Over time, your research might rely on both qualitative and quantitative data. It's important not to treat one as more important or better than the other. Instead, it will benefit your research if you know when and how to use both forms of data to address your research inquiries.
Quantitative data refers to any numerical data that can be used in statistical analysis or experimental research.
Researchers in quantitative research often collect data and analysis to generalize results or make conclusions about a particular phenomenon or subject. Survey researchers can sample a portion of a population and assert that the survey results are indicative of the perspectives of the whole population.
Collecting quantitative data
Generally, quantitative data collection methods are easier than their qualitative data counterparts. Suppose your research question involves measuring foot traffic around a city. In such a project, a researcher could place volunteers at selected places and have them count how many times they cross a street in their view. The volunteers' counts make the quantitative data they need to answer their research questions. Making assertions about the foot traffic in different places is a relatively easy task, given that the numbers are easily collected and readily available for comparison.
Forms of quantitative data
Quantitative data collection relies on structure and a clear understanding of what the numerical values mean to the research. Quantitative researchers can readily take a spreadsheet of test scores, for example, to generate descriptive statistics and inferential statistics. The shape of that spreadsheet (e.g., rows and columns) and its content (e.g., numerical data) ultimately make analyzing quantitative data feasible.
Limitations of quantitative data
Some phenomena cannot be reduced to mere numbers. For example, quantitative data may tell you the value of a particular product, but it has significant challenges in helping explain a product's inherent beauty or effectiveness. Such concepts can be difficult for quantitative data to define. After all, what is beautiful to someone will be less so to someone else, and vice versa.
Quantitative research may also have problems measuring people's perspectives. Survey research often relies on Likert scales or rating scales asking respondents to rate something on a numerical scale (e.g., one to five or one to ten). However, is one respondent's idea of a "4" on a five-point scale the same as another? Subjective concepts are thus difficult to capture with numerical data.
Finally, the role of quantitative research is primarily to confirm understanding. Think of quantitative research as a coin sorter dividing coins by size.
Qualitative research tends to look at the detail of a phenomenon rather than its numerical value. Qualitative research methods allow for theoretical development or exploration of a relatively unfamiliar phenomenon.
Think about a beautiful song. It might be beautiful because of the melody, singer, lyrics, or perhaps some combination of these and other factors. Collecting quantitative data on each aspect (e.g., "Give the melody of the song a score between one and five") might allow for some statistical analysis of a song.
However, what exactly does someone mean when they highly rate a song's melody or lyrics? Perhaps they mean the melody is relaxing or inspiring. Quantitative approaches alone are insufficient in allowing researchers to determine what people think is a "beautiful melody."
Coding qualitative data
Qualitative research relies on methods like interviews to explore social phenomena beyond the use of numbers. ATLAS.ti lets researchers code data, summarizing large sets of information more succinctly, so that gathering insights becomes easier.
When someone speaks at length about a song's melody being "relaxing," a researcher can apply the code "relaxing melody" to an entire segment of text in ATLAS.ti. That way, analyzing the data means looking at brief codes instead of lengthy paragraphs or pages where the meaning might be unclear.
Developing theoretical insights
Qualitative data analysis can also prompt us to look at a phenomenon from new and different angles. A researcher may conduct in-depth interviews where some respondents think a song is beautiful when accompanied by a visual presentation, like at a live concert. This may not fit our prior understanding of a beautiful song, meaning quantitative research wouldn't likely capture it.
The potential for qualitative research to further develop theory cannot be understated, particularly when it allows researchers to document new insights that quantitative methods cannot perceive. While the qualitative research process can be daunting, it can provide more detail than other approaches.
Forms of qualitative data
Qualitative studies often draw from the following data collection methods:
This is not an exhaustive list, as any unstructured data that can be organized might be considered qualitative data.
What is especially important is that qualitative data is not confined to text. Most forms of information can be analyzed for more insightful discussion. ATLAS.ti allows researchers to code major forms of qualitative data, including images, audio, and video.
Unlike quantitative data, qualitative data tend to lack an easily defined structure that could help with data analysis.
On the other hand, qualitative data often lacks the structure that makes understanding the data easy. Data can be neatly organized in a spreadsheet for easy data retrieval. In semi-structured interviews or focus groups, however, researchers may ask follow-up questions that can't easily be predicted.
The need for the interviewer to pursue deeper answers can impede the organization of data into neat rows and columns. However, the meaning of the answers can often be more useful than the organization of the data. Understanding this means taking the time to reorganize and interpret that data.
