English French German Portuguese Russian Spanish
You are here Home
PDF Print E-mail

Types of variables


When carrying out a research, a researcher is supposed to understand everything about variables and be able to talk about them. A variable can be defined as any unit which can have different values which means anything which can differ can be referred as a variable. A good illustration of a variable is age because age has the ability to take different values between different kind of people or even to the same individual but at different period.

Variables are not always numerical or quantitative (Trochim, K 2006). Gender is a variable that consist of two values which is the male and the female. It is not essential to allocate a number to something so that it can be a variable. Variables are not only the things which we measure in the traditional sense because in program evaluation, it is considered as a program to be made of one or more than one variables which means that a cause can be a variable. Education program has unpredictable amounts of time on task, the ratio of scholar and instructor, the classroom settings, and many other factors. Therefore a program can be a variable and it can consist of a number of sub variables.

There are four measurement levels which are ordinal, nominal, interval and ratio. Features in the ordinal measurement can be rank ordered whereby the distance between the attribute does not matter. In the case of education attainment a higher number mean having more education (Trochim, K 2006).  During nominal measurement the numerical values name the attribute in a unique manner. The distance between attributes in interval measurement has meaning. The distance from 30 to 40 when measuring temperature is the same as the distance from 70 to 80. This indicates that the interval between the values is interpreted.

The ratio measurement explains that an absolute zero must exist which is of meaning, which mean that it is possible to construct a meaningful fraction using a ratio variable. The weight is an example of a ratio variable. When carrying out social research, count variables are ratio. An example is the number of customers in the previous four months because it is likely to have nil customers and also because it has meaning to articulate that I had two times as many customers as in the past 4 months as I did in the previous four months (Marlow, C 2011).



How to operationalise

How to measure



Age in years

How old were you on your last birthday?



Male and female




White or black

How many people are black or white



Educational attainment

What level of education are you in



Amount of income

How much were you earning in the last month?

Marital status


Unmarried, married, divorce or widower

Are you married, unmarried, divorce or widower?



Weight in pounds

What did you weigh in the last month?

Blood pressure


Blood pressure level

What is your blood pressure?



Yes or no

Are you hypertensive

Body temperature


Body temperature in degrees Celsius

What is your body temperature?

Health insurance status


Uninsured and insured

Are you insured or uninsured

Smoking status


Yes or no

Do you smoke?

Cancer stage


Stage I II III and IV

What stage are you at?


When age is operationalized in this order, the researcher will just need to question the respondent in order to determine what age she turned in the last birthday. When gender is operationalized in this order, the researcher will just need to count the number of female and male respondents to determine the population. Operationalizing ethnicity in this manner helps the researcher to know the population of the white people and the blacks.

When education is operationalized in this order the researcher will only need to ask the respondent so as to know what level of education he or she is in. When the researcher operationalized income in this manner, he will just need to question the respondent in order to know the income she earned in the last month. When Operationalizing marital status in this order, the researcher will be able to determine if the respondent is married, unmarried, divorced or widower.

The researcher will be able to estimate the weight of the respondent in pounds by asking the respondent what she weighed in the last one month. When blood pressure is operationalized in this manner, the researcher will just need to ask the respondent so that to know the level of blood pressure of the respondent. By the researcher asking the respondent if she is hypertensive, the researcher will be able to determine if the respondent is hypertensive or not. Operationalizing body temperature in this manner helps the researcher to determine the body temperature of the respondent.

When the researcher asks the respondent this question, the researcher will be able to know if the respondent is insured or uninsured. This question helps the researcher in determining if the respondent smokes or not. When cancer stage operationalized in this order, it makes it possible for the researcher to ask the respondent the question and be able to determine what stage of cancer she is in.


A variable can be differentiated into two which is a dependent variable and independent variable. This kind of distinction is important when investigating the cause and effect relationship. Independent variable is what one manipulates like a program, cause or treatment while the dependent variable is what is affected by the independent variable which is the outcome or effects. The measurement level in a research refers to the affiliation that exists between the values which are allocated to the features for a particular variable. The measurement level is important because it help one in deciding the statistics from a variable. The second importance is that it helps one to choose which numerical analysis is suitable for the values which were allocated.


Babbie, E & Rubin, A (2011). Research methods for social work Cengage Learning

Marlow, C (2011). Research methods for generalist social work Cengage Learning

Trochim, K (2006). Variables Retrieved on 9th July 2012 from

Trochim, K (2006). Levels of measurement Retrieved on 9th July 2012 from

Trusted Site Seal SSL Certificate Provider SSL