When you think of the word "data," you might think about numbers and tables and organized spreadsheets. Qualitative data, on the other hand, can take on so many forms and serve so many purposes that it's important to examine the topic in greater detail.
Qualitative data is any unstructured or unorganized data that is difficult to define statistically or numerically. There are many uses for collecting and analyzing qualitative data, such as understanding social phenomena, gathering people's opinions on various subjects, and building evidence for recommendations.
Qualitative data can be transcripts of interactions with people, notes about observations, and even multimedia such as pictures, videos, and audio recordings.
Qualitative and quantitative data are almost always juxtaposed against each other. Quantitative data contributes to statistical analysis, while qualitative data contextualizes a phenomenon.
For example, consider the difference between comparing the average temperatures of two different cities and comparing the innate beauty of those two cities. The former can be quantified so that researchers can reach a quick conclusion about the differences in the climate. On the other hand, the latter is rather difficult to reduce to numbers. Even if beauty can be placed on a ten-point scale, what does "7 points" or "4 points" on the beauty scale mean? How does someone determine such a score? Understanding the beauty of a particular city requires qualitative data collection methods.
Anthropologists, professionals engaged in health services research, and market researchers collect and analyze qualitative data. While the data collection method in each field may differ, fields that commonly employ qualitative research have research questions that an analysis of numerical data cannot easily answer.
Regardless of your field, your needs, or your data collection method, ATLAS.ti can help you collect qualitative data easily and efficiently.
Researchers often perceive a divide between qualitative data and quantitative data and get into debates about which form of data is "better." The more important task is to collect relevant data for your research. Let's consider the pros and cons of qualitative data.
Qualitative data helps offer in-depth analysis and a more nuanced understanding of phenomena than quantitative data can provide. For example, statistics can tell us the average test scores for each school whose students took a standardized test. Comparisons of average scores can give us information about which schools are more successful or are struggling. Further statistical analysis can indicate a correlation between school funding and test performance.
However, these statistics are less likely to point out the causes leading to these test results. A qualitative study gathers data that can supplement those test scores with further context, such as teachers' instructional practices, students' opinions about learning activities, and funding for educational resources.
An analysis of qualitative data can allow researchers to draw relationships between ideas. This is accomplished by "coding" the data for ideas. Coding qualitative data involves looking at your data and applying short, descriptive codes to segments of text, images, audio, or video for later analysis. If done well, you can turn your raw data into an organized, meaningful data set from which you can draw insightful conclusions.
Suppose the study above involves interviewing students about their test performance and teachers. A researcher can code all of the instances where students describe their teacher as "nice," "helpful," or "strict." Qualitative data analysis software like ATLAS.ti helps researchers with the coding process so that themes emerging from the data become easier to understand. They can then conduct qualitative data analysis by determining whether the well-performing or struggling students have a specific set of keywords that describe their teachers' personalities. If so, the researcher can propose a connection between a teacher's personality traits and their students' test scores.
The main disadvantage of using qualitative data is that the analysis can be complex and time-consuming. On the other hand, quantitative data is relatively easy to collect and analyze. Because qualitative data cannot easily be reduced to numbers or statistics, researchers need to reorganize the data in more structured and meaningful ways for analysis. A qualitative researcher often has to read their data line by line to determine what to code and how.
ATLAS.ti can help save time with various tools that help you quickly and efficiently reorganize your data. The Word Cloud function, for example, determines which words are most common in your qualitative data. For example, if you are coding for words to describe a teacher's personality, the Word Cloud function can tell you that the word "funny" appears in your qualitative data more often than you might think. Using this knowledge, you can search for paragraphs with the word "funny" and then code them easily if they relate to the teacher's personality.
Another concern is that critics of qualitative research point to researchers' potential biases and subjectivities when analyzing qualitative data. The reorganization and analysis of the data must be presented clearly and transparently so that research audiences can easily understand the analysis and find the subsequent conclusions more credible.
The research objectives you want to pursue will dictate how you should collect data and what data you should collect.
Let's say you are conducting an experimental study to determine the effectiveness of a nutritional supplement in helping people lose weight. In this case, you will likely collect quantitative data such as weight, caloric intake, and time spent exercising. Quantitative data like these can be analyzed statistically to allow you to understand if research participants are losing weight because of the supplement.
On the other hand, you may also want to gather opinions on whether people are satisfied with the supplement. Qualitative data collection methods such as interviews or focus groups might ask research participants what they think the supplement tastes like, how they feel after taking it, and why they believe it is effective or not effective. Answers to these questions don't provide easy numbers or simple statistics. Still, they are just as important to product researchers because even if the supplement is effective, people may choose not to buy it if it leads to unpleasant experiences. Qualitative data is valuable to researchers in this comparison when they need to know more about an unfamiliar phenomenon and when understanding the phenomenon requires more complexity than a simple yes/no binary or a numerical scale can provide.
