Research Reports
Featured Research
Discover IA industry trends, software and education patterns among IAs, including our latest surveys:
Competency Survey, 2004
This survey was conducted July, 2004 to learn what information architecture competencies practitioners thought they should have and students should learn. Members of sigia-l and the Information Architecture Institute were invited to participate.
Download the Results
Numerical analysis: Excel | Acrobat PDF
Graphs: Excel | Acrobat PDF
The most frequently cited competencies were:
- Navigation Systems
- Labeling Systems
- Blueprints & Wireframes
- Collaboration Skills
- Client Relationship Skills
- Audience Modeling
- Behavior Modeling
- User Interface Design
- Interaction Design
- Search Engine Technology
- Metadata
- Indexing
- Search Systems
- Conceptual Maps
- Content Analysis
- Content Strategy
- Writing
- Ontologies
Data Analysis Notes
A total of 166 responses were collected, though some were incomplete and 60 were used for analysis purposes.
Correlations: Not surprisingly, everyone thinks that students should learn what they themselves know. The overall correlation between Self and Student Ratings is high (.85).
Note the difference between self and student ratings. When the difference is positive, it means that this is a topic IA’s want students to know more than they themselves know. A negative difference means the opposite. The highest positive differences are for Cognitive Psychology, Automated Classification, and Statistics.
Cluster Analysis
The dendogram is based on Hierarchical Cluster Analysis and shows how the IA skills can be clustered.
The MDS map simultaneously maps the data in the main two dimensions. The map uses only self ratings. As regards the axis, the analysis extracts those axis - they are not directly self or student ratings. The axis in MDS analysis are interpreted (you need to make sense what the axis are - the analysis can give you the axis, but cannot tell you what the axis are. This is similar to the clusters in cluster analysis - the clusters need to be interpreted.
Importance for each topic is shown through the size of the circle. The space can be broadly grouped into 5 groups: Content, Design, Users, Programming (or Technology) and Business. Some of the items do not fit.
While attribute data can be manipulated to show some trends, it is not as powerful as similarity data. MDS maps and cluster diagrams created through similarity ratings (such as through card-sorting) often show a clearer picture.)
Thank you to Victor Lombardi and Scott Robinson for help compiling the survey and Rashmi Sinha for the analysis.
This page was last modified on April 14, 2005 05:10 PM.