Manis Interface Project
Assignment #9 - Third Interactive Prototype and Final Write-up
"The SIMS213 team that undertook to assess and design a user interface for the NSF-funded Mammal Networked Information System (MaNIS) has produced a remarkably innovative and useful prototype. I believe that the team did a thorough job in investigating the expectations of a broad scope of potential users of the system - much more so than was done in the original design of the current MaNIS portal. In addition, the team was unhindered either by preconceived notions of what a useful interface could provide or by assumptions about the generic nature of the portal. The result is a prototype designed to aid both the knowledgeable researcher and the casual user to find relevant information. To achieve the functionality proposed by the team, new portal development (especially the ability to cache data indexing information) will be required, which is outside the scope of the current MaNIS project. Nevertheless, the design innovations apply equally well to at least two additional MaNIS-derived NSF projects, HerpNet and ORNIS, under which the designs can be further developed into a production system. Thereafter, an instance of the new portal based on the SIMS213 prototype could also be installed on top of the MaNIS network which inspired it. I offer my sincere thanks for the hard work, careful thought, and open-minded creativity applied by the team."
--- John Wieczorek (MaNIS Developer)
Contents
Problem Statement
The MaNIS application (Mammal Networked Information System) is a network of distributed databases of mammal specimen data. The project is a collaboration between 17 research institutions and natural history museums, funded by the National Science Foundation. MaNIS makes information available for nearly a million museum specimens. This information can be used to enhance conservation and research initiatives, both locally and globally. The back-end has been developed, collaborations are in place, and an initial interface exists.
The project goal is to perform a comprehensive usability assessment of the current MaNIS search interface and propose enhancements that will be considered for implementation by the MaNIS development team. These enhancements might also have an impact on other museum database query tools whose interfaces will be patterned after the MaNIS interface.
Solution Overview
Inspired by the well-tested approach Flamenco has pioneered, we set out to develop a search interface to the MaNIS data with browsing features effectively interwoven throughout the interaction. Four key features of the solution include 1) the presentation of the results within the same view where the search criteria are specified, 2) an emphasis on supporting the most common search tasks with simple and default interactions, 3) proposal of appropriate terms to narrow the current search, with a preview number indicating what the size of the new result set will be, and 4) feedback such as spelling suggestions, taxonomy tree navigation, and bold display of the search term in the results that steer users towards the terms that the system has results for. Unlike Flamenco, the MaNIS data is not cleanly faceted, and MaNIS users tasks suggest they may commonly wish to select two or more terms within a facet. Although our current prototype does not currently support this, we believe the design has laid the groundwork for an effective solution to this interaction.
Personas and Scenarios
Persona #1: Scientist with a Taxonomic Focus - Robert J. Mismer
Robert is a tenured professor of Conservation Biology at Stanford University. He works with bats from Austral-Asia. He particularly enjoys figuring out systematics questions, but knows there's no money in that anymore, so writes grants for questions with significance for conservation or agricultural economics. As a result, his research lab attracts grad students interested in these questions, but not interested in taxonomy and systematics. He has a reputation in the department for putting a disproportional emphasis on the taxonomic work for his students' dissertation projects.
Robert is in his early forties, 6 foot tall, brown-haired, balding with a bushy beard and twinkling blue eyes. His partner Sheila Lee is Australian, although her parents emigrated from Malaysia. Like her dad, Sheila is a dentist. Sheila and Robert have two kids, eight and 10 years old. Robert's dad fought in the Pacific in WWII. In college, Robert did a junior-year-abroad in Singapore and visited many of the places his father fought. He fell in love with Asia, started working on the bats, and spent as much time as possible doing fieldwork in that part of the world while collecting advanced degrees in Biology. He has contributed many specimens to the Smithsonian Mammal Collection, often from previously unsampled areas or under sampled species. He met Sheila on one of those fieldwork expeditions. He collects woodwind instruments, too, and frequently plays them for a laugh (for example, the Palauauan nose flute) on fieldwork social downtime or at department social events, though he doesn't claim to have any mastery.
Robert drives an old, dented-but-clean dark blue Ford station wagon, and often dresses in wool flannel plaid shirts and timberlands. He replaces the timberlands with tevas when it's warm. He's tidy and efficient, and expects his equipment to be. Technological techniques need to earn their stripes just like any field technique. His computer is just another field tool, supporting different research techniques. When a colleague at a conference shares a technological approach to data gathering that might apply to one of Robert's questions, they'll discuss the technique over dinner, and then get into the field together for Robert's direct tutorial in the technical aspects. He'd get training just like in any field technique. He learns what he needs to know to accomplish the desired task, and expects to be able to repeat it as needed to get the work done. He's impressed with the flexibility computers show, but seldom has time to explore new things. He believes in people-hours - putting in the time and effort in the field to answer questions - rather than technological shortcuts.
