The current NSF funded research project that I am working on for the Human Computer Interaction Institute at CMU in their Learning Technologies center focuses on redesigning a learning platform that aims to teach macroinvertebrate anatomy for amateur entomologists, facilitate trainers who teach citizen scientists, and acts as a resource and for expert entomologists in their insect identification process. Having done experiment design, data collection, analysis, synthesis, and low level prototyping\architecting, we are now in the phase of preparing to conduct workshops with stakeholders to vet design ideas, after which we will move into higher level prototypes.

Stakeholder Research

I was brought into the project to assist with designing a range of studies with stakeholders, conducting them and collecting data, and collating and analyzing it. Working with a co-investigator with a background in learning science, we wrote several protocols, and conducted 8 cognitive task analysis studies and touchstone tours with domain experts, and 13 remote interviews and usability studies using the current website with trainers. The object of the expert studies was to get a sense of how different experts used different approaches to identify insects, understand their tools and ID setup, what kinds of references, indexes and guides they used, and what they found challenging about the ID process. The trainer studies sought to understand how trainers used different training methodologies, how they incorporated different expert guides and reference materials into their teaching, and how they used the current website.

Analysis, Synthesis, and Ideation

The data revealed a number of insights crucial to the design of the platform. For example, the data revealed a wealth of knowledge about how experts and amateurs used specific terminologies, used scientific guides and keys to move through insect taxonomies and narrow down search spaces in order to correctly identify insects, how they prepared their work-spaces and tools, modifying and customizing their materials and reference books, the roles that different kinds of scientific illustrations, photographs and other graphical materials played, and what they found challenging about the current website.

Some of the key insights we derived:

• Most experts prefer to use common or colloquial terms in place of scientific
  terminologies,
• Less experienced experts tend to go back and forth through the couplets and diagrams in field guides whereas more experienced experts focus on key couplets,
• Experts create custom indexes, bench sheets, tally sheets and glossaries to be able to jump to couplets and keep track of moves made while trying to narrow down search spaces,
• Both trainers and experts made it clear that photographs and videos were helpful in understanding how insects looked and behaved in the wild, something vital to successful ID,
• There is a direct correlation between the level of detail and verisimilitude in a diagram and its usefulness at different stages of the ID process.

Collating data, we were able to come up with a set of questions that would be crucial to the redesign in three key areas: the macro-level architecture and flow or navigability of the site’s pages; meso-level content and interactions; and micro-interactions and site behaviors.

Some of the specific challenges that our design should aim to address:

• Should the site be structured along the lines of the strict linear taxonomic structures of a key, or the looser complementary structures of a field guide?
• How does the database expand to accommodate a large collection of over a 150 insects, while at the same time allowing for both ID and exploration or browsing?
• How do search and filter mechanisms efficiently narrow down search spaces while at the same time active scaffolding and aiding in the ID process?
• How do we drive attention to specific parts of the anatomy and differentiate between various taxa levels?
• How do we integrate information about best practices, toolkits and individualized expert processes?

Low Level Prototyping

Over the last few months, the team has created several sets of prototypes at various levels of fidelity following different pathways along several interaction paradigms; we began by mapping out the general architecture and workflows, and then proceeded to creating mockups with content. We are now in the process of designing generative exercises and usability studies with interactive prototypes to use in an upcoming workshop, where we will get citizen scientists and domain experts to sit together and co-design the website, using the prototypes we’ve created to facilitate conversations around the redesign.

Design Researcher & Interaction Designer: Ahmed Ansari

Learning Sciences Researcher: Lauren Allen

Research Assistants: Bo Kim (Design), Jen Liu (Learning Science)

Principal Investigator: Marti Louw