MaNIS Interface Project
Assignment #3 - Project Scenarios, Comparative Analyses, and Preliminary Designs
Contents
Scenarios
Robert Scenario: Specimen Availability Search
Robert has recently teamed up with an exciting project in Australia, where some agricultural scientists are looking for his assistance in developing bat-friendly approaches to reducing the impacts of the bats on the fruit industry. Robert needs to review the taxonomy of the Grey-headed Flying Fox, Pteropus poliocephalus, as a foundation for the ecological observations he'll be conducting in the field. He's also interested in any information about the changes in the fruitbats' ranges in the last 50 years. By examining DNA from museum specimens and the ones he catches in the field, he hopes to be able to say something about the robustness of the fruit bat population's genetic diversity.
He uses MaNIS to find out about specimen availability from museums. But when his search turns up only a few specimens, he remembers he needs find all the species synonymies for Grey-headed Flying Fox that may have been used over the last couple centuries. This will ensure that his MaNIS search results will include all appropriate specimens. He uses the authoritative list of synonymies available online at the Smithsonian's Mammal Species of the World website. In MaNIS, he directs his query to all museums and limits his search to specimens collected in Australia between the 1950s and the present time. He specifies the species and all synonymies he is interested in. He checks the search results for any specimens that include tissues samples, as those are the best candidates for collecting DNA. He makes a note to track down those specimens first. He specifies that the results include information about: the museum, museum contact, cataloging, field notes, preparation, collection locality, and collection timeframe. He downloads the results of his query onto his computer, one file for each museum. Finally, he e-mails the various museum contacts to request for a specimen loan or a site visit and attaches the results of his search to indicate to the curator which specimens he is interested in.
Helen Scenario #1: Endangered Species Identification
Helen's firm has received a request for an assessment of a local plot that is targeted for development of a new freeway bypass. Before the city can grant the developer a permit, this environmental assessment must be conducted.
To gain some historical background about the fauna that were present in the area, Helen decides to go to the MaNIS website. She wants to focus on a specific geographic location, so she must make sure that all the different ways used to refer to the location of the proposed bypass are included in her search. She uses MaNIS to pull up a list of mammals ever collected within 5 miles of the location. Helen wants to know which of the species displayed is on the current endangered species list, so she has MaNIS highlight those. She also views distribution maps for each species to gain a clearer picture of their geographical coverage. Helen downloads the results and maps and files them away for reference. She cross-checks the species lists with the field notes information available on the California state Fish & Wildlife website, so that when she goes to visit the location and do field surveys, she'll know what signs to look for.
Helen Scenario #2: Protected Area Management
Helen has been called in to work on a special project in the Yolumpite State Park. The State Park officials are considering adopting a "let-it-burn" policy for their oak grassland areas, where they have been suppressing fires for the past 40 years. They are concerned that without the fires, the native grasses are being overrun by the european exotic grasses. However, they know from other studies that the grasses have a close relationship with the small mammal populations. They want to know what the small mammal/grassland communities were like before the fire suppression policy started, and whether that community has a chance to re-establish if they start a let-it-burn policy.
Helen has plenty of fieldwork to do to answer this question, but she also needs the historical data. She uses MaNIS to create a list of all species ever collected in Yolumpite State Park. Then she explores some different time frames to see any obvious changes in the mammal community. She sees that Yolumpite State Park is located in a transition zone, and that the overlap of the maps looks strange. She suspects some of the important kangaroo rat species have been named and re-named over the past several decades so that it has become quite confusing. MaNIS allows her to access the Mammal Species of the World synonymies lists, so she knows which species names to group with which other species names to see maps of the collection localities according to the most recent taxonomy. She wants to check some ecology questions with the scientist who made this taxonomy, so she pulls up the records of the most recently collected specimens and looks at the collector's name. She googles it and contacts the scientist to get started on her work answering this big question!