Persona #2: Wildlife Biologist - Helen Williams
Helen is a 49-year-old wildlife biologist living in the Berkeley hills. She's a California native, born and raised in Sacramento. As a child, she played outside whenever she could. She was fascinated by the natural world and collected insects and wildflowers. Her family went on long camping trips every summer to remote areas in Montana and Wyoming, where her father had camped with his parents when he was a child. Love of the outdoors runs in the family, and for Helen the summer trips were formative experiences. She spent hours with her sister and brother playing in meadows and streams, and was able to observe deer, antelope, and moose in their natural habitat. She has never forgotten the wonder of those early experiences.
When it was time for Helen to go to college, she stayed close to home, getting her BS in Conservation Biology and MS in Ecology from UC Davis. She now works for a small environmental consulting firm in Berkeley, performing field studies on endangered species and their habitats. She's become an expert on the endangered salt marsh harvest mouse. Her projects generally require travel to various locations throughout California. When she's in the field she takes all her field notes by hand and occasionally uses a voice recorder. She transcribes these notes to the computer when she is back in the office.
Helen loves being outdoors and hates having to stay inside. However, as she gets older she is growing concerned about her sun-damaged skin and achy joints. She realizes that she needs to do more preparation work on the computer before she goes out into the field, to try to reduce the amount of time she spends outdoors. Helen considers computers a necessary evil-she uses them to research particular sites before she goes out to visit them, often looking at the historical distributions of specific species. She does not consider herself to be an expert user of computers, nor does she want to be.
Helen is patient, shows great attention to detail, and considers herself a "wash and wear" woman. She is not bothered at all by spending a rainy night in a tent, although she is always happy to return to her home in Berkeley. She drives a Ford Explorer, which she feels very guilty about, but is great for hauling all her gear when she goes into the field. Helen is single, but has a strong sense of community with friends and family in the Bay and Sacramento areas, many of whom also work in the field of conservation biology.
Persona #3: Curator of Mammals at Hunt University Natural History Museum - Debbie Twist
Debbie is passionate about collections. As a child, she loved to visit museums, but also the houses of older people, and libraries. As she poured over collections of all types, she loved to figure out what messages they had for her from the past. In high school, she had a wonderful biology teacher. As a result, Debbie took the premed track in college at University of Michigan. During her first two summers she worked in a hospital, but after her junior year abroad in London, she decided she preferred more static biological collections!
Debbie got a PhD in the systematics of South American marsupials from UC Berkeley, and worked as GSR in the Museum of Vertebrate Zoology. She did a post-doc at the Smithsonian, and soon took position as curatorial assistant at Hunt University. She was promoted to Curator when the previous mammal curator had a stroke and was forced to retire.
Now she responds to requests about the content of the collection, decides upon and arranges loans, supervises GSR's working on intake of specimens to the collection, and welcomes researchers who come to visit the collection for a bit of work. She also conceives of, seeks funding for, and supervises projects to improve the quality and accessibility of the collection, including most recently undertaking to get all the specimens in a database and georeference the locality information. She's constantly struggling to do a lot with very little funding and staff, but she and the other curators of the other collections can get pretty clever, and generally keep things ticking over well. She also enjoys working with the other researchers and ensuring the collection is supporting their projects. While the need for such collections seemed to be taken for granted during her predecessor's time, Debbie finds her tenure as Curator has focused on how to continually articulate and advertise the importance of the mammal collection.
She's 37, with curly brown hair kept pinned away from her face, and demonstrates a calm enthusiasm for her work and the beasties. Her partner is an academic at a nearby university, so they both have a mutual understanding of the need for long days. Debbie walks to the museum every day to get her exercise in. She tends to wear comfortable shoes (good for walking and standing over counters while working with the collection) and 50's style tailored skirts and blouses or dresses, now that fieldwork no longer shapes her clothing needs.
Scenario A: Which museums should I visit to review specimens?
You need to review the taxonomy of the Hairy-tailed bats, genus Lasiurus, as a foundation for ecological fieldwork observations in California [see persona Robert Mismer] . By examining DNA from museum specimens caught in the past 80 years and the ones you catch in the field, you hope to be able to say something about the robustness of the hairy-tailed bat population's genetic diversity.
- Task 1: Get a count of available specimens for Lasiurus.
- Task 2: Narrow your results to specimens collected in California since 1920.
- Task 3: Specify that the results include only the following information: Catalog Number, Institution, Scientific Name, preparation, and Collection year.
- Task 4: Download the results
What we looked for...
This scenario was created to illustrate a common interaction a user might have with the MaNIS application. It was used to determine how easily and effectively search conditions such as taxon, location, and date could be specified. It was also used to assess the usefulness of features such as "total record count", "show/hide fields", and "export."
