Quantitative data
What is Quantitative Data?
Quantitative data is defined as any information that is mathematically precise and allows for calculative operations.
For example, height, weight, distance, speed etc have numerical values and are mathematically quantifiable. Therefore, the height of a person, for instance, is quantitative data.
Quantitative data can also be derived from scales in surveys and questionnaires. For example, a survey question asks customers to rate their level of satisfaction on a scale of 1 to 10, 1 representing least satisfaction and 10 being the highest satisfaction. On such responses, mathematical operations are possible, such as finding the average customer satisfaction, which category of products have the highest satisfaction score etc.
As we see, quantitative data representation can also help transform subjective/ qualitative aspects into quantifiable information for mathematical analysis.
Key Characteristics of Quantitative Data
To better understand all the aspects of quantitative data, let us take a deeper looking into its key characteristics:
- Quantitative data has true mathematical numerics:
Quantitative data is always represented numerically and has a true mathematical value. This means the numbers it represents allows for mathematical calculation to be performed to derive meaningful insights, such as mean, median, average etc. This is unlike the qualitative data scale of nominal data where the number is merely a place-holder for the qualitative value.
For example, a vet clinic’s survey questions on gender of people’s pet animals (nominal data, qualitative), does not contain any mathematical data in the responses, other than the number of respondents themselves who selected an option. On the other hand, a survey question on the age of their pet will give precise numerical data that will allow for mathematical analysis on the responses.
- Quantitative data is objective:
Quantitative data is purely objective and eliminates subjectivity for the sake of analysis. This doesn’t mean that a researcher or reader cannot add subjective observations on the final information.
For example, a survey poll asks respondents to pick their favourite animal as pet from a multiple choice question. And the results show that dogs topped the list, followed by cats, rabbits, goats, deers etc. Now when the data was collected, it assumed pure objectivity in this classification. However, in reality, the respondents may have picked 2 or more animals, had there been an option to do so. And this may be subjectively pointed out by any reader of the survey results.
Objectivity and subjectivity are not as exclusive in real life as it may seem on paper. However, for research and analysis, quantitative data is objective in nature.
- Quantitative data has mathematical precision:
One of the cornerstone characteristics of quantitative data is its mathematical precision. If the data doesn’t have this quality, it is not quantitative data.
For example, when someone is rating their level of happiness, if the responses are on a numerical scale, say 1 to 5, then that data is quantitative. But if the response is in the form of a sentence, no matter how precise the answer may be, it is qualitative in nature.
- Quantitative data has scalability and range:
Quantitative data can always be plotted on a scale to represent a range. Even qualitative values can be made into quantifiable information and can be represented on a scale with a range of options.
For instance, a scale of agreement, such as very much agree (value 2), agree (value 1), no opinion (value 0), disagree (value -1), very much disagree (value -2). Here the response options may be qualitative in nature, but have been made quantitative to accommodate mathematical evaluation on a scale, within a range.
Quantitative Data Collection: Key Methods with Examples
- Surveys
Surveys are the most common method of primary data collection for quantitative analysis, especially through online survey platforms/ software that allows for complex logical branching of questions based on previous answers. This allows for customization and personalization of the questionnaire, allowing for more accurate and complete information gathering.
Application of surveys are seen everywhere today as companies and institutions become more aware of customer-centricity and the success that it brings. They are seen after most customer interactions, be it in e-commerce or in-store, aiming to understand their satisfaction.
For example, a delivery experience rating scale on Amazon is something most of us must have seen.
- Polls
The micro format of a survey is a poll which is typically limited to one question with simple multiple choice response options. They are very prevalent across social media platforms on a range of topics.
In many cases multiple polls may be a better method for capturing quantitative data than a survey, as it is a quick method for responses capture when the questionnaire doesn’t require much complexity and logical branching. However, multiple polls on the same audience may not be possible in many scenarios, and in such cases are best for capturing responses to single, unrelated questions only.
For example, polling is also common in the political analysis field where they are used for analyzing a candidate/ political party’s chances of winning an election. However, as we have seen in recent times, political polls may be biased, have inadequate sampling for responses or inaccurate since many respondents may choose not to respond accurately in lieu of privacy concerns.
- Documentations
Any existing documentation that is quantitative in nature, can serve as secondary data for quantitative analysis for any research. This can be in the form of published research papers, industry standards, market research surveys and polls, company documentation, public records etc.
For example, government’s public records, when trust-worthy, often serves as secondary data for quantitative analysis related to labour and job, population census, market and economy etc.
Similarly, such documentations can be any internal records of companies that may be utilized for internal analysis. For example, internal financial records may be used by the accounting team to understand profit and loss patterns and tax saving trends.
- Quantitative experiments
Quantitative experiments can be any experiment that results in quantitative results. These can be sales figures, marketing data, web analytics, physics, chemistry etc.
For example, a marketer wants to experiment with A/B testing of landing pages and the resulting data can be sales leads conversions from these pages to determine which one is most user friendly and has persuasive content.
Experimental conclusions need to be backed by documented and verifiable quantitative data that can help in final decision making.
- Quantitative observations
Observations are a method of collecting information by not engaging or interfering, but rather looking at patterns as an unbiased viewer. They are widely used in usability testing of products and software, spiritual therapy, psychology, pet care etc.
For example, in usability testing, the researcher watches how the tester is interacting with the software’s interface and notes how much time it takes for them to complete a certain task. The longer it takes, the less usability score is attributed and vice-versa.
Another example is in pet care where a new dog parent may observe how many times a puppy is eating, at what intervals etc, and documents them to derive patterns in digestion health.
Best Practices for Quantitative Data Management
Quantitative data management is simple when the parameters for data collection, documentation, quality management etc are set correctly.
Here are the key best practices to get you started:
1. Define quantitative data collection objectives and methods
Ensure that internally there is a clear definition of the objectives of your quantitative data collection effort. This includes identifying the specific questions you want to answer and the hypotheses you aim to test. Knowing your objectives helps you choose the most appropriate data collection methods and ensures that the data collected is relevant to your research goals.
Select the data collection method that best fits your research objectives and the nature of the data. Common methods include surveys, experiments, observations, and existing data sources. Each method has its advantages and limitations, so consider factors such as the type of data, required accuracy, and available resources.
2. Set data quality and documentation process
Setting data quality involves using standardized procedures for data collection, training data collectors adequately, and pre-testing instruments like surveys or questionnaires to identify and correct any issues.
Document how the data is being collected, how the audience is being sampled for response collection, any considerations and logic behind them etc. This ensures that the quantitative research process is transparent such that anyone referring to the data can verify and understand the process used to collect the information.
3. Set data privacy compliance
Ensure that the privacy and confidentiality of the participants are protected throughout the data collection process. This builds trust and encourages participation while complying with ethical standards and legal requirements.
There are several data protection laws across geographies and researchers need to ensure compliance to such rules and regulations.
4. Cite sources for any secondary data
For any secondary data being used, it is an essential practice to cite your sources. Always point to the final study papers rather than blogs or articles that have referred to it. For example, if you come across a useful article that has cited secondary data from a research paper, please cite the research paper itself, and not the article.
You can always directly cite news/ research articles that are first hand reports, and are not using other papers/ studies to draw conclusions.
