A Cost-Benefit Study of Text Entry Suggestion Interaction

Abstract

Mobile keyboards often present error corrections and word completions (suggestions) as candidates for anticipated user input. However, these suggestions are not cognitively free: they require users to attend, evaluate, and act upon them. To understand this trade-off between suggestion savings and interaction costs, we conducted a text transcription experiment that controlled interface assertiveness: the tendency for an interface to present itself. Suggestions were either always present (extraverted), never present (introverted), or gated by a probability threshold (ambiverted). Results showed that although increasing the assertiveness of suggestions reduced the number of keyboard actions to enter text and was subjectively preferred, the costs of attending to and using the suggestions impaired average time performance.

On Suggesting Phrases vs. Predicting Words for Mobile Text Composition

“…They also liked that text entry felt faster and easier in the Phrase condition, but some commented about spending a lot of time reading phrase suggestions (though there was no significant difference in ratings on "I felt like I spent a lot of time reading the suggestions": Phrase mean 3.0, Word mean 2.6, F(1,15)=1.5, n.s.). Participants commented that both Word and Phrase suggestions were often distracting, confirming the findings of a prior study [28].…”

“…Despite its utility and near-ubiquity on mobile devices, word prediction has received relatively little research attention in mobile text entry. With appropriate parameter choices, a system can offer both corrections for mistakes and completions for partially entered words [5], but offering suggestions can incur costs of perception and interaction that hinder entry speed [28].…”

“…While the suggestion bar plays a central role in commercial products used daily by billions of people, it has received limited attention from academic research [5,28]. Today's suggestion bars generally offer single words.…”

The effects of interaction sequencing on user experience and preference

“…Previous studies have shown asymmetries between objective performance and subjective experience of assistive user interfaces (Quinn and Cockburn, 2016;Quinn and Zhai, 2016), but not the influence of sequencing their successes and failures, as examined here.…”

“…Snap-to-grid eliminates many uncontrolled variables that occur in other assistive domains. For example, if text-entry autocorrect were used, it would be difficult to control the timing and presence of user errors, potentially requiring artificial constraints on error-free performance that may influence the results (Quinn and Zhai, 2016).…”