Project Update #2

Applying Qualitative Content Analysis to Study Online Communities

In this update, basic ideas about qualitative content analysis are reviewed and a bibliographic list is provided related to application of qualitative content analysis in online community studies.

Introduction of Qualitative Content Analysis

• Concept
  • Content analysis is a research technique for making replicable and valid inferences from texts (or other meaningful matter) to the contexts of their use (Krippendorff, 2012, p. 24). In other words, it is a method to understand the manifest and latent content of the texts, images, moving images, or other meaningful materials by categorizing, evaluating, or tabulating symbols in materials. The aim is to figure out meanings or probable effect behind the texts.
  • From communication perspective, Weber (1990, p. 9) pinpointed that valid inferences from text includes senders of messages, audience of messages, and message itself.
• Qualitative Content Analysis
  • The literature yields both quantitative and qualitative approaches of content analysis.
  • Quantitative approach, first brought up by Lasswell (1938, 1943, & 1952), focused on word frequency, space measurements, time counts (for video or television time) and keyword frequency. After reducing data into numbers and frequencies, statistical analysis will be applied to describe the phenomenon or test some hypotheses. The assumption is that words and phrases mentioned most are those reflecting important concerns in every communication.
  • By overcoming the rigorous nature and superficial analysis of quantitative content analysis (Kracauer, 1952), qualitative content analysis focuses on informal content and comprehensive understanding of a text (Hsieh & Shannon, 2005).
  • The nature of qualitative content analysis is to cope with interpretation problem of textual matters, which is rooted in hermeneutics (Janzuihof, 2013). According to Schleiermacher (1998), interpretation is the art of individuality. Each text is determined by individual historical context and its author. Content dependency makes each single “word” in hermeneutics has no permanent meaning, since factors of context cannot be fully understood. Therefore, full objective understanding can therefore never be achieved.
  • According to Krippendorff (2012, p. 32), qualitative content analysis has following characteristics:
    • Requires close reading of relatively small amounts of textual matter
    • Interpretation of given texts into new narratives, which are accepted within particular scholar communities
    • The systematic interpretation is required, and analysts should acknowledge that their own socially and culturally understandings constitutively participate in the interpretation.
  • Finding themes or systematic categories are the central task of qualitative content analysis (Krippendorff, 1980, p.76).
  • Mayring (2000) provided four basic rules about qualitative content analysis:
    • Fit the materials into a model of communication: analysts should be aware which aspect they’d like to focus on the inferences: to aspects of communicator (their feelings, experiences or opinions), situation of text production, to the social-cultural background, text itself, or the effect of the message.
    • Step by step rule: a systematic procedure should be followed to devise the material into content analytical units.
    • Category is the center of analysis, the aspects of text interpretation will be put into categories by following research questions, which is carefully founded and revised within the process of analysis
    • Reliability and validity concern: reliability will be maintained by comparing with other studies and inter-coder reliability. Validity will be guaranteed by trained project team members.
    • Three Approaches of Qualitative Content Analysis
      • Inductive and deductive (Graneheim & Lundman, 2004; Mayring, 2000)
        • Inductive approach starts from category criterion definition, which derives from theoretical background and research question. Following the derived criterion, texts will be worked through and categories will be tentatively step-by-step achieved. Feedback loop will be applied to revise and evaluate derived categories and their reliability.
        • Deductive approach starts from theoretical assumptions of categories from theories, and work through the texts to see if they fit the derived theoretical categories.
      • Conventional, directed, or summative (Hsieh & Shannon, 2005).
        • The three approaches differ in their ways to extract coding schemes and final coding categories
        • In conventional content analysis, researchers avoid using categories in previous theories or studies. Instead, they allow categories flow totally from the text. It requires researchers to emerge themselves into the data until insights come out.
        • Directed content analysis is based on previous research or theory, which is regarded as deductive approach in Marying’s classification
        • Summative approach is to understand contextual use of words or content frequencies. This approach usually starts from counting words or interested contents, and then interprets the context of those words (why they happen in certain ways), or evaluate quality of the content.
      • Discourse analysis, rhetorical analysis, ethnographic content analysis, and conversation analysis (Krippendorff, 2012)
        • Discourse analysis tends to focus on how particular phenomena are presented through texts. For example, how movies and TV shows manifest the stereotypical image of librarians.
        • Rhetorical analysis focuses on how message are delivered and with what effects. Structural elements, tropes, styles of argumentation, speech acts, and the like are researchers’ focus in analysis. Efforts to study negotiations to see what works and what doesn’t work is one example of this.
        • Ethnographic content analysis is an approach to encourage analysis emerged from the text. It doesn’t only work with categories and their narrative descriptions, but also deals with situations, settings, styles, images, meanings, and nuances presumed to be recognizable by the participants
        • Conversation analysis tends to analyze transcripts as records of conversational moves toward a collaborative construction of conversations.
      • Procedures of Qualitative Content Analysis
        • Zhang and Wildemuth (2009) proposed 8 steps of qualitative content analysis in library and information science from a particularly inductive approach with a focus of finding out themes and categories:
          • Prepare the data, for instance, transcribing interviews to analyzed data
          • Define the unit of analysis. Usually, individual themes rather than a sentence or paragraph is a unit for analysis.
          • Develop categories and a coding scheme, both inductive and deductive approaches can be applied. Constant comparative method will be applied in the inductive approach.
          • Test the coding scheme on a sample of text
          • Code all the text.
          • Assess coding consistency: usually, inter-coder consistency will be assessed.
          • Draw conclusions from the coded data
          • Report methods and findings