Helen Scenario #3: Public Health Planning
When the Hanta virus was first discovered, people died from it. Virologists figured out that it was associated with a particular species of small mammal, and that each species had its own related strain of the virus, not all of them harmful to humans. The public health officials called in Helen to help. While the virologists figured out which strains of the Hanta virus were noxious, Helen's job was to help the health officials figure out in which populated areas they would need to take precautions.
Helen uses MaNIS to call up distribution maps of the various small mammal species. She then limits the time-frame to the last 20 years, as habitat change has changed where many of the animals are found. Identifying which species are located near boundaries of human habitations, she focuses in on those specimens collected locally, saves those records, and contacts the museums which hold the specimens to arrange for DNA samples. The virologists will use the DNA to help see how closely related the various species are and how likely the Hanta virus is to spread to them. Meanwhile Helen is able to prioritize the public health officials' efforts on one part of the edge of a populous area, where humans and the small mammal species are most likely to interact. The public health officials posted signs in that area, and no more cases of the Hanta virus were reported.
Revision to Personas: Saying Goodbye to Valerie!
From interviewing educators, we recognized that the prepared lessons they need are not within the scope of the MaNIS project. We developed Valerie as a persona anyway, to represent both educators and any non-professional biologists who might use the search interface.
After doing our comparative analysis, however, we discovered that there are other sites we'd recommend Valerie use before we'd recommend turning MaNIS into a site that is good for her. For example, Animal Diversity Web has digitized, interactive 3-D images of mammal skulls, which would be a great substitute for actually taking the students to a museum and looking in the drawers, an image that has come up in multiple conversations with educators. Digimorph.org has similarly impressive 3-D content and some very public-friendly features for getting into the contents of their site. In fact, Valerie's interest in teaching "how science is done" and "environmental impacts" is not mammal-specific, so the broad coverage provided by these sites is a benefit.
Secondly, we are finally appreciating a very important distinction: the beneficiaries of the MaNIS system are not necessarily the people who will use the MaNIS search interface. After our project proposal, John Wieczorek clarified the audiences for the MaNIS system as a whole. MaNIS is being developed "as a clearinghouse for anyone in need of mammal specimen data," and as such will serve many folks, including 1) the participating museums, their curators and collection managers 2) Researchers - systematists, biogeographers, conservationists, etc. 3) Decision-makers and 4) Educators. However, this list is not the audience for the MaNIS search interface. Just as our interviews with museum curators indicated that they don't need to actually use the search interface, we suspect that neither will educators or decision-makers. Instead, educators will benefit from the MaNIS system when biology curriculum developers use it to create lessons using real data and maps. Similarly, decision-makers will benefit because the wildlife biologists they hire will have rapid access to important specimen distribution data.
Recognizing this distinction, we are now focused on creating a search interface that will support Robert and Helen, researchers with biological training who will use the system infrequently in order to answer parts of questions that arise in doing their work.
Comparative Analyses
Museum of Vertebrate Zoology
The Museum of Vertebrate Zoology website (http://www.mip.berkeley.edu/mvz/) provides access to data and images in various collections such as birds, amphibians, mammals, and reptiles. It is a good candidate for comparative analysis because its mammal collection is one of the collections currently aggregated by MaNIS. As such, its interface operates with data that is very similar to that of MaNIS.
Strengths:
- Uses representative icons to indicate collections available for browsing. Additionally, the name of the collection (e.g. mammal collection, bird collection, etc.) is displayed when the mouse is positioned over the icon.
- Uses radio buttons to denote mutually exclusive item selections (individual collections).
- While in a specific collection, a photo of a mammal is shown in the upper left corner of the page. This indicates to the user that the current collection being searched is the mammal collection as opposed to the bird collection, etc.
- Indicates to the user how the search criteria are combined to perform the search. At the bottom of the page is a message that says, "Partial words will match. If more than one field is completed, they will be AND'ed together. Case is not important."
- Provides a button for the user to clear the contents of a form in 1 stroke.
- Provides partial word matching so that users do not have to remember things like people's exact names for Collector and reduces the amount of typing required of the user.
