Variable: Defined based on how it will be used/treated
The Variables can be divided into Dependent variables and Independent variables based on how a researcher will treat them. Note that the scale that a variable has been measured does not influence if this variable can be treated as Dependent variable or as Independent Variable but it influences the statistical way that can be analyzed.
Dependent variables: Definition
Dependent variable is the presumed result that a researcher is waiting to see after the manipulation of other factors, usually of the Independent variable.
Independent Variables: Definition
An Independent variable is the variable that can influence the status of another variable, that of the dependent variable. The Independent Variable is the variable that the researcher is manipulating in order to see the presumed effect on the dependent variable. In summary, the Independent variable is the presumed caused and the Dependent variable is the presumed effect or result. That is, the effects of Independent variable on the Dependent variable are studied.
A researcher can be interest in to study the effects of Tea and Coffee (Independent variable / manipulated) on the Quality of Sleep of participants (Dependent variable / measured outcome).
A Researcher is interested in to study if different type of music (Independent variable / manipulated) can influence / affect the hours that you sleep (dependent variable, measured outcome).
Extraneous Variables are those variables that a researcher failed to control or to include into his/her research design but it can influence the final output / outcome. That is, extraneous Variables can alter the influence of the Independent variable on the Dependent Variable.
Extraneous Variables: Special case: Confounding variable
Note that when the levels of the independent variable varies according to some Extraneous Variable/s, then this variable is called Confounding Variable.
Extraneous Variables: Special case: Control variable
When the Experimenter is examining the relationship between the Independent and Dependent variable, and keep constant some other variable/s then these variables are called Control variables.
Extraneous Variables: Example Ι
A researcher can be interest in to study the effects of Tea and Coffee (Independent variable / manipulated) on the Quality of Sleep of participants (Dependent variable / measured outcome). Here, Extraneous Variables can be the mood of a participant as well what he/she has eaten or drunk, or how many hours he/she has slept before the experiment. Therefore, these extraneous Variables may have affected the levels of his/her Quality of Sleep. That is, Tea or Coffee may had an effect on the levels of well-being but these extraneous Variables may also have influenced the Quality of Sleep of the participants.
Extraneous Variables: Example ΙΙ
A Researcher is interested in to study if different type of music (Independent variable / manipulated) can influence / affect the hours that you sleep (dependent variable, measured outcome). The level of sound or the hearing quality of the participants can be such Extraneous Variables which may have affected also the hours of sleep. That is, higher level of sound or limited hearing may have tired some participants more / less than some other participants, and thus, their Quality of Sleep was negatively influenced.
Confounding variable: Example explained
If the Experimenter had divided the sample into age levels / categories e.g. young vs old, then, these older people may had limited hearing capabilities or they were more prone to have disturbed sleeps than some other older people that did not participate in that experiment. Then hearing capability is a confounding variable of this experiment.
Control variable: Example explained
If the Experimenter would like to test the Quality of sleep by Manipulating the type of Drinking before the sleep, then by keeping constant e.g. the room temperature, the noises inside and outside the building then, such variables are called control variables.
As it appears, a researcher may fail to take into consideration all the extraneous factors / variables that can influence the effect of independent variable on the dependent variable in an experiment. A well designed experiment will try to take control of most of such factors / variables.