Quantitative Data Collection and Analysis

Survey Research

What is quantitative research?

Quantitative research measures attitudes, behaviours, opinions and other variables to support or reject a premise. This is done by collecting numerical data, which is easily quantifiable to identify “statistical significance”.

“Numerical data” is collected using close ended questions, such as Likert scales or multiple-choice question types. Once you’ve shared your survey with your target audience, the responses for each answer option is easily quantified.

The questions you ask must be objective in order to collect and analyze respondent data.

Choosing between qualitative and quantitative data collection

Before designing a survey, you must determine your research goal. I.e. what you want the data you collect to help you achieve.

In academic research, this may be to prove or disprove a hypothesis. Whereas, in market research you might be gauging the popularity of product concepts.

After defining a research goal, you must ask yourself; “what are the expected results from my survey?”

If you’re expecting in-depth feedback or experiential data, you should consider running a qualitative survey instead.

If you’re looking to draw conclusions from a representative data set, then a quantitative survey is for you.

It’s important to determine your research goal and method before writing survey questions. This way, you ensure everything you ask contributes to your aims, rather than collecting data for the sake of it.

How to collect quantitative data

Typically, quantitative research requires a large respondent population. This is because the results of your research will be representative of a larger population.

Quantitative methods

There are several methods by which you can collect quantitative data, which include:

Experiments
Controlled observations
Surveys: paper, kiosk, mobile, questionnaires
Longitudinal studies
Polls
Telephone interviews
Face-to-face interviews

The method by which you collect quantitative data can impact your participant’s responses. For example, participants taking part in an experiment are unlikely to make the same choices as they would in a social context.

How participants respond to stimuli is context-bound, meaning they may respond differently to each of these methods.

If you plan to conduct your research using more than one of these methods, we’d recommend doing this in stages.

Firstly, analyze the data for each individual method to identify any contextual trends. Then, when you take a look at your collated data, you’ll have more information to explain any inconsistencies.

Quantitative data analysis

There are two types of data that require statistical analysis; descriptive and inferential. Descriptive statistics enables you to draw conclusions from data the data as a whole.

Inferential statistics highlight “statistically significant” differences between two or more groups of data. There are two types of inferential data: comparative data and relationship data.

Types of quantitative data

Descriptive data

Simply put, descriptive statistics allows you to easily quantify a variable (or variables) for a target group.

For example, transport surveys may look for data concerning how many people aged 18-25 use public transport to get to work.

In this example, the variable is ‘public transport’ and the target group is those aged 18-25.

Comparative data

This method of data collection allows a comparison two or more target groups based on one or more variable.

A comparative study may aim to compare the use of public transport between people aged 18-25 and those aged 26-35.

Relationship data

Relationship data highlights trends, associations and relationships between two or more variables.

In market research, you may set to find out what the relationship is between customer gender and loyalty to your brand.

Strengths of quantitative research

Objective analysis

Bias is less likely to occur in quantitative analysis, as the data are statistic in nature. Meaning there is less room for a researcher’s subjective reasoning to effect results.

Easy to analyze

Quantitative data is easier to analyze than its qualitative counterpart. Plus, if you’re using online survey software, you can apply filters for a more in-depth view of your data.

Conclusions are generalizable

If your sample population is representative of the study population, your findings can be generalized to that larger group.

Weaknesses of quantitative research

2 Dimensional data

Quantitative data is excellent at providing answers to what, who and when questions. But isn’t great for how and why questions.

Qualitative research is a better option for exploring participant experiences and opinions. Combining thise research method with quantitative data provides more context to participant answers.

Large number of participants needed

For the results of your quantitative research to be representative of a larger population, you’ll need to distribute to a large sample.

If you don’t have access to a list of respondents, you’ll have to spend time building one of an appropriate size. You’ll also have to ensure the sample is representative of a larger demographic or group. Otherwise your results will be meaningless.

Confirmation bias

This form of response bias can occur in quantitative studies, where researchers favour data that confirms their beliefs or hypothesis.

The researcher may not even be aware of this type of bias. They simply ignore or misinterpret data because they believe their research goal has already been met.

We find that quantitative and qualitative research methods often yield better results when used in tandem.

The quantitative approach allows you to draw fact-based conclusions to inform research or decision making.

Qualitative data can provide further context to participants’ answer choices.

For example, if you were to run a customer satisfaction survey you would ask: “how likely they were to recommend your product to a friend or colleague?”

This close-ended question will determine the average brand loyalty of your customers but achieves no more than that.

By following up with an open-ended question, e.g. “what one thing would you change about our product?”, you give context to those choices. In this way, your survey data is not only descriptive but actionable.