For example, one might use an average to represent a fact such as the amount customers are willing to pay for ice cream. However, if 95% of respondents stated that they would spend $5 or less on a pint of ice cream, and 1% of respondents stated that they would spend $100 on ice cream, an average would be skewed by the 1% who would spend much more. In this case, a researcher may decide that a different statistic, such as the median, would more accurately represent the findings. Making these judgments is an important step in the quantitative data analysis process, as are ensuring that data is properly cleaned and coded prior to analysis.

Quantitative Analysis Methods

Quantitative data is analyzed using statistical methods, as quantitative data represents numbers from which datapoints can be calculated. Data from a quantitative dataset, such as survey results, is usually loaded into a program such as Excel or the statistics software SPSS which enables researchers to quickly create tables and charts in order to examine findings. Often the first step in analyzing a dataset is to view top-level findings using descriptive statistics such as mean, median, and mode.

Descriptive Statistics

In the below definitions, we will use the example of a survey with 400 respondents who were asked to rate their opinion of chocolate ice cream on a scale of 1 ‘strongly dislike’ to 5 ‘strongly like’. The data indicated that 100 rated an ice cream flavor a ‘5’, 200 rated it a ‘4’, and 100 rated it a ‘3’.

• Mean or average: The numerical average of a set of numbers.
  • In the above example, the average rating would be ((5×100)+(4×200)+(3×100))/400= 4
• Median: The median is the midpoint in a set of numbers.
  • In the above example, the median would be the number in the 200^th row of data. In this case it would be 4, but depending on the dataset, the median can be different from the average.
• Mode: The number that occurs the most often in a dataset.
  • In the above example, this would also be 4 as it occurred 200 times, while 5 and 3 only occur 100 times each.
• Range: A statement that represents the lowest and highest numbers in a dataset.
  • In the above example, the range would be from 3 to 5.
• Distribution or Percentage: The percent represented by each category within a group, out of the total (100%).
  • In the above example, instead of looking at the dataset as a whole this would report that ‘25% rated the ice cream a ‘5’, 50% rated it a ‘4’, and 25% rated it a ‘3’ 

Cross Tabulations

After examining descriptive statistics, researchers may use cross-tabulations to dig deeper into a dataset. A cross tabulation or crosstab is a way to show the relationship between two variables and is often used to compare results by demographic groups. For the above example, we could create crosstabs to show results by age:

Crosstabs can also be created to examine one datapoint by another, such as if those who rate chocolate ice cream highly also rate vanilla ice cream highly, or if there is a different relationship between the two variables. Crosstabs are useful to better understand the nuances of a dataset and the factors that may influence a datapoint.

Calculating Statistical Significance

When researchers are looking to prove or disprove hypotheses, they will often also use measures to calculate the statistical significance of their findings. Measures of statistical significance demonstrate if a finding is merely due to chance or if it is a significant finding that should be reported on. In the above example, without calculating statistical significance we cannot be sure if the difference in results between those aged 18-24 and 25-34 is due to the difference in age groups, or if the findings are a coincidence based on the sample that was selected and not related to age.

Statistical significance is usually represented by a statistic called a p-value. A p-value is a calculated number between 0 and 1, and the lower the p-value is, the less likely it is that the results were due only to chance. Typically, a p-value of less than 0.05 is regarded as statically significant, as it means there is a less than 5% likelihood that the results were due to chance. While having a p-value of under 0.05 doesn’t necessarily mean that the stated hypothesis is true, it decreases the chances that any differences in the dataset are occurring by chance. Researchers who are running tests to make decisions, for example to determine if populations prefer vanilla or chocolate ice cream in order to make purchasing decisions, should use a test of significance in order to have more confidence in their decision making.

Programs including Excel, R and SPSS can calculate the significance of findings through a series of steps, outlined in more detail here. If you work with a full-service research agency such as GeoPoll, we can run statistical significance tests for you and include the resulting data in our data analysis.

Conduct Quantitative Data Analysis with GeoPoll

GeoPoll is a research company that gathers data for international organizations, governments, consumer brands, and media houses which enables better decision making. Our services range from study and questionnaire design to data analysis, including the creation of data tables, crosstabs, and full research reports. To learn more about our capabilities or get a quote for your next project, please contact us.

The post Quantitative Data Analysis appeared first on GeoPoll.

Difference Between Quantitative and Qualitative Data

The primary difference between quantitative and qualitative data is that quantitative data represents data that can easily be measured or quantified, such as the number of people who have bought a product. Qualitative data represent opinions or feelings and cannot be represented by a numerical statistic such as an average.

