What is interval data?

Interval data is a data type that is measured on a scale where each value is placed at an equal distance (interval) from one another. Simply put, it is a way of measuring something using equal intervals. It is always in the form of numerical values but cannot be multiplied or divided. They can be added or subtracted, but there is no true zero on an interval scale. 

In research, interval data is essential because it can support most statistical tests. Most forms of research, such as business, social, or economic, use this type of data.

As one of four types of data (nominal, ordinal, ratio, and interval), interval data is used in many quantitative studies that calculate demographic information, test scores, or credit ratings.

1. Interval data is measured using intervals that are consistent in the difference in their value.

2. Interval data shows order, direction, and the exact difference in the value.

3. There is no true zero in interval data.

4. They may not be multiplied or divided but can be added or subtracted.

5. Statistics that can be calculated using interval data include:

• Frequency distribution in numbers of percentages

• Central tendency: mode, median, and mean

• Measures of variability: range, standard deviation, and variance of a data set

• Parametric tests (e.g., t-tests, linear regression, ANOVA)

What's the difference between interval data and ratio data?

Both interval data scales and ratio scales both have equal intervals between values. But they are different in that only ratio data has a true zero. Interval data does not.

If the data that you are working with uses negative numbers, you will be working with interval data. Think about the thermometer. A Fahrenheit thermometer measures quantities below zero.

For this reason, only ratio data can calculate the ratios of your values because, unlike interval data, they can be divided and multiplied.

What is ordinal vs. interval data?

Ordinal data, like interval data, can be categorized and ranked. Unlike interval data, however, ordinal data is not evenly spaced between values.

An example of ordinal data is finding the top five homes sold last month in the area. They could be ranked one through five by sales cost. But with interval data, you could take it one step further. Out of the top five homes sold last month in the area, two sold between $400,000 and $500,000, two sold between $300,000 and $400,000, and one sold for between $200,000 and $300,000. 

Test scores are an example of interval data. SAT scores that range from 200 to 800 do not use an absolute zero because there are no scores below 200. Teachers use interval data when grading tests or calculating the grade point average. Even credit scores use interval data.

Time, using a twelve-hour clock, is another example. The distance between each number on the clock is equidistant and measurable, so the distance between four and five o’clock is the same as the distance between five and six o’clock.

Temperature, when comparing Celsius and Fahrenheit, is a type of interval data since there is no absolute zero. It is most commonly used as a tool in statistical research, particularly if performing measures of probability or studying a specific population.

There are various ways to collect interval data. Some of the most common are:

• Interviews can be done over the phone, face-to-face, or online. The interview is structured with a standard set of questions for data collection.

• Observation generally involves counting, either by using a naturalistic or standardized approach. An example might be the number of attendees at an event during different time frames.

• Surveys are conducted either through a web-based questionnaire, an online questionnaire, or a questionnaire conducted with pen and paper. They are usually short and easy to navigate.

• Document reviews are useful tools for gathering data from existing documents. These could be in the form of public records or other documents that contain historic data for your research.

• Probability sampling is done when researchers carry out a random selection to make probable conclusions based on the data they collect.

Interval data is used in nearly every major category of data analysis. From teachers grading papers to scientists plotting temperature changes, it is a common type of data measurement. Some of the most common uses are as follows:

Marketing and advertising

When a new market is introduced, a new company enters the marketplace, or a product upgrade or change is made, marketing and advertising departments will likely use a SWOT analysis. This is to determine the strengths and weaknesses before further investment is made.

Interval data is retrieved to examine internal and external factors affecting the product's viability or introduction. It can help determine pricing, customer demographics, and how a product or service stacks up against the competition.

Product development

By studying current users, product development can better grasp what improvements are needed and which ones will be well received.

During the product development stage, researchers use TURF analysis to investigate whether a new product or service will be well-received in the target market or not.

Education

The education sector has been using interval data for years to compute a grading system. Interval data is used in determining a student's GPA as well as scoring on normal test scores.

With some testing—for example, SAT scores— the digits 0 to 200 are not used when scaling the raw data to the section score. As a result, it is interval data.

Medicine

In the medical field, body temperature is often measured in Fahrenheit (interval data). Doctors and statisticians also use interval data in determining usage amounts, age, and BMI.

Interval data, being a numerical data type, can be analyzed by several quantitative methods. The level of measurement you use, however, will decide on the type of analysis you use.

In gathering information, you can either use descriptive statistics or inferential statistics, regardless of the scope.

Descriptive statistics summarize the characteristics of the data set, while inferential statistics compare different treatment groups and make generalizations about the larger population of subjects.

Descriptive statistics for interval data

If you are using descriptive statistics to summarize interval data, you are probably working with frequency distribution, central tendency, or variability. 

Read further to get a better understanding of each type of measure.

Frequency distribution

A common method for organizing data is through a frequency distribution.

Frequency distribution looks at how data is distributed, using both hierarchical and evenly spread categories. It may be organized in a graph such as a histogram or a line graph that easily lets a researcher view data at a single glance. Researchers can determine whether the observations are high, low, concentrated, or spread out.

Measures of central tendency: mode, median, mean

You are also able to measure central tendency using interval data. These three measures are mode, median, and mean.

The mode is the value that occurs most frequently in your dataset. The central value is the median, and the average is the mean. These values are easily identified when charted, with the mode being the largest or most frequent value, the median landing in the center, and the mean being the average of all the values, which sometimes takes some calculation.

Measures of variability: range, standard deviation, and variance

If you are measuring variation, you can extract different measures of variability such as range, standard deviation, and variance. Range is simply the difference between the smallest and largest values. Standard deviation measures the amount of variation, and variance measures the amount it varies from the mean.

Inferential statistics for interval data

When using inferential statistics, you draw comparisons, and the data is often analyzed using parametric tests, which use clearly defined parameters. 

Types of tests include the following:

• T-tests are used to compare the means of two groups using hypothesis testing. You only need to know the mean difference between values, the standard deviation of the sample, and the sum of data values in the groups. However, for many variables you have, the t-test will help determine the relationship between the dependent and independent values.

• Analysis of variance is much like a t-test but compares the variance of three or more data samples. While the t-test compares two samples, the analysis of variance (ANOVA) can determine the relationship between independent and dependent values, regardless of the number of variables.

• Pearson correlation coefficient measures both the direction and the strength of a linear relationship between two variables.

• Simple linear regression uses only two variables but can predict the relationship between them or can measure the impact that the independent variable has on the dependent variable.

Age is considered a ratio variable because it has a “true zero” (birth). A person who is 40 years old is half the age of someone who is 80.

ANOVA can be used to compare three or more sample means using interval data.

Unlike nominal- and ordinal-level data, which are qualitative in nature, interval- and ratio-level data are quantitative.

The Mann Whitney test is used primarily for ordinal data.