We recently conducted iPad user testing sessions from our lab at Brick Lane in London to assess user interaction with a retail customer service application. For those of you interested in how to get started, this post briefly explains our technical set-up and some key planning considerations for the research sessions.
Due to the nature of the application we had two user participants per session to fully observe the interaction between retail assistant, customer and iPad. To do this we had to ensure the users could stand or move around – much as they would in a store environment.
We quickly ruled out fixing cameras to the iPad as this would prevent users from handling the iPad in a natural way and would present issues with screen glare under lights in certain positions as the participants move around the room.
We moved on to screen recording equipment. If picture in picture recordings hadn’t been so important for user testing analysis, Silverback software recording the iPad screens only would have presented the most natural use for the participants. Thanks to a really useful post from CX Partners we found a solution that would allow us to capture the iPad screen and video of the session in one easily managed file.
This set-up would give us the recordings we required for analysis but not the flexibility of movement we wanted for our participants. So we adapted this set up to include:
- Apple MacBook Laptop with screen cam disabled.
- HD webcam mounted above the 1-way mirror to capture participant movements, gestures and interactions with each other as well as the iPad.
- Quicktime Player to record and show the HD webcam picture on the MacBook.
- Apple iPad 2 (with extended VGA cable – to allow greater movement – and VGA splitter).
- Epiphan VGA2USB frame grabber to see the iPad mirrored on the MacBook screen.
- Silverback software to record the MacBook screen, thus providing a picture-in-picture of the participants using the iPad.
In practice this worked well. Some limitations included:
- The extended cable with VGA2USB frame grabber attached added a little extra weight to the iPad.
- We taped the cable to the iPad as this had a tendency to slip out as the participants moved around the room and tilted the laptop.
- Having a cable attached to the iPad affected some, but not all, user interactions with the device (e.g. tilting).
One concern ahead of the research sessions was the possibility of each participant influencing the other during the research sessions. To mitigate this we:
- Prepared individual questionnaires ahead of each session.
- Provided users with unprompted ‘role-play’ tasks to view natural interactions before carrying out prompted tasks.
- Interviewed participants individually to ask closing questions.
Role-play was a very effective technique in encouraging natural interactions between both participants. The depth of their involvement in each scenario meant that certain technical constraints were overlooked or ‘forgotten’ by participants, giving us some great insights into the use of the iPad and its application as a customer service tool.
We’ll continue to develop the set-up in our lab for handheld devices and welcome your suggestions, hints or tips to improve the validity of iPad user testing and gain greater insight from results.