You may want to consider a mixed methods approach to research that combines quantitative and qualitative data collection methods. Researchers can best understand a complex problem from various data collected on the subject. In the example above, the successful launch of a nutritional supplement depends on its effectiveness and customer satisfaction. One is only particularly helpful if the other is also present. Ultimately, it is essential to consider whether you are collecting the right kinds of data for the research inquiry you want to pursue.
A researcher can employ various qualitative research methods to collect qualitative data. As a result, numerous forms of data can be used for qualitative data analysis.
Questionnaires or surveys are among the easiest methods to collect large-scale qualitative and quantitative data. In addition to capturing quantitative data for statistical analysis, questionnaires can also be used to collect open-ended answers from respondents. For example, researchers can ask respondents to rate their satisfaction with a particular product on a scale of 1 to 5 and then write down their reasons for their ratings. Qualitative data analysis can reveal sentiments about a product among respondents who are very satisfied with it and compare sentiments among unsatisfied respondents.
Qualitative data from interviews often involve transcripts and audio or video recordings. Transcription converts interviews into text that can be read and cited in documents and presentations when you want the audience to see what research respondents have said.
Recordings are also valuable as they allow researchers to see respondents' facial expressions and gestures or hear their non-verbal utterances. This qualitative data analysis helps researchers better understand how respondents feel (e.g., excited, upset, confused) during interviews.
The best qualitative data analysis software can assist you with all major forms of data. ATLAS.ti allows researchers to look at both recordings and their transcripts in one place so that you can code and analyze the words they say along with the visual and aural cues associated with those words. This provides opportunities for more insightful qualitative research when data allows researchers to view research participants through various lenses.
Focus groups are similar to interviews except that multiple respondents talk simultaneously with the interviewer. Like with interviews, qualitative data from focus groups can be found in transcripts or multimedia recordings. The recordings can have significant value for qualitative research because they can capture how focus group respondents interact or collaborate.
An observational research method can conduct data collection on a particular social phenomenon in a less controlled environment than where interviews or focus groups would be conducted. Qualitative methods in the field can help researchers who want to see the social world outside of a confined experiment. Researchers can collect various forms of data, such as recordings or transcripts of audio or video, the observers' field notes, and photographs. The type of research you want to conduct will help you determine which data collection methods to employ.
If you are at a train station, you may want to record audio of train station announcements or record field notes about how easy or difficult it may be to navigate the station. Additionally, taking pictures or videos while walking around the train station may be valuable to later analyze what you see.
ATLAS.ti allows researchers to conduct qualitative research on all types of data, from text to video to PDFs. This flexibility allows researchers to conduct the broadest variety of qualitative data collection. As time-consuming as the qualitative research can be, taking advantage of the full array of possible data collection makes the entire endeavor worthwhile.
Any textual data, such as medical records, journal articles, and website pages, can be analyzed qualitatively. Collecting documents is useful to researchers looking to conduct a comparative analysis, a thematic review, or user research. Researchers can analyze documents for their text, their images, or other visual features depending on the inquiry they want to conduct.
As mentioned above, ATLAS.ti can conduct qualitative analysis on text-enabled and non-text-enabled PDFs. The ability to code PDFs gives researchers flexibility in collecting all kinds of data useful for data useful for their research.
Content from Twitter, Instagram, and other similar platforms can provide abundant opportunities for qualitative analysis. ATLAS.ti allows researchers to import tweets directly into their project. Researchers can conduct complex searches of Twitter data to incorporate tweets as qualitative data instantly.
Tweets also include metadata such as a user's handle, location, likes, and retweets, all of which are included in imported tweets. This is useful if, for example, you want to confine your qualitative analysis to tweets from a certain location or by a particular user.
While there are numerous approaches to analyzing qualitative data, let's examine two broad approaches you may want to consider. One involves structuring the data so that it can be analyzed quantitatively. The other reorganizes the data to identify useful themes for theoretical development.
Content analysis is useful for analyzing qualitative data for patterns of words or features of language that commonly appear. One form content analysis involves looking for common words or phrases, which ATLAS.ti can do through the Word Cloud and Concepts tools.
For a more complex content analysis, you can use ATLAS.ti tools such as Concepts or Opinion Mining. These features employ machine learning to identify sentiments that may frequently appear in your data. The search for the frequency of these patterns in your data allows for quantitative research methods to produce statistical analyses.