For example, if a survey asked 500 respondents the question “Did you buy ice cream today?”, and 300 responded ‘yes’ while 200 responded ‘no’, we would know that 300/500 or 60% bought milk, a quantitative fact. If the same survey asked an open-ended follow-up question: “Why did you choose the brand of ice cream you bought?” you would receive qualitative insights that are unique to each respondent. One person may say, ‘I liked the packaging and label colors’ while another may state, ‘It was the first one I saw on the shelf.’ These descriptive insights cannot easily be quantified into numbers, so they are qualitative.

Qualitative vs Quantitative Analysis

Another difference between quantitative and qualitative research is how data is analyzed. While quantitative data can be analyzed statistically and calculated into averages, means, and other numerical data points, qualitative data analysis involves a more complex system.

To glean insights from qualitative data, researchers conduct a manual analysis of datasets and often code responses into categories. For example, to analyze focus group data, researchers could review transcripts or recordings and group similar sentiments together into categories. Due to this manual process, qualitative data analysis is a longer and more labor-intensive process than quantitative data analysis, which is another factor to keep in mind when deciding what type of data to collect.

While some methods such as focus groups typically collect qualitative data, other methods such as surveys often collect quantitative and qualitative data within one survey instrument, as outlined below.

Quantitative Data Examples

Quantitative data is collected through several methods, including surveys, controlled experiments, and certain observation types. Quantitative data types include:

• Yes/no questions
  • “Did you go to work today? 1) Yes 2) No”
• Single choice questions
  • “What is your favorite flavor of ice cream? 1) Vanilla 2) Chocolate 3) Cookie Dough 4) Peppermint 5) Chocolate chip”
• Multiple choice or ‘select-all-that-apply’ questions
  • “Which of the following products did you buy last week? 1) Toothpaste 2) Soap 3) Vegetables 4) Meat 5) Grains 6) Bread”
• Ranking questions
  • “Please rank the statement ‘I enjoy ice cream’ from 1: Strongly disagree to 5: Strongly agree”
• Numerical range questions
  • “How much money did you spend at the grocery store today? Please respond with a dollar amount”
• Quantitative observations
  • Observations that can be categorized or quantified, such as the number of times a person checks their phone in a given time. These observations often take place in a controlled environment.

As all of these question types collect data that fit into set categories or can be calculated into averages and other statistics, they are quantitative.

Qualitative Data Examples

Qualitative data can also be collected through certain types of survey questions, in addition to interviews and focus groups. Examples of qualitative data include:

• Open-ended survey questions
  • “Why is cookie dough your favorite flavor of ice cream?”
• Unstructured or semi-structured interviews
  • Unstructured and semi-structured interviews allow topics and questions to flow naturally, rather than only asking questions from a set question list in a specific order.
• Focus groups
  • In focus groups, multiple people have a discussion (in-person or via an online or mobile-based chat group) facilitated by a trained moderator who gives prompts to start conversations.
• Unstructured observation
  • Researchers can gather qualitative data through unstructured observations, such as observing participants as they partake in certain activities such as shopping.
• Documents or content analysis
  • Reviewing documents to better understand a particular topic or categorize elements of documents is a type of qualitative research.

Data collected from these methods and question types do not provide numerical statistics but instead, give insights that are often longer and more detailed than their quantitative counterparts.

When Should I Use Quantitative or Qualitative Research?

Once you understand the types of data provided by qualitative and quantitative research and the methods for each, it’s essential to understand how to utilize each type of data best. Generally, quantitative data is used to answer precise questions and prove or disprove hypotheses, while qualitative data provides richer insights on a smaller scale.

Qualitative research is often conducted at the beginning of a study when researchers are looking to gather broad, unstructured information on a topic to create a hypothesis, which can then be more clearly answered by quantitative research. Qualitative data collected through unstructured interviews or focus groups can also inform the development of a more structured questionnaire administered to a larger group.

For example, a focus on different ice cream brands may uncover that participants generally consider price and packaging first. That information can then be inputted into a quantitative question: “Which is more important to you when buying ice cream? 1) Price 2) Packaging” administered to a nationally representative sample.

Qualitative data may also be used as part of a mixed-methods research study to add additional context to quantitative data. A researcher may administer both a quantitative questionnaire and conduct a qualitative analysis of interviews with subject-matter experts to form a more robust conclusion.

Surveys can also be split between qualitative and quantitative; Many surveys are mostly quantitative questions that can be quickly analyzed, plus one or two qualitative questions that provide deeper insights into the topic being studied.

Quantitative vs Qualitative Data: Definitions and Uses Cheat Sheet

GeoPoll has experience designing and administering both quantitative and qualitative research studies around the globe. Our research methods include surveys with closed-ended and open-ended question capabilities, mobile-based focus groups, concept testing, and more. To learn more about GeoPoll’s capabilities, please contact us today.

The post Quantitative vs Qualitative Data appeared first on GeoPoll.