Steps

1. Define and limit the population – what are you studying? People? Assignments? Courses?

2. Select coding units: Course? Answer? Conversational turn? Test items?

3. Determine classification systems for study: What categories will each unit need to be coded into – inductively or deductively? (Hint: This is an iterative process J)

• Exhaustive – every unit should fit into a category (including “other”)
• Mutually exclusive – every unit should only fit into ONE category
• Coding rules/manual – how to determine which category fits

Example (1):

Use of fonts emoticons (coding online courses for nonverbal immediacy)

Unit: entire class: Categorized based on the majority of pages, headings, documents, and announcements

Categories:  Use of “fun” fonts: less formal and business like; use of emoticons, manipulation of punctuation to indicate emotions (“interesting . . .”; not emphasis for assignments etc.), so multiple exclamations count!!! (single do not); nonverbal vocalizations (“hmmm . . .,  soooo?”); and/or emoticons (as symbols J or created ;).

            Emotion expressive: Often uses fun fonts and/or other nonverbal emotive indicators such as, emoticons, vocalizations etc.(more than half the pages/announcements etc. use some form of nonverbal emotive indicators)

            Emotion low: Sporadic (not consistent) uses of fun fonts and/or uses some other nonverbal emotive indicators
            
            Emotionless: No use of fun fonts or other nonverbal emotive indicators

4. Sample messages: Get your sample

5. Code messages: At least two coders for all or part of the data – determine interrater reliability (Scott’s pi, percent agreement, etc.  - see resources).

6. Analyze the data

Resources:

Content analysis methods guides:

Overview of content analysis: https://writing.colostate.edu/guides/guide.cfm?guideid=61

Flow chart of content analysis process:  http://academic.csuohio.edu/kneuendorf/c63311/

Qualitative content analysis description:
http://www.qualitative-research.net/index.php/fqs/article/view/1089/2385%3E

Calculating interrater reliability:

Online tool to calculate reliability: http://dfreelon.org/utils/recalfront/

*Source for general information: Reinard, J.C. (2008). Introduction to Communication Research, 4th Edition. New York: McGraw-Hill.

(1) Dixson, M.D., Greenwell, M., Lauer, S., Rogers-Stacy, C., & Weister, T. (2014). Nonverbal immediacy behaviors and online student engagement: Bringing past instructional research into the present virtual classroom. Unpublished ms.

Dixson, Mack @ FACET, 2014

Other Web-based Resources

Overview of content analysis including calculating reliability

Qualitative content analysis, unlike quantitative, does not seek to count or categorize, necessarily, but to pull out recurring themes and trends from conversations, speeches/lectures, or other forms of "text." It is often an iterative process of working through the data multiple times to interpret messages.

University of Leicester - 10 concise steps to qualitative research

Indepth description of content analysis from Institute for Qualitative Research

The use of qualitative content analysis in case student research from Forum for Qualitative Research