The review argues that researchers should disregard eye movement behavior when trying to understand the role that peripheral vision plays in a variety of visual cognitive tasks. In some senses, one could argue that the decision whether or not to record eye movement behavior is a mere methodological concern. However, the reality is that by studying eye movement behavior, researchers can unlock the potential to confirm or disconfirm a wide range of hypotheses that may otherwise have taken years or even decades to resolve. Recording eye movement behavior enables researchers to gain fine-grained insights into moment-to-moment information processing as observers engage in a given task (Liversedge & Findlay, Reference Liversedge and Findlay2000; Rayner, Reference Rayner2009). Such insights are typically beyond the reach of behavioral measures such as response times (RTs) and accuracy alone. Eye movement metrics include fixation durations, the number of fixations, the likelihood of revisiting objects, the order of visiting objects, rate/direction of microsaccades, and many others (e.g., see Godwin et al., Reference Godwin, Hout, Alexdóttir, Walenchok and Barnhart2021). As a result, by recording eye movement behavior, researchers can better understand the spatiotemporal dynamics of information processing during visual cognitive tasks. Here, we provide some examples highlighting this point, focusing on how the recording of eye movement behavior enables the refinement, development, and resolution of theoretical arguments in the literature.

Decades ago, Rayner and colleagues examined the importance of peripheral vision in reading, and they did so by harnessing the insights offered to them by the recording of eye movement behavior (for a review, see Rayner, Reference Rayner2009). Such insights would have been impossible – quite literally – without the use of an eye tracker, because many of their studies used gaze-contingent displays. A gaze-contingent display changes what is presented to a participant based upon where they are fixating. A prototypical example might be a task wherein the letters in a sentence that are currently being fixated are visible, but beyond a small region (invisible to the participant), all other letters in that sentence are set to the letter ‘x’. Had Rayner and colleagues used only RTs and response accuracy, it would not have been possible to measure the attentional window during reading, nor would it have been possible to determine that, in English reading, the window is skewed to the right (Rayner, Reference Rayner2009).

In the field of visual search, researchers have long assumed that observers focus on objects that share featural similarity to the target (Duncan & Humphreys, Reference Duncan and Humphreys1989; Wolfe, Cave & Franzel, Reference Wolfe, Cave and Franzel1989). This assumption was confirmed by eye movement research which demonstrated that searchers are more likely to fixate objects which are similar to the target object (Becker, Reference Becker2011; Luria & Strauss, Reference Luria and Strauss1975; Rayner & Fisher, Reference Rayner and Fisher1987; Stroud et al., Reference Stroud, Menneer, Cave and Donnelly2012; Williams, Reference Williams1967), and this is true both in terms of visual and semantic similarity (Godwin, Hout & Menneer, Reference Godwin, Hout and Menneer2014). Without the recording of eye movement behavior, such confirmations of long-standing assumptions by theories and models alike would not have been possible.

So far, we have provided examples wherein utilizing eye movement analyses either confirmed or extended what was assumed by theories and models. But there are other examples which we can draw upon which demonstrate that, by paying attention to eye movement behavior, we can also adjudicate between competing theoretical arguments. For example, Wolfe and Horowitz (Reference Horowitz and Wolfe1998), on the basis of a paradigm wherein a search display regularly changed the location of all objects, argued that ‘visual search has no memory’ for already-inspected objects. This was a strong and controversial theoretical claim. In response, eye movement studies served to discount the claims made, finding that the rate of revisits to already-inspected objects was low (e.g., see Kristjánsson, Reference Kristjánsson2000; Peterson et al., Reference Peterson, Kramer, Wang, Irwin and McCarley2001), and that Wolfe and Horowitz’s (Reference Horowitz and Wolfe1998) paradigm instead caused participants to ‘sit and wait’ – i.e., hold their eyes still – rather than to actively search for a target (Geyer, Von Mühlenen & Müller, Reference Geyer, Von Mühlenen and Müller2007; Von Mühlenen, Müller & Müller, Reference Von Mühlenen, Müller and Müller2003). In this example, the study of eye movement behavior actively contributed to correcting the scientific record.

There are many other examples that demonstrate the value of paying attention to eye movement behavior, particularly in the context of inquiry into peripheral vision. For example, Tatler and Vincent (Reference Tatler and Vincent2009) reported the surprising finding that individual differences in eye-movement behavior are more predictive of information extraction from a scene than are bottom-up salience maps. Elsewhere, findings regarding change and inattentional blindness, which were deemed surprising or even shocking, were often explainable in terms of where participants were fixating (e.g., see Slattery, Angele & Rayner, Reference Slattery, Angele and Rayner2011). In other cases, eye movements ruled out fixational (and peripheral crowding) explanations for these phenomena (Memmert, Reference Memmert2006; Smith, Lamont & Henderson, Reference Smith, Lamont and Henderson2012). In all cases, unique conclusions were drawn that have a direct bearing upon theory that would not have been possible without studying eye movement behavior.

The examples we have included here highlight that the study of eye movement behavior is not simply a methodological choice and should not be disregarded, as is advocated in the review. Instead, we argue that by studying eye movement behavior, we can gain nuanced insights that can confirm or disconfirm hypotheses in a manner that would otherwise have been impossible. To take another final example, researchers explaining the function of the brain would struggle to do so without examining the brain itself; researchers studying human vision may struggle to do so to the extent that they desire without studying eye tracking. Indeed, by deciding to ignore eye movement behavior, our field opens itself up to future crises. Those crises may arise because we collectively chose not to check (at least in part) our theoretical predictions and assumptions by examining how, where, when and why observers acquired visual information from the displays presented to them. We are not advocating that all visual-cognitive researchers engage in eye-tracking. Rather, we argue that our field disregards eye movements at its own peril.