Open Context

Major Research Insights

Design

By analyzing and categorizing the research findings, I iterated and refined the design solution with a principle of maximizing space usage and minimizing interaction (such as scroll back and forth). I mainly focused on optimizing 3 aspects : visual hierarchy, filter, and data export.

Design Philosophy

Redesign the search screen to maximize info while minimizing effort/interactions.

Phase One: Existing Filter Design Mismatched with Users' Mental Model

By analyzing usability test data, I discovered that the old filter function was not intuitive to users: people cannot find filter function, especially time filter in the old searching page.
One of main reason is because users view the time filter as a static visualization like the map because the two are visually grouped together. Failure to find filter causes users to pause and abandon their task, result in decreasing loyal users.

Hence, I put time filter and general filter in close proximity to each other to conform user's mental model. Users believe that grouped sections have similar functions and that the time filter belongs to the general filter. Conform to the existing mental model, instead of adding a tutorial to on-board users, could also save time and labor costs.

I further add detail function to improve filter functionality, supporting by the interview insights and deck research, including type-in feature and multi-select checkbox.

Phase Two: Optimized Visual Hierarchy of Search Page

The old search pages has a mismatch between the priority of function and the user needs.
Thus, inspired by multiple visualization based web application, we aimed to prioritize the most frequently used functions by positioning them in a central and prominent location, while also utilizing space effectively.

Biggest change was we placed Filtering Options menu and Search results on the same screen for users to view them simultaneously and decrease the difficulty of navigating between features.

The Old Export window is not intuitive as its huge cognitive load and unpredictable exported documents, results in multiple failure export for user.
In the initial design round, we add a navigation bar to separate options in categories and in readable format, decrease the density of information.

In the next iteration, we add a preview of the data in the export tab that could be presented with export options to  provide an overview about the data to solve the problem of false download.

Phase Three: Data Export

Data Results

We conduct the 2nd usability test to evaluate our design solutions and got positive feedback. After we re-ran the experiment, we were able to positively impact all metrics including complete time, complete rate and user rating

Design Handoff

Main Design Changes

Design Handoff Documents for Clients

I arranged the design components and created hand-off documents for the develop team and project manager, including responsive design guidelines with demo, user flow and specs for core interfaces. A hand-off meeting is also held to on-boarding the develop team effectively.
Throughout the project, I received positive feedback that the developers felt they had a clear understanding of the interactions and UI design decisions.

A Design style guide is also created for developer and following designer. Component was categorized by color, typography, icons, buttons, bars, etc.

Hi-fidelity Prototype: Go Through the Open Context Search Journey

Design Hand-off: Responsive Design Demo

Detailed Research Process and Insights

Go back to research overview

Goal: Understanding the system and core audience

1. Understand the motivations of Open Context’s primary user groups in accessing archaeological data.
2. Determine if Open Context users could get the information they seek in the format they need.
3. Identify areas in Open Context that can be improved based on foundational usability best practices.

Research Method

Over the course of five months, we collected, analyzed, and synthesized data regarding Open Context’s usability through the following research methods:
1. Interaction mapping of the main site pages
2. User Interviews across users and stakeholders
3. A comparative evaluation of similar archaeological databases and platforms
4. Surveys
5. Heuristic Evaluation
6. Usability testing
In my portfolio, I will focus on 3 of these methods that closely related to our design changes.

Phase one: Interaction Map

To understand the goals of the Open Context website and identify user experience issues, our team started testing and recording current website interactions. As a result, we provide interactive maps that visually show how to navigate through the website.

Through this exploring activity, our team has a deeper understanding of the functionality of Open Context and find out 2 break-down points related to search function.
1. Entry points to the site's search function are hard to find.
2. Differences in the highlighted Search types is not immediately clear, and individual Search pages include similar functions.

Phase two: Survey

To get insights supported by quantitative data and find volunteers for further research activities, we sent out surveys to current and potential Open Context users. We piloted the survey with the client to test survey functionalities, validate that questions covered the entire landscape of potential choices, and to ensure that the survey complies with client requirements. Survey distribution was reliant on the client who sent it to target populations through various channels such as social networks (Twitter, personal social posts) and professional distribution lists.

Our final survey obtained 108 responses, of which 56 were fully complete. The target number of responses for was derived assuming a Confidence Interval of .50 and a Confidence Level of 95%.

1. Open Context users are “expert” users that already have some form of experience with open data repositories and are comfortable in working with data. 
2. According to 80% of respondents, Access to Open Data is the most important function of archaeological databases, while Data Export remains high in frequency and relevance(ranked #2 and #3, respectively).

3. Usability test

In the interactive map, we found that the search function is not clear enough for users. We would like to further explore the usability of Search and Filter tools in Open Context.
Participant sessions were conducted over Zoom by two team members over 60 minutes. Participants were provided “task sheets” with instructions, followed by a post-task questionnaire.  Results were captured on an annotation spreadsheet with qualitative (quotes, behaviors) and quantitative (time to completion, success ratings) data. Results were individually then holistically synthesized to understand trends across demographics.

Overall, the usability tests showed common user tendencies across all demographics and ability levels, such as:
1. Back and forth navigation between search results and filters, since both cannot be viewed on the same screen.
2. Reliance on iconography to drive navigation.
3. Though instruction is placed, the design of data export is intuitive and inconsistently displayed on the site.

Go back to research overview>>>

Lesson Learned