Project Summary

Gallery is an open-source photo album organizer, allowing users to imbed album management functionality into their own websites. Despite its powerful functions in organizing photos, there are too many buttons to choose from, too many descriptions to read through and too many steps to go into. Users are thus overwhelmed by the information provided to operate Gallery. Having conducted a sponsored student project in regard to increase the usability of Gallery, Jakob Hildon –  a 2nd year HCI student, entrusted a 1st year HCI student project team – PhototiC to further probe into this open-source ware in order to make it more user-friendly and usable.

In the beginning, Gallery was exhaustively investigated and thoroughly experimented for obtaining a better understanding. Discovering current problems were the key to our next step. Preliminary research was then conducted in order to penetrate present users’ acceptance and expectations of the new Gallery. After obtaining responses, the stage of data analysis began. The affinity diagram and work models were then built. Following that, four personae were wholly developed to describe the desired user's daily behavior patterns. The next step was to generate a low-fidelity prototype for further discussion and insights on the redesign. Afterwards, the persona-based scenario was created based on each user's working environment, their background and frequency of using Gallery. Finally, all previous conclusions and observations were combined to construct a high-fidelity prototype of the new Gallery, which was then used to conduct user testing.

The final product of this project is a testable prototype of the new Gallery, which will then be handed to the Gallery developer team. Hopefully the developers of Gallery can find values in it and incorporate the insightful redesign ideas into the current version of Gallery for the purpose of transforming Gallery from an overloaded, complicated and developer-centered system into an intuitive, simple and user-centered one.


Review of Milestone 6

We received many valuable feedbacks from classmates on our hi-fi prototype. We are glad that many of you like the skit, and how we use it to guide our present the prototype.  And you like that we put many features in the prototype, which actually costs us more time to make sure them work properly before we go on to user testing.

Later on, we find many of your critiques coincide with what the users are saying; and they refresh our opinions on the prototype in various ways. Thank you!


Getting Ready for User Testing

According to Alan Cooper's explanation of user testing, "it is a collection of techniques used to measure characteristics of a user’s interaction with a product, usually with the goal of assessing the usability of that product." 

Typically, usability testing is "focused on measuring how well users can complete specific, standardized tasks, as well as what problems they encounter in doing so". Therefore, we design a task list to guide the user testing, which is meant to test users on all the features within the prototype.

The results "reveal areas where users have problems understanding and utilizing the product, as well as places where users are more likely to be successful". And we intentionally make the task list very generic, so as to observe how the users explore the functions.


User Testing--What we did

With the feedback from clients and classmates, we modified our Hi-Fi Prototype and make sure everything is functioning properly. Here is a list of instructions that users are asked to follow. As what we did in Contextual Inquiry, the users are required to sign the Informed Consent Form. Background information and observation sheets are also used collect data.

Besides the face-to-face user testing, we also reach out for some Gallery developers at distance. With the prototype posted online, our client helps send out the testing invitation to the Gallery Developers. An online survey follows up so as to gather their comments and insights.


Findings and Recommendations

Finding 1: DROP-IN BOX sounds good, but HOW?
We observed that very few users intuitively dragged photos into the DROP-IN BOX. e.g. Almost all users just clicked on the Browse button, trying to choose certain directory to upload photos. They couldn’t tell there are two ways for uploading, one is stated above, and the other is to drag photos directly from personal folders.

Recommendation 1A: Change the term "DROP-IN BOX" to a more explicit phrase "Drop your photos here!"

Recommendation 1B: Slightly distance the Browse button from the box.

Finding 2: Can I… click?
We found that most users were not quite certain about what can be clicked on and what cannot, which indicates that there aren’t enough affordances for clicking.

Recommendation 2: All icons, tabs, buttons, albums, photos should be highlighted in a way that shows this function has been chosen. e.g. when hovering on a photo, the photo will be framed. Besides, all contextual links should also be emphasized to imply that this function could be clicked. e.g. when hovering on “delete” or the depth of Gallery, the text will be either underlined or shaded.

Finding 3: It’s a MAP!
We discovered that many users like the concept of position photos on the map pretty much, but they thought that it should offer an easier way to find the exact place of the photo and indicate more clearly that they can drag photos from the photo stream directly onto the map, i.e., the map should provide more flexibility and affordance. e.g. some users even double click on the photo.

Recommendation 3A: Put a search box near the map for users to search locations on the map and the address will automatic fill-in so that users don’t have to drag and drop many times to find the exact place.

Recommendation 3B: When hovering on the photos in the photo stream, the hand-shaped cursor will appear to show that these photos afford dragging.

Finding 4: What does that mean?
Some users didn’t understand the word we use for certain functions such as “Permanent Link” and “Public” or thought that there’s a better way to name them. e.g. some users suggested that we should use “ok” instead of “enter”.

Recommendation 4: Change the term “Permanent Link” to a more explicit phrase like “link to this photo” or “embed code”. Also change “Public” to “everyone can see it” and “enter” to “OK”. Besides, place [?] by each function to explain the usage of it if needed.

Finding 5: I wanna BATCH EDIT my photos!
Some users stated that although we had batch edit function for photos, it’s not flexible enough because batch edit all photos within one album means that they still had to edit certain photos individually. Users expect a way to freely choose photos and then batch edit them.

Recommendation 5: Provide an explicit link for Batch operations where users can choose photos freely from any album and then drag photos to the drag-in box for preview and batch edit the chosen photos.

Finding 6: Photo stream? I don’t like it!
Some users expressed that they disliked random photos of the photo stream, claiming that it shouldn’t take that much space and place.

Recommendation 6: Use the photo stream in a more useful way, i.e., most recently uploaded or the highest ranked photos.

Finding 7: Tag cloud? I am not a heavy tagger.   
Many users didn’t consider tag cloud to be the primary interface for navigation but it’s still good to have it because it would prompt more people to view your photos.

Recommendation 7: Provide a separate tab for Tag cloud where users can sort the tag by popularity or alphabet.

Finding 8: It’s jammed together!
Some users stated that it’s hard to grasp the idea behind each module on the left side bar in that the modules were cramped and jammed together, which hindered the actual important operations they could handle.

Recommendation 8: Disperse each module and have tab-like text accompanied with it. If needed, extend the box of that module by clicking on “+”.

Future Work

We’re now awaiting feedbacks from these three places and expecting most of Gallery developers’ comments on our hi-fi prototype. A large proportion of our future works will be highly involved with their valuable replies. Through our client, we would like to have more contacts with the core developers of Gallery and also follow up with them on how they view our redesign and what parts of our prototype could be integrated into the new Gallery. 

 © PhototiC, 2008-2010