The term data refers to a collection of information. A more detailed definition includes types of data that combine to be the collected information such as numbers, words, pictures, video, audio, and concepts. Many definitions of data include the word fact, or facts, but this implies an inference about the data and not the data themselves. This occurs more often in the physical sciences. One may also see the word raw as a descriptor of the data. This description is used to separate data such as the number 42 from information such as at 42 one is 5 years older than one's sibling. Once data are gathered, they are typically put into a format that can be analyzed by machine or human. The format can be a spreadsheet, notecards, or literary analysis software and serves to increase the ease of data analysis.
In the generic split of quantitative versus qualitative research, quantitative research gathers data that are in numerical form. The original data can be in nonnumerical form such as statements that are recoded on some specific numerical scale. Quantitative data separate into categories based on their measurement type—nominal (e.g., gender), ordinal (e.g., law school class rank), interval (e.g., degrees Fahrenheit), and ratio (e.g., degrees Kelvin).
Qualitative data are generally nonnumerical but have a greater variety of sources. Those data sources are generally categorized as verbal and nonverbal. [Page 186]Data are verbal if the majority of what is being analyzed is words. Verbal data sources include items such as personal diaries, letters, media reports, surveys/interviews, and fieldnotes. Within the group of interviews the data can come from in-depth/unstructured interviews, semi-structured interviews, structured interviews, questionnaires containing substantial open-ended comments, focus groups, and so on.
Nonverbal data sources include items such as student concept maps, kinship diagrams, pictures, video, film, art, and print advertisements. Each type of data and how it was collected has different strengths and weaknesses in relation to the research questions and analysis techniques. For example, nonparticipant observations from video collected through surveillance cameras potentially allow the researcher to collect data without influence in the field, but there are issues with the ethics of these observations.