Measuring complexity 4. Exercise
1. Review the inspection results
Review the Inspection activities from L2SCA and LxGrTgr. Use your notes and the output files to answer the following questions.
-
Which L2SCA index did you inspect? Define it and state whether your manual calculation matched the automatic result.
-
Which LxGrTgr tag did you inspect? Define the structure it represents and identify any missing or unexpected tags.
-
Compare the two tools. What does each tool measure, and what kind of output does each produce?
-
Based on your inspection, what are the strengths and limitations of automated complexity analysis? Give at least one example from each tool.
-
Write a short summary that explains why researchers should inspect automated annotations before using the results in an analysis.
2. Explore proficiency and complexity
Analyze the texts in sample_proficiency with L2SCA and/or LxGrTgr. Choose one or two indices or tags that you expect to be related to writing proficiency.
-
State your prediction. For example, do you expect higher-proficiency writers to use more complex nominals per clause, longer T-units, or more instances of a particular LxGrTgr tag? Explain why.
-
Compare texts at different proficiency levels.
-
Describe the relationship. Does the selected feature appear more often, less often, or in a different form across proficiency levels?
-
Evaluate the evidence. Explain whether your observations support your prediction. Consider alternative explanations such as text length, genre, topic, or individual writing style.