VARIABLES IN RESEARCH

Identification of variables is an essential step in research. Variables help researcher to design theoretical model. They pave the way towards the research objective. Variables are various target points in research framework. They are like buttons on Joy Stick which cause overall situation under investigation. Having good understanding and control over these points will help researcher to develop convincing arguments in favor of the research objective.

Word “variable” is used to represent something which is moving or which keeps changing. It is a component which appears in different shapes and names. A variable may have different names depending on the angle researcher looks at the subject. That, it may be found on different investigating routes at the same time.  For example, the “gender” can be defined as variable from various angles in same study. In the discussion of child performance in class the gender of teacher could be an important factor. Gender as variable can be a factor effecting “child's performance” in the class. Gender can be a factor which defines the degree of relationship between teacher and student.  Sometimes, researcher defines variable from the point of view of its determinants which are present inside or outside the model.  In the discussion of “child’s performance in class” the gender of teacher is determined by some factors outside the model. That, the “gender” of teacher depends on other factors which are not included in the study e.g. employment regulations about gender of teacher. In each relationship a variable may take different angles. Therefore, for a researcher,  it is very important to clearly identify the positioning and labeling of the variable.  Following section will explain various types of variables and their classification.

1.                  Dependent and Independent Variables

2.                  Endogenous and Exogenous Variable

3.                  Intervening or Mediating Variable

4.                  Latent(confounding) and Manifest Variables

5.                  Quantitative and Qualitative Variables

6.                  Categorical or Measurable Variables

7.                  Continuous and Discrete Variables

In above example, the variable “gender” is taking position of an independent variable, if performance of child in class is affected by gender. For example, in primary school, female teachers are more effective than the male teacher. The reason is that female has natural disposition towards child development. Therefore, gender is a factor or independent variable for the research of child’s performance in class. The variable “gender” will become dependent, if we are treating  “gender”  of teacher as a result of the job description of the teachers at the time of appointment. Note that dependent variable is always considered as the objective of the research. If our objective is to understand the reason of downfall of economic growth, then GDP growth rate may be considered a dependent variable.

Another group of variables is known as Intervening and mediating variable. When a relationship between two variables is depending on the presence of some other variable then it is called intervening variable. For example, the gender-child’s performance relationship does not exist if teacher is not there. Therefore, the presence of teacher is necessary for the existence of the gender-child’s performance relationship. Here variable “availability of teacher” is an intervening variable. Similarly sometimes, we can find some variables which exist inside the model to define the quality of the relationship but its presence is not mandatory. For example, the age of a teacher can play an important role in the relationship between “gender” and “child’s performance”. Thus variable “age” is a mediating or moderating variable, which moderates the degree of relationship. This is also called sensitivity variable. Note that the presence of intervening variables is mandatory for the relationship, but not for the moderating variables. The presence of moderating variables can give more control to the researcher to define the dependent variable.

Another angle to look into the variables is the presence of Observed and Unobserved variables. Observed variables are also called manifest variables. And unobserved variables are called Latent or confounding variables.  The variables which are discovered and included into the model are called manifest variables, whereas the variables which are influencing the relationship in the model but yet to be discovered are called latent or confounding variables. In above example,  child’s performance may affect due to  some hidden factor (like income level of child’s family), but it has not yet been discovered or identified-called Latent or Confounding variable. But all other variables dependent or independent which are evident and measureable in the model are called manifest variables.

Quantitative and Qualitative variable is another classification of the variables. Quantitative variables are measureable variable which are measured through some scale of measurement. For example, age is quantitative variable and it is measured by an interval scale (number of years).  Whereas, gender is a qualitative or categorical variable, as it is measured by using nominal scale. Nominal means naming the values of variable. Qualitative variables are not measureable ( means that we cannot add two gender values to make third), rather we only count them (that how many male and how many female gender values are there). Qualitative variables are discrete in nature, as counts are only appeared in whole numbers. Discrete means that variable is only assigned with some values (whole numbers 1, 2, 3, 4 …). Whereas quantitative variables mostly appear as are continuous variables. Continuous means that variables can be assigned with any value within certain range. For example, age of a teacher can be any value between 25 years to 40 years. It may be 35 years and 11 months (35.11) or 26 years.
The term endogenous and exogenous is another angle to look into the variables. Endogenous variables are those variables whose values can be determined inside the model. Or in other words, the factors determined these variables are present inside the model. Exogenous variables, on the other had are those whose value is determined outside the model.
(Author: Muhammad Azeem)