User testing

Oct 14th test:

User Testing Results:

1. User had trouble figuring out what to do on initial screen. How do you enter a word? What do you do? ~ Goal: Find a way so user knows exactly how to enter the word both on the very first screen and on a screen with a preexisting graph. ~ Some initial ideas: 1) Text box like Google on both initial page and preexisting graph pages 2) Arrow pointing to an area with instructions "Double click here to type a word"

2. Once expanded, what can user do? What options does the website give the user? ~ Goal: Clear way for user to identify what he can do one a simple graph has opened up. ~ Some initial ideas: 1) Instructions (little arrows which tell a new user what all he can do next) 2) Simple Tutorial-like instructions 3) Use hovering

3. How does user know the graph's nodes can be expanded? ~ Goal: Clear way to show user nodes can be expanded further. ~ Some initial ideas: 1) Instructions 2) Nodes have "link" style: underlined and blue 3) Hovering over a node causes the graph extensions to appear in grey - this symbolizes that clicking on the node will expand the graph permanently.

4. User deleting a node: The x in the corner of each node seemed to be a clear indication of deleting a node to the user.

5. User was pleasantly surprised by the cross-connections made between an opened node and an unopened node. Said it got him thinking along different lines.

6. Heart symbol was misleading: ~ Goal: Prevent confusion while still having a "like" button. ~ Some initial ideas: 1) Change symbol: maybe a star or a "thumbs-up" 2) Change location 3) Instructions on first screen

7. User was expecting more than just a shaded background after "liking" the node: ~ Goal: Make "like" option more useful. ~ Some initial ideas: 1) "Liked" word list appears on the side (suggested by user) 2) Change background color to stand out.

8. The user was confused when clicking a node, but it didn't have any neighbors listed in the database, and thus no new words to add. They weren't sure if their clicks were registering. Idea: Some sort of feedback (such as flashing red, perhaps?) when the user clicks a link that doesn't have any neighbors to add. Or, conversely, have some sort of indicator when a node does have more words ready to be added.

9. The user wished to enter phrases, not words. For example, they wanted metaphors for phrases like "in a hurry".

10. The users complained that many of the words were not relevant to their product, and at the same time wanted more words for some nodes (where our demo only had a few, or no, words to add to a node.)