Variable is key term in research. Every research involves variables to be measured. When the variables are not clear, it is difficult for the researcher to conduct the research. The important thing is to define the variables and to discuss the types of variables. Variable is the focus in any research. Without knowing what variables are involved, what variables refer to, and what the role of the variable in the research is, a research project cannot be conducted. In fact, research always starts with variable. So, here I will share article about variables. This article describes variables and discusses types of variables including continuous vs discrete variables, independent vs dependent variables, and confounding variables.
Defining variables
Variable is something that can be measured.Variable is defined as “characteristics that tend to differ from individual to individual, though any two or more individuals may have the same variable trait or measure”. (Charles, C.M, 1995:29).
Example:
- Unclear variable: “School children achievement” and “Beauty”
- Clear variable: “School children’s achievement in English at the end of grade 8 in Tulungagung” “physically like weight, height, and speed. Others like intelligence, achievement, and personality are mental constructs which require operational definitions to measure.
Unclear variable
“School children achievement” is not clear because it can refer to any subject matter achievement. While beauty, it is not a measurable variable because it is difficult to get objective measures of “beauty”.
Types of Variables
Based on the characteristic there are two types of variable, namely continuous and discrete. Based on the roles in research there are three types of variables, namely dependent, independent, and confounding variables. It is very important to know the types of variables because as a certain research design requires a certain type of research.
Continuous Variables
Continuous variables are variable that show gradational differences in the same trait processed by individuals (Charles, C.M, 1995:29). For instance, height, weight, and speed. The adjective modifier can be used to explain these variables. The word very, and more like very high, very fast, very tall, very big, or taller, bigger, faster. It can be used in the correlation research, like the correlation between reading strategies and reading comprehension. This variable cannot be used in causal designs to measure, for example, the effect of students’ reading skills to their writing.
Discrete variables
Variables are classified as discrete variable if they are naturally categorical, like sex or handedness which gives the choice of either male or female, right handed or left handed (Charles, C.M, 1995:29). This variable group people of the same traits. In sex we divide people into male and female group. Variable intelligence is not discrete variable although we can group people into high, medium, and low IQ. It is because the members of high IQ group still vary in their intelligence. Besides, intelligence can be modified into very smart, smarter, and the smartest. A variable cannot be discrete and continuous at the same time. Discrete variables can be used in causal designs to measure the effect of, for example, sex to the student’s achievement.
Independent and Dependent variables
Independent and dependent variables are used in causal designs measure the effect of independent variables to the dependent variables. For instance, the effect of sex (independent variable) to the students’ language achievement at school (dependent variable). The independent variable, which turn may change as it is affected by the independent variable (Charles, C.M, 1995:29).
In the correlation research which does not measure cause and effect relationship, the terms independent and dependent variable are not use in the variables. In correlation which measure the degree of the relationship between student’s reading skills and their writing skills, the two variables are not dependent or independent because one of them does not exert influence on the other. The terms are Predictor as variable X and Criteria or it can be called Criterion for Correlational research design.
Confounding Variables
Confounding refers to interfering unexpectedly. Variables are considering confounding if their presence unexpectedly interfere research outcomes, and therefore needs to be controlled. The example is on causal research. As like, the female Junior high school students achieve better significantly in language learning than male Junior High School students may be intervened by unexpected variables like intelligence and motivation.
Analyze this situation;
If the sample representing the female group happens to have higher intelligence and higher motivation than the sample representing male group, the students’ achievement in language learning (dependent variable) is not only affected by sex (independent variable) but also by the unexpected presence of the variables intelligence and motivation (intervening variables). Which have not been successfully controlled during the process of sample selection. Or when the sample representing the male group which is found to achieve significantly lower in language learning than the female group happens to consist of male students who have a lot of eyesight and hearing problems (Organism variables), then the students’ achievement in language learning (dependent variable) is not only affected by sex (independent variable) but also by the unexpected presence of the variables eyesight and hearing problems (organism variables).
Charles, C.M. (1995: 30) defines that another confounding variable involves physical or environmental conditions, for instance fatigue, distraction, over excitement, discomfort, or test anxiety.
That is about variables of educational research. In conclusion, a researcher has to know the variables involved in the research. In fact, when the variables are not clear the research cannot be conducted. In the simple terms, the research is failed.