Back to Search
Start Over
Characters as groups: a new approach to morphological characters in phylogenetic analysis
- Source :
- TAXON. 56:479-492
- Publication Year :
- 2007
- Publisher :
- Wiley, 2007.
-
Abstract
- A new method for working with morphological characters is described and explored in experiments using human participants. The method uses direct comparison and sorting of images to produce hierarchical character-cladograms. A character-cladogram is a graphical representation of a single character that serves as a hypothesis of phylogeny based on that character. Each dichotomy in the character-cladogram represents a character state. Character states are unnamed, thus avoiding problems that arise through the application of verbal labels. Experiments with human participants are used to explore the conditions under which direct comparison produces reliable (consistent from investigator to investigator) and valid (in agreement with an independent estimate of phylogeny) characters. Participants were drawn from students taking a course in plant diversity at UNC Greensboro, and professional plant morphologists attending the Botany 2004 meetings. The students were randomly assigned to trained and untrained groups. Training was carried out using a method that has been shown to change a participant‘s mode of visual processing from analytic (the mode used by visual novices) to holistic (an additional mode only employed by visual experts). Morphologists (no specialists of the taxonomic group) were included in the study to investigate the effects of disciplinary expertise on the ability to describe character-cladograms. They received no additional training beyond that available to them as disciplinary experts. The results suggest an improvement in both reliability and validity after the training regime. We found no support for the idea that the morphologists differed from untrained students in their ability to produce reliable or valid character-cladograms. Disciplinary expertise may not translate into the ability to make reliable and valid assessments of similarity in an unfamiliar visual domain. Based on these results, we suggest a method for creating morphological characters and character states.
- Subjects :
- business.industry
Sorting
Plant Science
computer.software_genre
Domain (software engineering)
Visual processing
Character (mathematics)
Similarity (psychology)
Artificial intelligence
Representation (mathematics)
Psychology
business
computer
Discipline
Ecology, Evolution, Behavior and Systematics
Natural language processing
Reliability (statistics)
Subjects
Details
- ISSN :
- 00400262
- Volume :
- 56
- Database :
- OpenAIRE
- Journal :
- TAXON
- Accession number :
- edsair.doi...........8d363885197f072e163901f1eb391da3