Comment définir ses variables de recherche - English version
Comment définir ses variables de recherche - English version
Modifié le 24/02/2025 à 09h14

Variables (or Factors)

A variable is an object, an idea, or a condition that can have multiple values or formats. Each variable has a specific form within a possible set. Ex: variable reading medium > digital, text, etc.
In other words, it is a measurable characteristic that can take on different values or modalities.
Examples: height, age, gender, income, country of birth, years of education, type of housing, etc.
In a research study, we always seek to observe the effects of a variable whose values do not change (independent variable = IV) on another variable whose variations are measured (dependent variable = DV).
In a research study, we will formulate a hypothesis. The researcher will establish the hypothesis that the independent variable (with fixed values) will impact the dependent variable (the one being measured).
Variable 1 = what explains, what is fixed. This is the cause.
Variable 2 = what is explained, what is measured. This is the effect.
There is an optional third type of variable in a research study. It is the controlled variable: a variable that must be considered because it could impact the variation of the dependent variable.
Controlled (or Extraneous) Variable = a factor that could influence variations in the Dependent Variable (DV).
To simplify understanding and organization of variables, it is often said that Variable 2 (which is explained) depends on Variable 1 (which explains, which varies). Hence the name dependent or explained variable.
dependent and independent variables
The dependent variable (DV) is the variable tested and measured by the experimenter in a scientific experiment.
For better understanding:
Example 1
Julie wants to learn how to ski. She observed skiers to imitate them, but she thought she could improve faster by taking lessons. So, she enrolled in private lessons to enhance her level.
We can hypothesize H1 that the number of lessons Julie takes will influence her skiing level at the end of the vacation.
The dependent variable is "Julie’s skiing level at the end of the vacation". That is what we will measure!
Example 2
Researchers want to know if students' learning performance can be improved by guidance in an e-learning platform.
We can hypothesize H1 that the presence of guidance on the platform will lead to better learning performance.
In this hypothesis, the measured variable is the students' learning performance: the dependent variable is learning performance.
The independent variable (IV) refers to what the experimenter identifies and controls. It is manipulated by the researcher through different conditions, as well as through different groups that are set up.
An IV corresponds to characteristics identified as potential causes of the DV, such as:
  • Individual characteristics (e.g., gender, age, level of expertise, etc.).
  • Physical or social environment (e.g., presence/absence of others, wall color, noisy vs. quiet environment).
  • The task (difficult vs. easy, familiar vs. unfamiliar).
  • The nature of presented stimuli (e.g., ambiguous vs. unambiguous; subliminal vs. supraliminal).
These characteristics are manipulated by the researcher to control or analyze their impact on the behavior, state, or mental process of participants.
An IV has at least 2 conditions or states (also called levels or values) identified by the experimenter.
Example 1
According to the working hypothesis H1, the number of lessons Julie takes will influence her skiing level at the end of the vacation. The skiing level is the DV (which is measured).
The independent variable (fixed, explanatory variable) is the number of lessons:
IV: number of lessons = N3
  • n1: 0 lessons
  • n2: 1 lesson
  • n3: 2 lessons
Here, the IV has 3 conditions: 0, 1, and 2 skiing lessons.
For example, the modalities of the variable family status could be: Single, widowed, married, in a relationship, etc.
Important
To better distinguish between types of variables (independent and dependent), considering the cause/effect relationship can help.
If an independent variable is modified, an impact should be observed on the dependent variable. In other words, the evaluation of the dependent variable (measured) should not be the same depending on the modality of the independent variable. In the skiing example, Julie's skiing performance should not be the same if she took no ski lessons, one lesson, or two lessons.
The controlled variable is a “extraneous” variable, meaning that it might interfere with the dependent variable (the one being measured).
This variable is fixed, just like the independent variables (which are manipulated intentionally), but here, the researcher wants to control the effect of the controlled variable on the dependent variable. Controlled variables are not always identified, nor necessary.
Example 1
The number of lessons Julie takes will influence her skiing level at the end of the vacation.
Possible controlled (extraneous) variables:
  • Snow quality
  • Julie’s sleep quality
  • The ski instructor’s mood
  • Etc.
Example 2
Students' learning performance improves with the implemented guidance.
Possible controlled (extraneous) variables:
  • The learner’s psychological state
  • Their motivation
  • Their prior knowledge of the subject
  • Etc.
ski example
summary table