Wiki

PERSON

Task 1: Used heatmap to solely determine number of shoppers when the user should have used the actual map. Lack of finite numbers on the map may be the reason. For the last usability study #3 posed a problem, with this interface, the person was able to identify the top three stores a person visits after Safeway quickly and without confusion.
Task 2: Quickly identified a general range, but took more time to pinpoint the actual value. Didn't use filter until prompt.
Task 3: Used every control we had minus zoom, asked about demographics. Used binning to see if different time periods were effected.
Suggestions: Didn't like double slider because they wanted to select disjoint times. No label for out flow and combined flow, confusing. Wish they could have seen the actual locations on the map of the people. The scales for the top map and bottom map are different and should be labeled separately. Wish they could have toggled the actual data point for the store on the map.

Task 1: 1) Do not trigger panning on the map, easily figure out Safeway. 2)3) The user did hover on the Safeway and tell the traffic flow out of Safeway easily without any hint.
Task 2: 1) The user noticed the filter on the control panel and left Macy's alone. Nevertheless, the user noticed a tiny bug on heatmap: The color of Macy's is dark green but it is not displayed in the legend aside. Afterwards, the user hover over the cell in the heatmap with bug, gladly finding the correct number of customers at that time and answered the exact number of people visit at 11:00. 2) When conducting the second task, another bug appeared: when the hour filter is fixed at a certain point(e.g. 11am), it's only available to expand the filter to one direction. The user had to filter to 10am-11am, then 10am-3pm, then 11am-3pm, which is not reasonable. 3) With the help of the heatmap and the time filter, the user clearly identify there are more people at 3pm.
Task 3: The user separately filtered the time to 4PM and 10AM, as well as filtered Apple and Macy's, finding 4pm at the center of the map is better via the map: there is just no one at 10am coming out of Apple to Macy's. By the way, the user generally regards the flow as out-flow.
Suggestions: Fix the two bugs described in Task 2). The user does not like the color scheme.

##PERSON

Task 1: Did not trigger panning on the map, easily figured found Safeway. Did not hover over the store. Followed paths to see paths going to Capital One, Macy's.
Task 2: Said no. Didn't use filter.
Task 3: Said center because of the amount of paths there and used heatmap to compare nearby stores and 4pm intensity.
Suggestions: Use brighter colors. Use colored paths for distinguishing intensity. Use actual logos of stores for easier identification.

##PERSON

Task 1: Found safeway instantaneously. Didn't bother panning. Used the heat map to figure out if people were in safeway. Tried for a while to understand if people were in safeway by the paths. After explaining direction we got TMobile as an answer
Task 2: Said no, not many users at 11. Unable to determine Macey customer number due to granularity. Didn't see tooltip
Task 3: Tried to figure out by paths. Wanted Inflow. Used the filter to filter paths. He put it in center due to high traffic in the center at 4pm vs smaller arcs near Apple in the morning.
Suggestions: Bothered by grey as a color value. Thought it was empty. Didn't think that hover was very important.

====

##Overall

People were bothered by colors, they didn't recognize that this is a double ended problem