Types of Data in Quantitative Research
When collecting any quantitative data for your research, you will end up with two types of data, that is, categorical and/or numerical data. It is essential to understand the similarities and differences between the two data types since they are an important aspect of statistical analysis. This will make it easy for you to appropriately collect quantitative data and correctly apply statistical methods. Perhaps you have already come across the four levels or scales of measurement (nominal, ordinal, interval, & ratio) and you are a bit uncertain or confused about them? Don’t worry! This article will explain the two data types and the different levels of measurement in simple terms with useful examples.
What is Categorical Data?
Categorical data is a type of data that can be placed into groups or categories using names or labels. Each of the categories is mutually exclusive and the grouping is made based on data characteristics or unique qualities. For instance, gender is categorical data since it can be categorized into male and female. Essentially, categorical data is qualitative data that can be assigned numbers, such as 1 for males and 2 for females. There are two types of categorical data, namely; nominal data and ordinal data.
a) Nominal Data
Nominal data is a categorical data type describing groups or categories but with no natural order or rank between them. In other words, all categories have the same value and it is not possible to rank one above another. Examples of nominal data include;
- Gender (Male, Female)
- Ethnicity (White, Black, Hispanic, etc.)
- Marital status (Single, Married, Divorced, etc)
- Favorite mobile brand (Samsung, iPhone, Google Pixel, etc.)
- Political party (Republican, Democrat, Independent)
b) Ordinal Data
Ordinal data is a categorical data type that includes elements that can be ordered, ranked, or has a rating scale. It is not possible to numerically measure the difference between categories but one can logically order or rank them. Examples of ordinal data include;
- Socioeconomic status (low income, middle income, high income)
- Education level (high school, undergraduate, masters, doctoral)
- Level of agreement (strongly disagree, disagree, neutral, agree, strongly agree)
- Satisfaction rating (strongly dislike, dislike, neutral, like, strongly like)
- Political orientation (far left, left, center, right, far right)
What is Numerical Data?
Numerical data is a type of data that can be expressed in numbers and are quantitative in nature. This data is naturally measured as numbers and can only be collected in number form as opposed to assigning numbers in categorical data. For instance, age, height, and weight are numerical data. Also, for this data type, it is possible to carry out arithmetic operations like addition and subtraction. There are two types of numeric data, namely; discrete data and continuous data.
a) Discrete Data
Discrete data is a type of numerical data with countable elements, that is, the elements can be mapped one to one with natural numbers (whole numbers). Discrete variables can either take a finite number of values (i.e. can be counted from beginning to the end) or an infinite number of values (cannot be counted completely). Examples of discrete data are; number of students in a class, pages in a book, units sold, or the number of candidates in an election.
b) Continuous Data
Continuous data is a type of numerical data with uncountable elements. This is because they can take many values and can only be rounded to the nearest whole number or decimal places. For instance, the height of a person can be represented as 192 cm, or 192.2 cm, or 192.25 cm, and so on. For this reason, continuous data is represented as a set of intervals on a real number line. A good example of such data is participants’ age in a survey questionnaire which is collected using intervals, such as 20-29 years, 30-39 years, 40-49 years, etc. Similar to discrete data, continuous data can be finite or infinite. Continuous data is further divided into interview data and ratio data.
– Interval Data
Interval data is a numerical data type that can be measured in numbers, has an order/rank, and the differences between the measurement points are equal. Examples of interval data are;
- Temperature in Fahrenheit or Celsius
- IQ score
- GMAT or SAT score
- Credit score
- pH
Even though this type of data is naturally quantitative and you can measure the difference between points, it does not have a meaningful zero point, that is, the zero is arbitrary. For instance, a temperature of 0°C does not mean that there is no heat or temperature at all. Similarly, there is no zero point for IQ which would indicate that a person has no intelligence.
– Ratio Data
Ratio data is a numerical data type that has all the properties of interval data, and also has an absolute zero. This means that it can be ordered or ranked and the distance between numerical data points is consistent and can be measured. However, unlike interval data, the zero point is absolute. In other words, when a variable measurement is equal to zero, it means that there is none of that variable. Examples of ratio data are;
- Weight
- Length or height
- Temperature in Kelvin
- Concentration
- Pulse
Also, it is worth noting that for ratio variables, the numerical distance between two measurements has a meaningful interpretation. For instance, a weight of 10 kg is twice as heavy as a weight of 5 kg. However, a person with an IQ of 120 cannot be considered to be twice as intelligent as someone with an IQ of 60 since IQ is not a ratio variable. Similarly, a temperature of 10°C is not considered twice as hot as 5°C.
| Characteristics | Nominal | Ordinal | Interval | Ratio |
| Categories | Yes | Yes | Yes | Yes |
| Order/ rank | Yes | Yes | Yes | |
| Equal spacing | Yes | Yes | ||
| True absolute zero | Yes |
Does Data Types / Scales of Measurement Matter in Data Analysis?
Yes, it does. Knowing the level of measurement for a variable is important in choosing the right statistical techniques to use in the analysis. Each statistical test works with some types of data, where some work with categorical data, others with numerical data, while some work with a mix of both data types. For instance, you can code nominal variables with numbers but any calculations, such as computing the mean, median, or standard deviation, would be meaningless. Also, some statistical software like SPSS and R may “allow” you to run tests with the wrong data type but the results will be flawed and meaningless. Hence, it is important to know and understand the different scales of measurement before planning your data collection or designing your survey so that you can decide on the type of data to collect and the statistical analysis techniques to apply once you have got your data. For instance, instead of offering categories for income level and have an ordinary scale, you can get the actual income and have a ratio scale. Need help with your survey design or analysis? Order now or reach us at [email protected] to discuss your requirements.
| Computation | Nominal | Ordinal | Interval | Ratio |
| Frequency distribution | Yes | Yes | Yes | Yes |
| Mode | Yes | Yes | Yes | Yes |
| Median and percentiles | Yes | Yes | Yes | |
| Mean and standard deviation | Yes | Yes | ||
| Add or subtract | Yes | Yes | ||
| Ratios and coefficient of variation | Yes |