Qualitative Content Analysis
Aim: describe a phenomenon
Situation to Apply: When the existing theory or research literature on a phenomenon is limited.
Collect data through interviews, and use open-ended questions.
Read all data repeatedly to achieve immersion and obtain a sense of the whole.
Highlight the exact words that capture key thoughts or concepts, and take notes.
Initial codes emerge.
Store codes in categories.
Organize codes into clusters.
Develop a tree diagram to organize categories into a hierarchical structure with definitions of every part.
Advantages: gaining direct information from study participants without imposing preconceived categories or theoretical perspectives.
Limits: Fail to develop a complete understanding of the context, thus failing to identify key categories.
Aim: validate or extend conceptually a theoretical framework or theory.
Situation to Apply: existing theory or prior research is incomplete or needs further description.
Procedure (More structure than Conventional Approach):
Identify key concepts or variables as initial coding categories.
Define each category using existing categories.
Coding with two strategies
When identifying all instances of a phenomenon, read the transcript and highlight all text that, on first impression, appears to present an emotional reaction. Then code all highlighted passages using the predetermined codes, then give new codes to text not categorized by the initial coding scheme.
Code the data with predetermined code immediately and then analyze the data that was not coded by initial codes as they may represent a new category or a subcategory of an existing code.
Advantages: In the direct approach of content analysis, existing theory can be supported and extended.
Researchers approach the data with a stronger bias, they are more likely to find supportive evidence rather than non-supportive evidence.
Participants might get cues to answer.
Overemphasis on the theory can blind researchers to contextual aspects of the phenomenon.
Aim: explore usage, not infer meaning.
Situation to Apply: Identify and quantify certain words or content when researchers want to understand the contextual use of the words or content.
Search for occurrences of the identified words by hand or by computer, including word frequency and source or speaker.
Interpret the context associated with the use of the word or phrase.
Advantages: It is an unobtrusive and non-reactive way to study the phenomenon of interest.
Limits: Inattention to the broader meanings present in the data.
Conventional Content Analysis:
The study starts with Observation
Codes are defined during data analysis
Codes are derived from the data
Directed Content Analysis:
The study starts with Theory
Codes are defined before and during data analysis
Codes are derived from existing theory or relevant research findings
Summative Content Analysis:
The study starts with Keywords
Keywords are identified before and during data analysis
Keywords are derived from interest of researchers or reviews of literature
Hsieh, H.-F. and Shannon, S.E. Three approaches to qualitative content analysis. Qualitative health research 15, 9 (2005), 1277–88
Quantitative Data Analysis
Quantitative analysis involves a large variety of statistical tests. Knowing which test to use depends on what type of data you have collected (source):
Nominal: You can categorize your data by labeling them in mutually exclusive groups, but there is no order between the categories. For example, place of birth, gender, ethnicity, brands, or game genres.
Ordinal: You can categorize and rank your data in order, but you cannot say anything about the intervals between the rankings. For example, beginner, intermediate, advanced
Interval: You can categorize, rank, and infer equal intervals between neighboring data points, but there is no true zero point. For example, temperatures or test scores
Ratio: You can categorize, rank, and infer equal intervals between neighboring data points, and there is a true zero point. For example, height, weight, or age.
What Descriptive statistics you can use are determined by your data type:
Nominal: You can only calculate the mode (most common)
Ordinal: You can calculate the mode, median, range, and interquartile range
Interval: You can calculate the mode, median, range, mean, interquartile range, standard deviation, and variance
Ratio: You can calculate the mode, median, mean, range, interquartile range, standard deviation, and variance
What statistical tests you use will be determined by your data type and your hypotheses. Here are some common calculations:
T-tests: For comparing 2 samples of numerical (interval or ratio and sometimes ordinal) data. Use paired when the data being compared came from the same people (within-subjects) and independent when it came from different (between subjects)
ANOVA: For comparing 3 or more samples of numerical (interval or ratio and sometimes ordinal) data
Chi-Squared: For comparing 2 or more samples of categorical (nominal or sometimes ordinal) data
There are also equivalents for non-parametric (non-normalized) data (advice on how to check for normal distribution here):
Wilcoxon Signed-rank test: Use in place of paired sample t-test
Wilcoxon Rank-Sum test: Use in place of independent sample t-test
Kruskal–Wallis H: Use in place of ANOVA
Instructions and tools for how to perform these tests are available online. You can use spreadsheets, code, or apps.