By Laura Mora / Instagram: @implementingscience / Email: @
Variables and error sources in Epidemiology
Epidemiology is the study of how diseases are distributed and spread among populations and the factors that
influence their occurrence and impact. Therefore, variables are needed to conduct and solve a hypothesis or a
research question. Regardless of the study design, you will always need variables, thus the importance of knowing
the differences and their categories.
A variable is a characteristic of interest that can vary among individuals or populations. Variables can be
classified based on their nature and role in the study, such as:
According to their nature
● Categorical Variables: Gender (male, female), smoking status (smoker, non-smoker), blood type (A, B,
AB, O).
● Continuous Variables: Age (in years), blood pressure (in mmHg), weight (in kg), income (in dollars).
or Level or scale of measurement:
● Nominal Level: Categories with no inherent order or ranking. Example: Blood type (A, B, AB, O).
● Ordinal Level: Categories with a natural order or ranking but no fixed interval between categories.
Example: Education level (high school, college, graduate).
● Interval Level: Numeric scales with equal intervals between values, but there's no true zero point.
Example: Temperature in Celsius or Fahrenheit.
● Ratio Level: Numeric scales with equal intervals between values and a true zero point. Examples: height,
weight, and age.
This is an example of how to organise or present the operationalisation of your variables.
Variable Definition Nature Level Example
Sex Two main categories Categorical Nominal Female, male
(male and female) into
which humans and most
other living things are
divided on the basis of
their reproductive
functions
Another important concept to consider in epidemiology is errors and their sources, starting from variability.
Variability is a natural difference or fluctuation in data due to inherent variability in individuals or measures.
There are two types of variability.
Inter-observer variability occurs when different observers measure the same data differently.
● Example: In a study assessing the severity of skin lesions, dermatologists may assign different scores
based on their subjective assessments of their characteristics, such as size, colour, and texture.
Intra-observer variability occurs when the same observer measures the same data differently on separate
occasions.
● Example: A radiologist reviewing chest X-rays may interpret the presence or absence of lung nodules
differently when reviewing the same set of images at different times, leading to variability in the reported
findings.
Systematic Error: Bias introduced by flaws in study design, data collection, or analysis that consistently skew
results in one direction. These are some of the types:
• Selection bias occurs when there is a systematic difference between those selected for a study and those
not, leading to an incorrect estimate of the association between exposure and outcome.
Variables and error sources in Epidemiology
Epidemiology is the study of how diseases are distributed and spread among populations and the factors that
influence their occurrence and impact. Therefore, variables are needed to conduct and solve a hypothesis or a
research question. Regardless of the study design, you will always need variables, thus the importance of knowing
the differences and their categories.
A variable is a characteristic of interest that can vary among individuals or populations. Variables can be
classified based on their nature and role in the study, such as:
According to their nature
● Categorical Variables: Gender (male, female), smoking status (smoker, non-smoker), blood type (A, B,
AB, O).
● Continuous Variables: Age (in years), blood pressure (in mmHg), weight (in kg), income (in dollars).
or Level or scale of measurement:
● Nominal Level: Categories with no inherent order or ranking. Example: Blood type (A, B, AB, O).
● Ordinal Level: Categories with a natural order or ranking but no fixed interval between categories.
Example: Education level (high school, college, graduate).
● Interval Level: Numeric scales with equal intervals between values, but there's no true zero point.
Example: Temperature in Celsius or Fahrenheit.
● Ratio Level: Numeric scales with equal intervals between values and a true zero point. Examples: height,
weight, and age.
This is an example of how to organise or present the operationalisation of your variables.
Variable Definition Nature Level Example
Sex Two main categories Categorical Nominal Female, male
(male and female) into
which humans and most
other living things are
divided on the basis of
their reproductive
functions
Another important concept to consider in epidemiology is errors and their sources, starting from variability.
Variability is a natural difference or fluctuation in data due to inherent variability in individuals or measures.
There are two types of variability.
Inter-observer variability occurs when different observers measure the same data differently.
● Example: In a study assessing the severity of skin lesions, dermatologists may assign different scores
based on their subjective assessments of their characteristics, such as size, colour, and texture.
Intra-observer variability occurs when the same observer measures the same data differently on separate
occasions.
● Example: A radiologist reviewing chest X-rays may interpret the presence or absence of lung nodules
differently when reviewing the same set of images at different times, leading to variability in the reported
findings.
Systematic Error: Bias introduced by flaws in study design, data collection, or analysis that consistently skew
results in one direction. These are some of the types:
• Selection bias occurs when there is a systematic difference between those selected for a study and those
not, leading to an incorrect estimate of the association between exposure and outcome.
