@MASTERSTHESIS{pgi2018021, author = "D. Mijatovic", supervisor = "M. Dunlop", title = "The Impact of Vibrational Directional Feedback on Text Entry on Mobile Devices", school = "Department of Computer and Information Sciences, University of Strathclyde", year = "2016", abstract = "One of the key disadvantages keyboards on mobile touchscreens have in comparison to standard personal computer keyboards is their lack of physical feedback. A user on a physical keyboard is made aware of immediately of when they have hit the edge of a key, making them more aware of a potential mistake. Furthermore, this feedback helps users to adjust their typing so they more often hit the centre of a key, improving their accuracy over time. These effects are not present with keyboards on flat touchscreens. This project investigates a vibrational feedback model that seeks to mimic the feedback available to physical keyboard users. Two actuators plugged into an Android device via audio jack and attached to either side of an Android device, provide vibrational feedback that notifies users when they have hit the left or right edge of a key. This feedback takes the form of vibrations in the individual actuators, on the left and right of the device. Users holding this device will know from the actuators whether they have hit the left or right side of a key, as either the left or right actuator will vibrate. If they have hit the centre of the key, both will vibrate. An application was developed in Android where a custom keyboard was built that could take advantage of this feedback model for experimenting. Experimentation was carried out with participants, who tested this feedback model, as well as a simpler design to mimic standard typing interfaces on Android phones, that fired both actuators whenever a key was pressed, regardless of where the key was hit. An analysis of objective data relating to participants' speed and accuracy while typing found that there was no substantial improvement in their performance when using the directional vibrational feedback. This was found to be in contrast to participants' opinions on the feedback, with a majority reporting that they found the feedback model useful, and believed their performance had increased by being able to spot their mistakes more consistently, as well as a smaller perceived amount of edge presses on the keyboard. A system that users find easy to adjust to and learn has merit, and as such further research and experimentation with this feedback model would be extremely worthwhile for improving users' speed and accuracy when typing on smartphone devices.", }