This form of qualitative research also looks at patterns but for the purposes of creating a hierarchical thematic structure. Researchers who analyze qualitative data this way can help create new theory rather than merely confirm existing theory. Conducting qualitative research in this manner involves looking for patterns among codes and grouping those codes under larger themes (which may also be grouped into even broader categories).
The tools in ATLAS.ti allow researchers to organize their codes in a hierarchical network, especially when related phenomena (e.g., "excitement," "happiness") can be grouped under larger themes (e.g., "positive emotions"). Using networks, ATLAS.ti lets researchers turn their qualitative data into useful theory to guide future research and practice.
Quantitative data collection lets researchers create tables and figures to provide a useful statistical analysis. Think about representations like these as ways to reduce large sets of data into compact forms that are easy for your audiences to understand quickly.
Qualitative data research has the same responsibility of representing data in a form that is accessible to audiences. Where possible, you may consider using tables for quantitative analyses of qualitative data. However, other representations are just as useful.
Quantitative data analysis makes it easy to represent certain phenomena in numbers through statistical tables. However, qualitative researchers can accomplish a similar goal by presenting the most impactful quotations from a set of interviews or focus group sessions in reports or presentations. As a result, text excerpts from qualitative data research are a common use of the data collected. Directly reporting what a person says lends a certain credibility to a report or presentation.
ATLAS.ti lets researchers easily export sets of quotations based on the criteria they define (e.g., selected codes, documents) into Microsoft Excel. Rather than look at long documents or large numbers of documents one at a time, you can view all the text that belongs to a certain code in one easy spreadsheet. The resulting quotes in the exported spreadsheet can then be used in research reports or presentations.
Especially when you are seeking to define or expand a theory, a visualization of that theory makes it clear to your audience what you are trying to explain. Think about a casual "theory" for how people decide where they should go on vacation. They may think about the cost of the trip, the places to visit, and the food to eat when they get there. All of these individual items fall under the larger theme of "vacations." You can then visualize this in a code tree of sorts.
Networks in ATLAS.ti help to visualize these code trees. By creating codes for these aspects that you find in your data, you can form and organize networks that visualize a hierarchy representing your qualitative data in the form of a working theory.
Researchers should always be careful with collecting and handling qualitative data, especially if it contains personal information or is obtained without consent. People's perspectives can often be reduced to numbers if quantitative analysis is possible but qualitative data collection often preserves the words, circumstances, and behaviors of people who may feel uncomfortable with how such data might be used. The main consideration is how the researcher should prevent presenting the data to their audiences in a way that provides unwelcome clues about participants' identities.
Let's think more specifically about the different concerns and proper practices associated with qualitative data.
Medical records are especially sensitive as people can connect names to health conditions that patients may prefer to keep secret. In observations, people may not want their pictures taken if they don't want to be associated with being in a particular place. Respondents in interviews and focus groups may choose to withdraw from research after they say something they think might be embarrassing or uncomfortable if their words are used later.
Before collecting any data, researchers should obtain informed consent from participants to ensure that they understand their rights and how their privacy is protected. This might be a challenge in observations, especially when they occur outside a controlled environment. When collecting data in the field, researchers might consider avoiding taking pictures or videos of people's faces or recognizable clothing or possessions. As a result, field notes might be the most appropriate form of data to collect while observing in the field.
Even if research participants give informed consent, there is always the possibility they might say something sensitive or unintentionally provide identifying information. A respondent to a paper questionnaire might write their name on a page by accident. A focus group member might end up saying another member's name during the course of a discussion. A researcher might leave a video camera recording an observation idle for a second, at which time someone walks into view by accident.
There are also ethics involved with how the data is analyzed. A researcher may look at an interview respondent's utterance and interpret it in a way that may be different from what the respondent intended. While there is flexibility in qualitative research methods regarding researcher biases, it's the researcher's responsibility to categorize qualitative data to reflect research participants' perspectives and utterances as accurately as possible.
As a result, researchers should conduct an analysis process called member checking when and where possible. This involves going back to the research participants and confirming what they said and if your interpretation of what they said is accurate. This member checking is important for data validation in interviews and focus groups, particularly if your research question explores the opinions of research participants.
Finally, how you represent the data should reflect as accurately as possible the actual data collected while also protecting the privacy of research participants.
Just as member checking is useful in data analysis, it is also important to have participants confirm that what you write about them is accurate and respectful of their perspectives and opinions. The necessary thing to keep in mind is that participants remain involved throughout the research process.
Visual data can help give away a research participant's identity. The sound of someone's voice or nearby landmarks can provide hints about who a person is and where they live or work. Researchers should consider if they are using segments of audio or parts of images that are especially distinguishable to the extent that your research audience can make unwanted inferences from the data when the focus should be on the research inquiry.