Statistical Analysis - Concerned with the development and application of methods and techniques for organizing and analyzing data so that the reliability of conclusions based on the data may be evaluated objectively in terms of probability. (Powell pg.261) There are two different types of statistical analysis: theoretical which deals with mathematical aspects of statistics, and applied statistics deals with practical applications of statistics.
Steps involved in statistical analysis (Powell pg. 263-264):
1. Establishing of categories
- The set of categories for any one variable should be derived from a single classifactory principle, which is determined by the research question or hypothesis being investigated.
- Each set of categories should be exhaustive.
- The categories within each set should be mutually exclusive.
- The development of categories should be based on a sound knowledge of the subject matter and an anticipation of likely responses.
2. Coding the data
- Once a category has been established and data assigned to them, it is necessary to convert the new data or responses to numerical code, so that they can be tallied.
- Coding needs to be reliable and consistent.
3. Analyzing the data
- Decide whether to use descriptive statistics, inferential statistics or both.
4. Cautions in testing the hypothesis
- Statistical inferences are based on probability, and one can never rely on statistical evidence alone for a judgment of whether a hypothesis is true. It is also important to remember that a single statistical acceptance of a hypothesis does not prove it to be true with absolute certainty.
Wildemuth:
Content Analysis - the systematic, objective, quantitative, analysis of message characteristics. (Neundorf 2002) The "message" refers to information that travels from source to destination. Content analysis was originally developed to analyze texts such as journal articles, newspapers, books, responses to questionnaires, and transcribed interviews (Wildemuth pg.297).
Identifying the units for analysis - There are two basic types of units of content to define after you have chosen a research question: sampling units and recording units (Riffe 2005). Recording units are the elements of content that are coded (Wildemuth pg. 299). The types of recording units are: physical, conceptual, and temporal. Sampling units are usually collected from the overall population of the text or other media of interest.
Qualitative analysis - A research method for the subjective interpretation of the content of text data through the systematic classification process of coding and identifying themes or patterns (Hsieh & Shannon 2005). Qualitative content analysis goes beyond merely counting words or extracting objective content from texts to examine meanings, themes, and patterns that may be manifest or latent in a particular text (Wildemuth and Zhang pg. 308).
Discourse analysis - The analysis of discourse. Tends to focus on either the particular types of conversations (the reference interview) or formal texts (professional literature). Such units of information comprise social texts (expressions of our society) and function to support interpersonal relationships, institutions and ideologies within that society (Wildemuth and Perryman pg. 320).
Analytic induction - A specific form of inductive reasoning used to analyze qualitative data. It is a formalized method for developing and refining a theory or hypothesis, directly from the data (Wildemuth and Spurgin pg. 329). The hypothesis and definitions must apply to all cases of your sample.
Variable - A property of an object, person, or event that can take on different values (Howell 2004). They can also be defined and operationalized at different levels of measurement. Nominal variables are those whose possible values are categories, with no true numerical value that can be assigned to them (Wildemuth pg. 339). Ordinal variables are those for which the values can be rank ordered (Bernard 2000). Ratio-level variables are ordered and have values at equal intervals, they have a true zero point (Bernard 2000).
No comments:
Post a Comment