The qualitative data collection process may be assessed through two different points of view—that of the questionnaire and the respondents. A respondent may not care about the classification of data he/she is inputting, but this information is important to the questionnaire as it helps to determine the method of analysis that will be used.
There are different methods of analysis that vary according to the type of data we are investigating. In statistics, there are two main types of data, namely; quantitative data and qualitative data.
For the sake of this article, we will be considering one of these two, which is the qualitative data.
Qualitative data is a type of data that describes a piece of information. It is investigative and also often open-ended, allowing respondents to fully express themselves.
Also known as categorical data, this data type isn’t necessarily measured using numbers but rather categorized based on properties, attributes, labels, and other identifiers. Numbers like national identification number, phone number, etc. are however regarded as qualitative data because they are categorical and unique to one individual.
Qualitative Data can be divided into two types namely; Nominal and Ordinal Data
In statistics, nominal data (also known as nominal scale) is a classification of categorical variables, that do not provide any quantitative value. It is sometimes referred to as labeled or named data.
Coined from the Latin nomenclature “Nomen” (meaning name), it is used to label or name variables without providing any quantitative value. This is not true in some cases where nominal data takes a quantitative value.
However, this quantitative value lacks numeric characteristics. Unlike, interval or ratio data, nominal data cannot be manipulated using available mathematical operators.
For example, a researcher may need to generate a database of the phone numbers and location of a certain number of people. An online survey may be conducted using a closed open-ended question.
E.g: Enter your phone number with country code.
The best way to collect this data will be through closed open-ended options. The country code will be a closed input option, while the phone number will be open.
Ordinal data is a type of qualitative data where the variables have natural, ordered categories and the distances between the categories are not known. For example, ordinal data is said to have been collected when a customer inputs his/her satisfaction on the variable scale — "satisfied, indifferent, dissatisfied".
Thus, ordinal data is a collection of ordinal variables. For example, the data collected from asking a question with a Likert scale is ordinal.
An organization creates an employee exit questionnaire that primarily highlights this question: “How will you rate your experience working with us?”
Other examples of ordinal data include the severity of a software bug (critical, high, medium, low), fastness of a runner, hotness of food, etc.
In some cases, ordinal data is classified as a quantitative data type or said to be in between qualitative and quantitative. This is because ordinal data exhibit both quantitative and qualitative characteristics.
Various Qualitative data examples are applied in both research and statistics. These examples vary and will, therefore, be separately highlighted below.
1. Qualification: When filling job application forms, the applicant is usually required to fill in his/her qualification. This data can be collected in different ways — open-ended or closed questions.
a. Open-ended question approach: What is your highest qualification? _____
b. Closed-ended Question approach
What is your highest qualification?
2. Competitive analysis: During competitive analysis, brands send out questionnaires to their target market to access the popularity of their competition.
For example: Which of the following payment platforms are you familiar with?
They may even take it further by asking questions like, "How did you hear about them?". This may even help them improve their marketing strategy.
3. Surveys or Questionnaires: Researchers use surveys and questionnaires to carry out investigations and collect data. Below is an example of a questionnaire that collects nominal data.
Where is your country of residence? _____
4. Bug severity: When testing for bugs on a website or software, security researchers also check for bug severity. The extent to which a bug can cause damage is what is termed its severity.
The severity of a bug may be said to be critical, high, medium, or low. This data can be collected on either a nominal or ordinal scale.
5. Likert scale: A Likert scale is a point scale used by researchers to take surveys and get people's opinions on a subject matter. Consider this example:
How will you rate the new menu?
This is a 5 point Likert scale, a common example of ordinal data.
During the voting process, we take nominal data of the candidate a voter is voting for. The frequency of votes incurred by each candidate is measured, and the candidate with the highest number of votes is made the winner. In statistical terms, we call this mode.
Each embassy in every country has a database of the immigrants coming into the country. For example, the Nigerian embassy in the US has a database of all the legal African migrants to America. This way, the US Government will have an estimate of the population of Africans in the US. Not only that but also personal details like gender, countries, etc. that may help in proper statistics.
During an event, organizers take nominal data of attendees, which include name, sex, phone number, etc. An example question like "Where did you hear about this event? " may help them determine which is the most effective marketing platform.
When trying to build a database of people with diverse backgrounds like different genders, races, classes, etc. we use qualitative data. For example, when employing people, organizations that care about having equal female representation take statistics of the number of male and female employees to balance gender.
Ordinal data is a data type that has a scale or order to it. This order is used to calculate the midpoint of a set of qualitative data.
For example, qualitative data on the order of arrangement of goods in a supermarket will help us determine the goods at the center of the supermarket. This may even be a factor in determining whether the position of good influences the number of sales.