Moreover, interpreting non-numerical data is a significant challenge for qualitative researchers. The relative quantitative value of different things can be relatively easy to interpret. For example, if someone takes the temperature of New York and the temperature of Chicago on the same day and gets two different values, asserting that one city is warmer than the other would be uncontroversial.
However, people may disagree about what makes a city interesting or exciting. To take from our example about music, people may even disagree about whether the visual or performance elements of music should be considered. Thus, the researcher needs to clarify the potential differences in understanding between people.
Analyzing qualitative data to answer such research questions requires transparency in qualitative analysis. Qualitative researchers analyzing socially constructed, subjective concepts should clearly define their concepts so their audiences understand the data analysis.
People can make the mistake of thinking both forms of data are mutually exclusive. Both approaches are useful in determining cause and effect relationships and drawing conclusions based on key takeaways from analysis.
Choosing research questions
Your inquiry will determine whether quantitative data or qualitative data are more appropriate for your research. In a nutshell, think about how your research question guides you.
A quantitative research question seeks to confirm something that researchers already know exists. On the other hand, a qualitative research question looks at something unfamiliar and seeks to understand it more deeply. The research question you ask is more important than deciding whether one approach is generally better than the other.
Quantitative methods rely on concepts like statistics to accurately measure a phenomenon. This is essential in the "hard" sciences where researchers have to rely on a significant degree of accuracy.
However, social scientists may also employ a quantitative method to reaffirm theories regarding social constructs.
Perhaps the opinion poll is the first example that comes to mind regarding statistics. Pollsters conduct interviews with members of the public to gather their perspectives on a particular person, product, or service. The numerical information from aggregating all the answers can clarify whether the population supports or opposes the subject of the opinion poll.
ATLAS.ti lets users create statistics for their research. Tools such as Code Co-Occurrence and Code-Document Table count the codes employed in a project, allowing researchers to generate statistics around the frequency counts of codes.
The words used in a text and how often they are used tell a story about how people write and speak. The Word List function in ATLAS.ti can give frequency counts on the different words found in text documents.
Content analysis can also apply to groups of words (e.g., positive and negative words). Sentiment Analysis in ATLAS.ti uses machine learning to guess which segments of text are positive or negative, allowing researchers to code sentiments that can be analyzed statistically.
Analyzing qualitative data may be more challenging and time-consuming, but several analysis methods can help make the process smoother. Here are some qualitative methods to consider.
Researchers can look for the most apparent themes in their qualitative data. Imagine you're examining a crowd of people. Can you tell which clothing color is the most common? Which type of clothing is seen the most? What are the most typical possessions people hold?
Parsing the data for themes is a common analytical method in qualitative research. After coding your data in ATLAS.ti, you can use the Code Manager to view which codes representing your themes appear the most often in your data and make assertions based on those themes.
The qualitative research process often involves relaying to your audiences what you see in the data. Instead of having an external theory guide your research, you can generate themes based on your understanding of the data and clarify that understanding to your audience. Qualitative methods such as grounded theory create a useful distinction in the qualitative vs. quantitative debate. You can uncover new insights from grounded theory rather than merely confirm if something is true.
The data collection methods you employ may guide the form of data produced. Still, it's important to understand that the method does not exclude qualitative data from being converted into quantitative data and vice versa. For example, qualitative data from interviews may take the form of transcripts or audio recordings, but researchers can easily reorganize the interviews to generate quantitative data.
ATLAS.ti has several tools that allow researchers to generate quantitative data from their qualitative research project. The Word List tool, for example, provides the frequency of different words and tells researchers which words appear the most in their research project. This can easily help researchers conduct content analysis. Moreover, the quantitative data can be exported to Microsoft Excel for statistical analysis.
Research questions may restrict your data collection methods, but the larger research agenda can benefit from a more holistic understanding that requires both quantitative and qualitative data. The most productive inquiry pursues both quantitative research and qualitative research.
The mixed methods approach seeks to gather in-depth insights through multiple methods. As a result, qualitative and quantitative research work hand in hand so that each approach covers the shortcomings of the other.
When choosing between qualitative or quantitative data or both, the bigger question is what you want to know, which determines the data collection methods and data analysis that are most effective for your research project. Researchers can benefit from understanding the strengths and weaknesses of quantitative and qualitative data and deciding how both can benefit their research.