Abstract:

Today Information and Communication Technologies (ICTs) are reaching in the hands of people in the remotest corners of the world, from mobile phones and PCs to handheld tablets. If we consider mobile phones, as of 2014 there were 6.9 billion accounts in the world and 78\% of the subscribers lived in developing countries. However just having access to ICTs does not mean being able to use them to one’s advantage. There might be various intermediating factors that impact the use of these devices-- low-literacy, language barriers, lack of technology experience, lack of ICT maintenance infrastructure, etc.

In this thesis we start by focusing on one of the factors—textual low-literacy. About 775 million people in the world are completely non-literate, and even more are able to read only with great difficulty and effort. Many such users avoid complex functions, and primarily use phones for synchronous voice communication only. There is a significant body of previous work that looks at UI design for low-literate users, focusing on graphical and voice UIs to help low-literate users overcome the need to read text. While some of this work shows that low-literate users prefer non-textual interfaces, there still remains other cognitive challenges that impede use of UIs even when they are Text-Free and do not require any reading. One of these challenges, as suggested by anecdotes in related literature and our own previous work is navigation of hierarchical UIs. The other challenge is transferring learning from instructional videos and applying to actual practice.

In this thesis we study how transfer of learning of video-based skills can be enabled through presentation of instructional videos, and how navigation of ICT UIs can be enabled through appropriate Information Architecture design, even where the UIs are Text-Free. We focus on first-time usage scenarios with minimal training. We conduct controlled usability studies of variations of instructional videos, with first-time, low-literate users from urban slum communities in Bangalore, India. This is in the context of training for use of a vacuum cleaner. We follow this up with controlled usability studies comparing different Information Architecture designs of graphical UIs—a list design, a shallow and a deep hierarchy– presented on a PC and mobile phone, again with first-time, low-literate users from the same communities. Our second and third experiments are conducted in the context of finding 40 familiar household items.

The main contribution of our thesis is concrete proof of hypotheses from three controlled experimental studies that skills required for transfer of learning of video-based training, and for navigation of hierarchical UIs, are correlated with levels of textual literacy, even when the UIs are Text-Free. Based on results from these three experiments, we make recommendations for presentation of training videos for technological appliances and information architecture design for PC and mobile phone UIs for search tasks of familiar items. In addition, we have some surprising findings that complement or disprove our experimental hypotheses and we list selected ones in our thesis, with directions for future work.