Date Published: February 6, 2019
Publisher: Public Library of Science
Author(s): Gemma Fitzsimmons, Mark J. Weal, Denis Drieghe, Veronica Whitford.
There has been debate about whether blue hyperlinks on the Web cause disruption to reading. A series of eye tracking experiments were conducted to explore if coloured words in black text had any impact on reading behaviour outside and inside a Web environment. Experiment 1 and 2 explored the saliency of coloured words embedded in single sentences and the impact on reading behaviour. In Experiment 3, the effects of coloured words/hyperlinks in passages of text in a Web-like environment was explored. Experiment 1 and 2 showed that multiple coloured words in text had no negative impact on reading behaviour. However, if the sentence featured only a single coloured word, a reduction in skipping rates was observed. This suggests that the visual saliency associated with a single coloured word may signal to the reader that the word is important, whereas this signalling is reduced when multiple words are coloured. In Experiment 3, when reading passages of text containing hyperlinks in a Web environment, participants showed a tendency to re-read sentences that contained hyperlinked, uncommon words compared to hyperlinked, common words. Hyperlinks highlight important information and suggest additional content, which for more difficult concepts, invites rereading of the preceding text.
One of the main differences between reading on and off the Web is that the materials that are being read on the Web contain hyperlinks embedded within the text. Two differences between reading hyperlinked words compared to plain words are explored in this study: Firstly, hyperlinks are typically coloured and therefore salient compared to the rest of the text and secondly, a hyperlink links one piece of information to another, perhaps on a separate page of the same website, or a different website all together. Hyperlinks are a tool to navigate the Web and the word chosen to be hyperlinked often represents the page the hyperlink is linking to. This paper systematically explores these features in order to understand the impact of hyperlinks on reading text on the Web.
Experiment 2 follows on from Experiment 1 by including multiple coloured words in a sentence to investigate if additional salient words will have an impact on reading behaviour compared to the single coloured word presented in Experiment 1. If a single coloured word causes a reduction in word skipping for that word, will it occur for all salient words in a sentence when there are multiple? Or will an effect of “over-signalling” occur where the signal of importance is reduced when words are coloured seemingly randomly [19,20]. Note that on the Web the presence of multiple hyperlinks across the screen will likely be the default as opposed to a single coloured word.
Experiment 3 explored whether perceiving the coloured words as hyperlinks influences reading behaviour and whether they are processed differently to coloured words in plain text. Previously it has been suggested that the hyperlinked words would be fixated for longer due to the reduced visual discriminability of the blue words [2,13]. However, because blue hyperlinks are so commonplace in webpages, the processing may become automatic [4,31]. This could mean that in the case of hyperlinks, blue text is automatically processed as being a hyperlink without an additional cost because blue text tends to always be a hyperlink. We implemented a word frequency manipulation of the hyperlinked words in order to explore whether common lexical effects are present in hyperlinked text and to investigate if they are modulated by the word being hyperlinked. For word skipping, we predict that high frequency words are skipped more often than low frequency words . Whereas Experiment 2 has shown that when multiple words are coloured there is no effect on word skipping of the colouring (compared to reduced skipping when only one word is coloured), the signal function of colouring which indicates the presence of a hyperlink might draw attention to the word and as such increase the chance of fixating the hyperlinked word, reducing skipping for these words.
Combined these three experiments allowed us to examine the potential differences between reading text with and without a coloured word embedded in the sentence and when this coloured word was a hyperlink or not.