Research variable is the fundamental building blocks of research. It is the basis for a research hypothesis. A research study examines the relationship of a research variable in a particular context. A variable can be defined as anything that takes different values. The same applies to research variables as well. In research, variables are observable characteristic of an event or object. They can take one or more values. The common characteristic of all variable is that they do not have a fixed value. A research variable should have three characteristic features: it should be defined, it should have a function, it should have a measurement. (Stommel & Wills, 2004)
Every research variable has an operational definition, that is specific to the context of the study. The operational definition provides meaning for the variable. The variable specifies a specific operation in the research. For example, behavioral variable ‘x’ is defined as the response of an organism to green light, in a study that examines the effect of green lights on animal behavior. Operational definition should also mention, how the variable will be measured in the study. It will ideally provide, also the standardized instrument that will be used to measure the variable. Operational definition will also state the direction in which the variable will be manipulated through experimentation. Variables are frequently used in research for the purpose of inquiry or for stating a presumed causality. Based on the function of a variable, in the context of presumed causality: it can be classified as an independent and dependent variable. (Sommel & Wills, 2004)
An independent variable is under the control of the researcher. He can manipulate and measure it. The independent variable is not affected by the outcome measured by the research tool. It is not influenced by factors that cause the outcome. Dependent variable on the other hand, is a measure of the effects of the independent variable. It is also called the output or response variable. (Cottrell & McKenzie, 2005)
Variables used to define a purpose, are usually classified as control variables and moderator variables. Control variables or blanks, are used to cancel out the influence of exogenous factors on the measured variables. Moderator variable on the other hand, helps to introduce relationship among different variables used in the study. (Cottrell & McKenzie, 2005)
Variables are also classified based on the scale used to measure them. They can be continuous or categorical and nominal, ordinal, interval, or ratio. Continuous variables take values from a range of values. For example, a continuous variable can have its value as 4, or 4.5 or 5. Any value between 4 and 5 is acceptable. However, categorical data can take only the value of a category. For example, it can take have a value 4 and 5, but not a decimal figure like 4. 5, 4.2, etc.
Nominal variable represent names, symbols or classes. They do not have any set order. Ordinal variable on the other hand have a numerical order. Intervals and ratio, take a specific order, but they have numerical data with an equal meaningful interval. (Stommel & Wills, 2004)
The variables can be classified into different groups called factors. The number of groups under each factor are called levels. Studies with more than one independent variable are examples of different factors. Multivariate analysis methods handle more than one independent variable. (Stommel & Wills, 2004)
Data collection method: Data are the values taken by the variables. They are information collected, using suitable tools or research instruments. The task of data collection starts, when the research problem is identified and the research plan has been worked out. Data can be broadly classified into primary and secondary data. Primary data is one that is collected by the investigator himself for the purpose of the research study in question. Secondary data, is when the researcher collects data from other studies. Examples of secondary data include: vital statistics published by different public and private sector, hospital records, school records, etc. While using secondary data, the investigator must determine the date of the data, and also verify that it is relevant in demonstrating recent trends. (Jackson, 2008)
Primary data can be directly collected by the investigators through different methods like survey, questionnaire, interview, experiment and observation. Unlike secondary data, collecting primary data can take more time and money. A researcher, who ventures into collecting primary data, should ensure the suitability of the research instrument be used to collect data, population characteristics, sample, standardization of the instrument, pilot study and ways of using the instrument to collect data. (Jackson, 2008)
A data which is collected for the first time is unique and may lack cohorts to compare with. The choice of data collection method largely depends on the research question. The choice of the research instrument and its accuracy is very critical in obtaining correct information or data. Example of research instrument includes: questionnaire used in an interview, a test used to measure an observation for a variable, etc. Instrument is somethings that will helps to collect data. (Jackson, 2008)
The instrument is designed by the researcher to provide measurable and observable data. The data thus collected, is suitable for further analysis and provides answers to the research problems. The instrument for collecting data is designed or selected based on the objective of the study. Data is usually collected in a systematic process, ensuring that the sample chosen for data collection represents the study population. Data is usually collected under controlled environment. The process of choosing the variable and collecting data is very critical in addressing once hypothesis. (Jackson, 2008)
References
Cottrell, R. & McKenzie, J. (2005). Health promotion and education research methods (pp. 80- 85). Sudbury, Mass.: Jones and Bartlett Publishers.
Jackson, S. (2008). Research methods (pp. 84-90). Belmont, CA: Thomson Wadsworth.
Stommel, M. & Wills, C. (2004). Clinical research (pp. 11-13). Philadelphia: Lippincott Williams & Wilkins.