Scenario B: Where is the best place to capture a bat?
You're beginning a study on Lasiurus cinereus in New Mexico [see persona Helen Williams]. You'd like to capture a few animals, and will have the best luck going where they have been found before. By looking at the locality information for where specimens of these bats, you can plan where in New Mexico to go to set up mist nets. Ideally the collection record will be relatively recent, to avoid traveling to someplace where the correct habitat no longer exists.
- Task 1: View a list of all Lasiurus cinereus from New Mexico.
- Task 2: Organize the records by when they were collected.
What we looked for...
This scenario was created to test the "sort" feature of the application.
Scenario C: What species is this?
As a curatorial assistant [see persona Debbie Twist] at the Museum of Vertebrate Zoology (MVZ), you are bringing a new specimen into the collections, and need to identify what species it is. Field notes indicate that it is a bat of the genus Lasiurus, and was collected in Pasadena, California. Once you see which Lasiurus species' distributions include Pasadena, you'll be able to go to the specimen cabinets and compare the new specimen with the identified specimens, and determine which species the new specimen belongs to.
- Task 1: Pull up the records of Lasiurus bats collected in Pasadena, California.
- Task 2: Limit the results to those specimens held in the MVZ collection.
What we looked for...
This scenario was created to assess whether it was clear to the user that the location panel could be used to conduct two types of location searches - by state or locality. It was also used to verify the effectiveness of the insitution search.
Final Interface Design
Functionality
Our prototyping effort focus on implementing features that allowed us to test our basic interaction model ((1) and (2) in the summary). We also implemented examples of key browsing and feedback features of our design ((3) and (4) in the summary) as demonstrations of these behaviors.
The screen is divided into three main sections: a top panel that provides navigational links and feedback on the current search criteria, a left panel that provides the controls for setting the search criteria, and a right panel that displays the search results.
Clicking a navigational link in the top panel (Taxon, Location, Date, Institution, Advanced) shows the selected search criteria controls in the left panel. For example, clicking Location in the top panel displays the Location search criteria panel on the left.
Location Panel: In the left panel, typing a location name and clicking the Find Location button creates a Location search condition. An example of an appropriate entry is provided below the text entry box. The program retrieves all results where the text matches within the State/Province or Locality fields. The search results display in the right panel, where bold text highlights exact text matches within the results. The Location search condition indicator is displayed underneath the Location link on the top panel. Clicking the X on the condition removes it.
Clicking the Taxon, Date, Institution, and Advanced links in the top panel opens the corresponding left panel and adds additional search conditions that further limit the number of results. These panels operate similarly to the Location panel. Clicking New Search clears all conditions and begins again.
Taxon Panel: In the left panel, you can type a taxon name from anywhere in the taxonomic tree. An example of an appropriate entry is provided below the text entry box. The program retrieves all results where the text matches within the Scientific Name field. The Taxon search condition indicator is displayed underneath the Taxon link on the top panel. Clicking the X on the condition removes it.
The program attempts to match the text entry to a known taxonomic name, and provides a "Did you mean..." feature that offers one or more hyperlinks on the Taxon panel, which re-set the Taxon search condition to the known name. This helps correct for spelling mistakes and guides users towards effective search terms. When the search condition is a known taxonomic name, the chain of names larger than it in the taxonomic tree are displayed, where many of them can be clicked to implement a new Taxon search condition with that term. The names for smaller categories than the currently selected name are displayed as a list of clickable terms, with a preview number indicating the number of records that will be returned in that search. The ISIS taxonomic name web service is queried for synonyms of the taxonomic name, and links are also provided to include those names in the search results if there are any records with those names.
Date Panel: In the left panel, typing one or two four-digit years limits the search results by the year the specimen was collected. An example is provided. The program retrieves results where the year entry is within the entered range. The Date search condition indicator is displayed underneath the Date link in the top panel. Clicking the X on the condition removes it.
Institution Panel: In the left panel, a list of participating institutions appears. Clicking on any name limits the search to records from that one institution. The Institution search condition indicator with the initials of the institution is displayed underneath the Institution link on the top panel. Clicking the X on the condition removes it. If any other search conditions have been set, a preview for the number of records from each institution displays on that institution's link.
Advanced Panel: The search options here are currently for show only.
Clicking the Show/Hide Columns link opens the controls for the visibility of the results columns in the left panel. Checkboxes specify if columns should be displayed in the results window. You can click X in the column header in the results window as a shortcut to hide the column. The "sort" link in the column header displays the records sorted alphabetically by the contents of that column.
Clicking the Export link opens a dialogue box and selects the desired file format for downloading the results. An option is available to divide the results into separate files by institution, obviating the need to accomplish this with multiple searches.
Clicking the catalog number of any record shows a mock-up of the detail view of the database entry for that record.
