This sketch is intended to provide a step-by-step procedure of corpus-based analysis of a syntactic construction using part-of-speech tagged corpora.
The main stages:
- Initial stage
- Retrieval of a target construction
- Categorization of retrieved instances
- Quantitative and qualitative analysis
1. Initial stage:
- Literature review on the research problem: up-to-date state of affairs
- Framing a research question and working definitions
- Preparing a retrieval:
- Defining the information to be extracted from corpora
- Deciding how many corpus instances need to be coded
- Choosing annotated corpora and corresponding software
- Defining corpus query syntax
2. Retrieval of a target construction:
- Design of a grammatical query using part-of-speech tags
- Training and testing of the query
- Creating an initial query in order to identify possible patterns
- Developing a prototype query according to the research question
- Training the query in order to identify invalid patterns
- Refining the query
- Creating the final query
- Testing the query: estimating a precision rate and a recall rate of the query
- Performing the final retrieval with the final query
3. Categorization:
- Selecting the categories for analysis
- Establishing operational definitions of the relevant variables
- Choosing approach and software:
- in-corpus-tool approach: concordancers (WordSmith, Monoconc, BNCweb, etc.)
- database approach:
- database programs (Filemaker, Microsoft Access, Microsoft Excel, etc.)
- statistical analysis packages (SPSS, Minitab, etc.)
- Classifying the retrieved instances according to the established categories and variables
4. Quantitative and qualitative analysis
Because, in their chapters, the authors Hoffmann and Smith & Seoane demonstrate and discuss only the first three stages – initial stage, retrieval and categorization of a target structure - no further information will be provided on quantitative and qualitative analysis at this point.
Created with the Personal Edition of HelpNDoc: Write EPub books for the iPad