- Provides an immediate link that allows the user to return to the search criteria page from wherever he or she is located.
- Defines 3 different views of the data to help users generate a narrower and more defined result set (e.g. all specimens, species count, species parts).
- Provides display options for users and also indicates the wait time for each display option:
- A tabular summary of the results with links to more specific data about each record
- An option to expand the tabular summary to include more fields
- An option to download the information
- An option to view information on a per-record basis
- Indicates the number of records that matched the search criteria at top right corner.
- Provides dropdowns for selecting items such as country. This reduces data errors and typing.
- Provides a consistent list of display options that behave in a predictable manner no matter what query "View" is selected (e.g. all specimens, species count, species parts).
Weaknesses:
- Inconsistent with user expectations. If a user navigates from the home website to the Mammal information page, selecting the link to the database takes the user back to the entire MVZ collection that includes birds, fish, etc. This means the user has to re-select the mammal collection.
- Page where the user selects collection to browse contains 2 distinct groupings of radio buttons but neither one of them is labeled.
- Page where the user selects collection to browse provides display options for user such as sort, limits to number of records, and limits to the type of records returned. These display options are located above the search criteria form which makes them look more prominent. This breaks the visual flow of the user and interferes with the user's main task of entering search criteria.
- Search criteria page doesn't indicate which fields are required and what formats are required for the free text entry fields.
- After a user fills out the selection criteria form top to bottom, the user has to go back up to the top of the page to click the search button. This disrupts the natural flow of top to bottom execution.
- Button that clears the search criteria in a form is positioned very closely to the search button. A user may accidentally hit clear instead of search and lose all entered search criteria entered.
- Displays the SQL database query at the bottom of the page. This jargon is too technical to be useful and also can distract users.
- An option to download results is offered but functionality is inconsistent with how other applications execute downloads. Selecting the option opens up a browser that displays a tab delimited text file. The user then has to save the text file using the browser menu option File>Save As.
IUCN Red List
The IUCN Red List is the internationally authoritative list of endangered species. The Red List has been online since 2000, replacing the print publication. Similar to MaNIS, the search website is designed to support researchers and conservationists in locating species-centered information.
The Red List provides two search interfaces, "Search" and "Expert Search." The primary functionality on the "Search" interface consists of a text search box, with various other settings defaulted so as to support the search for specific, properly spelled taxonomic or common name terms (screenshots). Adjusting the settings allows for narrower searches within common names, taxonomic terms, authorities for the information, biomes, and various facets of the term hierarchy developed to describe the status of the species on the Red List. A nice feature, which is rather cryptically presented, supports search based on common group name terms, such as "bats" or "game birds." Unsuccessful searches show the terms of the search and a short sentence about the ambiguous meaning of the result, but provide no feedback about spelling or similar terms, nor assistance with revising the search to be broader or narrower. The words "database" and "Red List" are used interchangeably on the interface, although their equivalence may not be intuitive to the user. Successful searches show a list of species and provide a tool for sorting them by any of the displayed fields. No feedback is displayed for why these records match the search terms. Each search results page displays the appropriate citation for the search results at the bottom, which is a lovely feature.
The "Expert Search" interface is identical to the "Search" interface, with the addition of a feature at the top of the page to dynamically drill down the Linnean taxonomic name hierarchy, while at the bottom of the page are links to authority files for the Red List-specific terms. While it is true that "experts" are more likely to feel comfortable accessing information by drilling down the taxonomic name hierarchy, it seems counter-intuitive that the "expert" search interface is the one which provides the most feedback about what may actually be found within the List. An improvement would allow any user to enter the taxonomic name hierarchy at any term they knew, and browse up or down from there. Ironically, the presence of a species on the taxonomy drill-down list is indicative of its existence but not of its presence on the Red List. However, the search results screen doesn't distinguish between species not found because of spelling mistakes and species not on the Red List. Thus the absence of a species from the Red List (if a species is not "red-listed", it is not threatened in any foreseeable manner) is impossible to verify from the normal "Search" interface, and confusing to verify from the "Expert Search" interface.
