 40 per cent minus 30 per cent, or 10 per cent.
Again, with a variable such as the ages of pupils in a class, the following set of ages may be possible: 13, 12, 13, 14, 16, 12, 13, 15, 13, 15, 15, 16. The differences in their ages can be calculated quantitatively by subtracting one student's age from another's, as follows: (15 - 12) = 3 years.
Neither measurements nor variations can be made numerically for qualitative variables. In the case of these variables, variations occur in the form of attributes, organized into groups or categories. For instance, the gender variable is measured by specifying the categories, i.e. male and female. The variable educational level or level of education cannot be measured quantitatively. Instead, measurement is conducted by observing the categories of the attributes, i.e., primary, junior secondary, senior secondary, bachelor's, master's, doctoral, and post-doctoral.
Quantitative or qualitative variables affect the type of data generated from measurements. Quantitative variables provide ratio and interval data, whereas qualitative variables generate nominal and ordinal data.
Quantitative or Qualitative?
Occasionally, a quantitative variable may be transformed into a qualitative variable. For instance, when attempting to measure the ages of persons in a study quantitatively, it may be difficult if respondents are unwilling to provide their true ages. In such an instance, it would be necessary to change age from a quantitative to a qualitative variable. You will do this by entering age in categories, e.g., 15 – 20 years, 21 – 25 years, 26 – 30 years, 31 – 35 years, and over 35 years. Additional instances are shown in the table below:
Thus, respondents would feel at ease selecting a category rather than providing a specific number or figure.
b) Independent or Dependent Variables
This is the most prevalent variable category, as nearly all studies must include at least one variable from this group.
In research, the independent variable is the variable that influences the dependent variable. It is considered the "cause" in experiments, while the dependent variable is considered the "effect."
In regression, the independent variable is referred to as the regressor or explanatory variable, and as the manipulated, experimental, or treatment variable in experiments, since it is the variable that is manipulated or given a treatment. Other terms for the independent variable are exposure or input variable.
The dependent variable is the variable that is affected by or impacted by the independent variable. Normally, we are interested in the outcome. In some contexts, it is considered the problem the researcher is attempting to address.
In experiments, this variable is also referred to as the outcome, the criterion, or the post-test variable. In regression, this variable is known as the response variable or the regressand. Measured variable, observed variable, responding variable, and output variable are other terms for the dependent variable.
Variables that are influenced by other variables are dependent, and variables that are not influenced by other variables are independent.
Note:
A variable that is independent in one context may be dependent in another.
Typically, arrows are used to illustrate the link between independent and dependent variables in diagrams. For instance, the relationship between customer service and customer satisfaction can be illustrated as follows:
The concentric circles represent the variables, while the arrow illustrates their link. The arrow's endpoint points to the dependent variable, thus, independent variable --> dependent variable.
The relationships between independent and dependent variables are utilized to inform policy decisions. For instance, if a government is confronted with the issue of a rising crime rate in a country, it can rely on research to assist discover solutions to the problem. Typically, the issue at hand is the "dependent variable," also known as the "effect." The current task is to identify the "cause" of the problem, which will serve as the independent variable. Existing research indicates a correlation between crime rate and unemployment in this instance: as the unemployment rate in a country rises, so does the crime rate. As a tool to combat the escalating crime rate, the government can then concentrate its efforts on improving the employment rate. It makes sense to assume that occupied individuals would have less time to plan and commit crimes.
c) Moderating Variables
A moderating variable alters the initial relationship between an independent variable and a dependent variable. It can influence the direction or strength of an established relationship. They are depicted diagrammatically as shown in the diagram:
Examples include work experience (independent), salary level (dependent), and gender (moderating). Another illustration would include workforce diversity (independent), organizational effectiveness (dependent), and managerial expertise (moderating).
Without competent management, workforce diversity cannot attain the intended organizational effectiveness.
Moderating variables can be qualitative (non-numerical values such as race, socioeconomic status, and gender) or quantitative (numerical values like weight, reward level or age).
d) Mediating or Intervening or Intermediary variables
These variables establish a link between the independent and dependent variables. In other words, they describe the connection between the independent and dependent variables.
In a study examining the effect of income on longevity, for instance, a diagrammatic representation of the independent and dependent variables may not convey much information.
How does income impact longevity? Incorporating a mediating component, such as medical treatment, provides an answer to this question.
This demonstrates that a person's income level impacts his or her access to a certain degree of medical care, which in turn determines his or her longevity.
e) Extraneous variable/ Confounding Variables
An extraneous variable influences the dependent variable without the researcher's knowledge and is therefore uncontrolled. It can also be considered as any variable that is not the focus of your research but that could potentially influence your findings.
Uncontrolled extraneous variables can result in erroneous conclusions on the link between the independent and dependent variables.
Confounding variables are extraneous variables that are connected to both the independent and dependent variables.
Conclusion
Before beginning your study, it is crucial that you set out all the variables you plan to employ and the roles they will play. The variables can then be linked in a conceptual framework, a basic diagram that establishes the links among all variables. An illustration is presented below:
--ENDS--
About the Author
Comments
Please can we also have an article on the five research process ?
Thanks