1. Types: Qualitative data is of two types, namely; nominal data and ordinal data.
2. Numeric Values: Qualitative data sometimes takes up numeric values but doesn't have numeric properties. This is a common case in ordinal data.
3. Order: Ordinal data have a scale and order to it. However, this scale does not have a standard measurement.
4. Analysis: Qualitative data is analyzed using frequency, mode and median distributions, where nominal data is analyzed with mode while ordinal data uses both.
When collecting qualitative data, researchers are interested in how, i.e., specific details around the occurrence of an event, with a particular interest in the perspective of the subject of study. Some of the techniques used in collecting qualitative data are explained below:
This is the process of studying a subject for a given period to access some information. This may be done with or without consent of the subject that is being observed.
Observation may be done in several ways. It is not necessarily done by looking at the subject for a long period. It may be through reading materials written by or about the subject, stalking on social media, asking about them, etc.
An interview means a one-on-one conversation between two groups of people where one part interrogates the other party. The word group is being used because at times we may have two or more interviewers and two or more interviewees.
In recent times, we now have phone interviews and Skype (video) interviews. The subject may be interviewed to collect qualitative data directly from them.
This is a very common technique for collecting qualitative data from a group of respondents. Traditional questionnaires are printed on paper and given to the respondents to fill and handed back to the researcher.
Researchers can now create online surveys and send them to respondents to fill. This is better than the traditional method because it automatically collects the data and prepares for analysis.
Quantitative data analysis is the process of moving from the qualitative data collected into some form of explanation or interpretation of the subject under investigation. There are two main stages of qualitative data analysis.
This is the first stage of qualitative data analysis, where raw data is converted into something meaningful and readable. This is done in four steps:
Coding is a major step in analyzing qualitative data. It is the process of classifying data by grouping them into meaningful categories to easily analyze them.
Things to note when developing codes:
Closely review the developed categories and use them to code your data. Having teamwork on data coding will accommodate different perspectives.
Don't be afraid to include or remove subcategories as you move on. This may turn out to be needed in the case that the codes are too broad or too detailed.
The Coding Process
This is the point where you take a break from the hard work. Step back and observe the coded data for emerging themes, patterns, and relationships.
Here, you check for similarities and differences and see what each group is depicting.
This is the process of streamlining the remaining chunk of data and keeping it brief. All parts of the data should be summarised to get them ready for analysis.
After completing the first stage, the data is ready for analysis. There are two main data analysis approaches used, namely; deductive and inductive approaches.
The deductive approach to qualitative data analysis is the process of analysis that is based on an existing structure or hypothesis. Researchers pick an interesting social theory and test its implications with data.
This approach is fairly easy since the researcher already has an idea about the likely results of the analysis before conducting the research. It is usually associated with scientific investigations.
The inductive approach to qualitative data analysis is the process of developing a new theory or hypothesis for data analysis. Researchers find themes, patterns, and relationships in the data and work to develop a theory that can explain them.
This is a more difficult and time-consuming approach compared to the former.
By collecting qualitative data using Formplus, researchers have access to tools that will make their research simple and easy. Data analysis is made easy with an efficient data collection tool that records real-time data.
This online form builder helps businesses conduct better customer satisfaction surveys with qualitative data. The data collected through Formplus can be exported in different formats that are compatible with data analysis tools.
Formplus eliminates stress and needs for manpower that may arise from dealing with qualitative data. No matter how big the sample size is, Formplus makes collection easy for both respondents and questionnaires.
To collect qualitative data using Formplus builder, follow these steps:
Formplus gives you a free plan where you can create basic forms and surveys. To collect vast qualitative data from premium online surveys, you can subscribe to a paid plan for as low as $20 monthly, with reasonable discounts for Education and Non-Governmental Organizations.
We will be creating a sample qualitative data collection form that takes the name (nominal data) and bug severity level (ordinal data) from a respondent.
Qualitative data in statistics is similar to nouns and adjectives in the English language, where nominal data is the noun while ordinal data is the adjective. This comparison is an attempt towards breaking down the meaning of qualitative data into relatable terms for proper understanding.
A proper understanding of what qualitative data is aids researchers in identifying them, using them, and choosing the best method of analysis for them.
In qualitative data analysis, we break data into smaller bits and group them before analysis. This is to properly understand the data and ease the analysis process.
You may also like:
Coefficient of variation is an important concept that allows you to predict variables within and outside data sets. While it has its roots ...
Aside from consulting the primary origin or source, data can also be collected through a third party, a process common with secondary data. ...
Research and statistics are two important things that are not mutually exclusive as they go hand in hand in most cases. The role of ...
The underlying need for Data collection is to capture quality evidence that seeks to answer all the questions that have been posed. Through ...