9  Variable types

Author

Samuel Blay Nguah MD

9.1 Variable types in research

Questionnaires are designed primarily to collect data. The data types (variables) collected can be broadly grouped into Qualitative and Quantitative variables.

9.2 Quantitative Data type

These are data types that can be numerically recorded and analysed as such. It includes:

9.2.1 Continuous data

This numeric data type can attain any number and to almost and expressed to almost an infinitesimal precision based on the accuracy of the measuring instrument. A typical example is weight, which can be measured to the nearest kilogram, gram and even milligrams. Others include height and length.

9.2.2 Discrete data

These numeric data types can only attain certain values. These include the number of cars, eggs etc. One can only speak of three eggs and 5 cars. Ther cannot be a fraction of a car for that matter.

9.3 Qualitative Data

Qualitative data types on the other hand do not have specific numbers, rather they are considered as numbers. These include:

9.3.1 Nominal data

This is a qualitative data type which has categories not ordered but simply have names. Examples include blood groups (A, B, AB, and O), race Black, Hispanics, Caucasian, etc).

9.3.2 Ordinal data

Ordinal qualitative data on the other hand have a specific order to it. These include staging disease as mild, moderate, and severe,

9.3.3 Binary (dichotomous) data

These are specific qualitative data types that take up just two possible options. A well-known one is sex (male and female) and any question that requires a yes or no answer. Others include the presence of disease, which could be present or absent.

9.4 Other data types

9.4.1 Percentages

These data types are usually stored as quantitative but are fractions. Purely analysing these data types as numeric types may be erroneous. For instance, 24% of 200 is not the same as 24% of 300. A typical example can be seen in an asthmatic patient’s FEV increasing by 24% following the administration of a bronchodilator. Note that this increase is not absolute and depends on the initial value.

9.4.2 Ratios or quotients

Ratios describe the relationships between two variables, both of which are usually numeric. We can for instance say the ratio of males to females is 1.4, meaning males are about 1.4 times the number of females. Common examples of ratios are the Incidence ratio of a disease, and body mass index (BMI)

9.4.3 Scores

Scores are arbitrary values with no meaningful numeric variable used to grade another variable. For instance, the pain score is used to grade how much pain a person experiences while the coma score is used to grade the severity of the coma present in an individual.