Interaction Flow
Features Not Implemented
The biggest feature that we have left undemonstrated by our prototype is the ability to specify two search conditions from within one facet, where they will be combined using a logical OR. We also would need to have an effective way of displaying this situation for the user.
A whole variety of other parts of our paper prototype design were not implementable given the time & programming resources available. The details can be found in ToBeImplemented (.xls).
Tools Used
We built our prototype using HTML, CSS, JavaScript, and PHP. We created a MySQL database that contains a subset of real data from the current MaNIS system. We used PHP to perform the server-side interaction with the database, and we used JavaScript to perform the client-side interaction.
The tools allowed us to create a realistic prototype in a relatively short time frame. However, we had to overcome a large hurdle in that we had varying degrees of knowledge about these technologies. Some of us knew PHP, some of us knew JavaScript, and some of us knew neither, but none of us knew both of the technologies. It was a challenge to figure out how they could work together, and many long hours went into figuring out the interactions.
Design Evolution
Over the course of the semester, our design evolved from rough sketches to a working interactive prototype. The following sections describe the changes the interface went through.
Initial Sketches
Our first sketches for our prototype were based on comparative analyses of other search interfaces. We were influenced in particular by the Flamenco interface. The sketches tried to incorporate some ideas from Flamenco, and we had many different screens that were used to specify the search criteria and display the search results.
Low-fi Prototype
We created a paper mock-up for our low-fi prototype. The mock-up was constructed with simple materials: paper, pen and pencil, and Post-it notes. Most of the UI controls were hand-drawn. Despite the simplicity of the materials, the prototype simulated most of the functionality of the interface in detail. We simulated real results from database searches by downloading selected result sets from MaNIS into Excel and printing them out. The most important change from our initial sketches was that the prototype now displayed the controls for specifying the search criteria and the search results all on the main screen. The prototype consisted of a search criteria panel on the left, a search results panel on the right, and a navigation panel at the top for switching between the search criteria panels.
The results of our usability testing with our low-fi prototype indicated some serious usability problems, such as problems with the navigation, terminology, and interaction model. We addressed as many of these issues as we could in our First Interactive Prototype. This prototype simplified the interaction model and improved the terminology and navigation. The most important changes we made to the interface as a result of the low-fi usability testing were the following:
- Removed the unintuitive central Update Results button and added Update buttons closer to where search conditions are entered
- Improved terminology, such as removing the term Curation and replacing it with more specific terms such as Institution and Preparation
- Added a Sort link to each column heading in the Search Results
Heuristic Evaluation
Heuristic evaluations were conducted by Marti and the Paparazzi team on our first interactive prototype. These evaluations raised many more issues, especially the lack of documentation and the poor use of color to provide visual cues to the user. We addressed as many of these issues as we could in our Second Interactive Prototype. The most important changes we made to the interface as a result of the heuristic evaluations were the following:
- Created a Taxon hierarchy to help with searches that return zero results
- Made color improvements to relate the top navigation panel to the left panel and the results panel
- Added features to help the user with text entry, such as examples of the required formats
- Added more documentation throughout the interface
- Added a "New Search" feature
Final Usability Test
Our final evaluation used our second interactive prototype for a series of usability tests. These tests suggested further refinements that could improve our interface. We addressed as many of these issues as we could in our Third Interactive Prototype. The most important changes we made to the interface as a result of the final usability test were the following:
- Displayed preview numbers on hyperlinks that add search conditions
- Added searching in multiple database fields for location
- Highlighted terms in search results that were part of the query
Usefulness of the Evaluation Techniques
The various techniques were valuable to the prototyping effort in very different ways, yet in the end, each was instrumental in the development of a viable prototype. Each evaluation method yielded information that progressively improved the next version of the prototype. More specifically, the low-fi prototype exposed core design issues, heuristic evaluation yielded feature refinements, and the pilot study assessed the overall usability of the design.
The low-fi technique was the most valuable because it highlighted core design issues that caused users some major problems. It exposed navigation problems (users kept clicking on the sort hyperlink instead of the facet hyperlink to specify a search condition), misconstrued terminology (users were unclear on what "Curation" meant), and non-intuitive application behavior (users thought that specifying new search conditions would nullify prior ones). The low-fi technique was also valuable because it highlighted the most useful design features that needed to be carried over to successive iterations such as instantaneous feedback, sort and column selection tools, "delete" shortcuts (x's), synonyms, and "Did You Mean..." to address user errors.
The heuristic evaluation further refined the prototype that resulted from the low-fi technique by correcting the lack of documentation and poor use of color in order to provide better textual and visual cues to the user.
Finally, the pilot usability test confirmed the viability of the prototype and offered some insight into the type of audience for which the prototype would be most appropriate (beginner to intermediate user).