In general, the search interfaces have appropriate defaults and some nice features. However, the search results screen provides inadequate feedback to the user about the reason for the success or failure of their search, and more browsing capability should be supported for all types of users.
Calflora
Calflora is "the on-line gateway to information about California's wild plants". It strives to provide information about California's plant biodiversity to professionals, scientists, educators, students, and amateurs.
Strengths
http://www.calflora.org/species/index.html (screenshot)
- Has indices of common names, scientific names, and family names
- Allows search by county and elevation
- Parameters default to "any"
- Shows geographic distribution based on Calflora Plant Observation Library on map
- Does have option for Rare or Special status plants, which includes Endangered species
- Has option for CalEPPC Listed Noxious Weed
- California Plant Name Library (screenshot) gives synonyms for entered scientific names
- Lists examples next to search boxes
- Results include photos, distribution map
- Advanced query form available http://www.calflora.org/advanced.html (screenshot)
http://www.calflora.org/occ/ (screenshot)
- Allows results to be shown as HTML table or Static Map
http://www.calflora.org/about-cf.html (screenshot)
- "About Calflora" page has lots of info
Weaknesses
http://www.calflora.org/xcreate_account2.html (screenshot)
- Must register to use site. "Register as Guest" option is kind of clunky. Should probably have its own hyperlink.
http://www.calflora.org/species/index.html (screenshot)
- Difficult to see grayed out hints
- Search listed as "Species Information"
http://www.calflora.org/occ/ (screenshot)
- Plant Observation Library page is overwhelming, similar to MaNIS.
http://www.calflora.org/xwalk/ (screenshot)
- Disclaimer in California Plant Name Library: "Please note that the information returned is EXPERIMENTAL in nature, and is known to have omissions and inaccuracies."
http://www.calflora.org/occ/instructions.html (screenshot)
- Separate "Instructions" page and "Advanced Query Form" for Plant Observation library
- Queries are case sensitive!
Compare http://www.calflora.org/species/sci-D.html to http://www.calflora.org/species/index.html. (screenshots)
- Navigation is disorienting and bizarre. Easy to lose your place. Changes from page to page.
- I can't tell what the site architecture is.
http://www.calflora.org/advanced.html (screenshot)
- Advanced query form link found only on "Instructions" page.
- Simple query form link on page leads to homepage, not simple query page.
http://www.calflora.org/about-cf.html (screenshot)
- About Calflora page has lots of links to information that should be placed in other areas of the website. For example, "Take a Tour" should probably be on the homepage.
Flamenco
The Flamenco project, led by Marti Hearst at UC Berkeley, is a search interface framework that allows users to move through large information spaces in a flexible manner without feeling lost. The interface exposes faceted metadata that helps guide users through the collection. The Flamenco interface differs from most other search interfaces because it successfully combines both browsing and searching. The interface allows users to refine and expand their searches by using a combination of browsing and searching.
Flamenco can search over collections of images or documents. The demo that uses a tobacco documents archive is probably closest to the way the interface might be used for MaNIS, although MaNIS will return database records as the end result.
Initial Flamenco screen (http://orange.sims.berkeley.edu/cgi-bin/flamenco/tobacco/FrankenMatrix?stylesheet=frank)
From the start page you can either run a keyword search or click on a link within one of the metadata categories. The italicized numbers after each term indicate how many documents in the collection have the particular descriptor assigned to them (query preview).
If you click on a link, you are taken to either a grouped results page, an ungrouped results page, or an intermediate refinement page.
Next Flamenco screen (http://orange.sims.berkeley.edu/cgi-bin/flamenco/tobacco/FrankenMatrix?q=demographic:1&group=demographic)
In this case I?ve clicked on Age group under Demographics on the first screen, and the second screen lists the documents grouped by age categories. I can further refine my search using the metadata on the left side of the screen or keyword searching.