What to choose: eye-tracking, biometrics vs neurometrics?
We are sometimes asked what the added value of applied neuroscience is. In the bid or preparation for projects, we sometimes hear: “why can’t we just do an eye-tracking study” or “we are mainly interested in biometrics.”
Granted, neuroscience may seem as more cumbersome and it is traditionally a quite intrusive measure. Not the least when using methods such as functional Magnetic Resonance Imaging (fMRI), in which people have to lie perfectly still inside a claustrophobic tube and an extremely noisy environment, doing very repetitive tasks. Even with electroencephalography (EEG) the traditional method is using odd looking bathing caps, not something you’d like to meet your neighbour with strolling down the street…
Yet, when measuring unconscious responses, the differences between eye-tracking, biometrics and neurometrics are substantial. Here, I will demonstrate just what the differences entail. Before we do more, please consider the following chart:
By using infrared cameras to capture the eye’s movements, eye-tracking allows researchers to measure and analyse exactly where people are looking. Some immediate outputs from the eye-tracker include the heat map, which is often a powerful tool to visualise differences between groups or conditions.
Compared to biometric methods, NeuroMetrics has a superior precision and many added measures and insights.
However, to do proper analysis of eye-tracking, one should do what is called Area Of Interest (AOI) analysis. This is a method where one pinpoints specific areas that is within the primary interest of the study, such as certain products or packages, signs, pricing, or even whole shelves and even people. The AOI analysis allows us to compare things like how long it takes for people to look at the area, how much they spend overall, as well as how many times they look back and forth to this AOI.
In terms of emotions and cognition, eye-tracking can indeed measure pupil dilation, which (besides responding to brightness) is both related to emotional arousal as well as cognitive load. However, using eye-tracking in this way requires that brightness is kept constant, which rules out the use of it as an arousal measure for movies or in mobile studies. Therefore, eye-tracking cannot be seen as a measure of arousal in most conditions.
Furthermore, even when measuring emotions, arousal is only one dimension of emotional responses. Arousal can be seen as the intensity of emotional response, but it does not tell you the direction of the emotion: are people disgusted/fearful or are they happy? Therefore, any measure of emotions that only measures arousal is bound to be limited to only one dimension.
Biometric methods include methods such as galvanic skin response, pulse, and respiration. Pupil dilation is also considered a biometric measure. Generally, these measures allow researchers to assess and thereby understand what people are aroused by. Biometric methods are therefore measuring arousal.
As we saw above, this is a good measure of the intensity of emotions, but it is a poor guide to understand the direction of that response. We simply cannot see if people are responding strongly because they love something or hate it, only that they have a strong emotional response to it.
Arousal is a good measure of the intensity of emotions, but it is a poor guide to understand the direction of that response
So there is a need for combining with other measures, and some researchers and companies are using automated facial coding to address the direction and content of emotions. This may seem all good and well, but there are profound limitations to facial coding. Not the least they are often under conscious control, and can be highly influenced by the social setting and belief. For example, if a person believes that he is being observed, or in the company of others (even if he cannot see them), he will often tend to show stronger facial expressions. Facial coding in itself is a valuable method for understanding the social dimension of emotional responses, but it does not address the need for a core measure of emotional direction and motivation.
Furthermore, biometric methods are limited by their low temporal resolution and time lags. The typical measures used are often several seconds delayed, the responses observed are typically low and noisy, and the measures are affected by noise factors — for example, the GSR signal never shows a baseline, but a constant fluctuation, or “drift” in the response. Therefore, the analysis and interpretation of biometric signals are suboptimal for measuring responses to specific events. They may be good for addressing emotional states such as moods, but they are too crude to measure emotional responses to single events, at least with any high precision.
NeuroMetric methods are typically the insights offered by EEG, sometimes by fMRI. I will focus on EEG here, as it has the far most advantages compared to any other method.
Indeed, there are no real alternative to NeuroMetrics
As noted above, EEG has been relatively clunky and odd to wear, but with recent advances in technology and analysis methods, it is possible to measure emotional and cognitive responses during consumer behaviour and other types of responses.
Importantly, NeuroMetric can do everything that biometrics does, only better, and with many additional advantages:
*** NeuroMetrics measure arousal. This means that as a minimum, NeuroMetrics does the same as biometrics.
*** Granularity (time): NeuroMetrics have a much higher temporal resolution: while biometrics measure at several seconds at a time, NeuroMetrics are accurate down to tens of milliseconds (sometimes even lower).
*** Granularity (space): biometrics can only be used to assess general arousal responses to a time point (e.g., whether people are aroused during a time frame of a few seconds), while NeuroMetrics has a millisecond accuracy, and this allows us to measure responses to individual items in a scene. This can be emotional and cognitive responses to products, prices, people, events, scenes, and key communication. If multiple items are visible in the scene at any time, it is possible to determine consumers’ responses to each item individually, as well as the scene as a whole.
*** More measures: With NeuroMetrics, researchers have a larger toolbox of measures to understand consumer responses. The Neurons Inc NeuroMetric toolbox includes these metrics:
MOTIVATION – What makes people seek out more information? What drives their choices? This measure has more than 20 years of scientific support, and is closely linked to approach and avoidance behaviours.
COGNITIVE LOAD – Is an ad too complex or too simple? Does it demand too much active processing? This is a measure of the mental effort that a person is going through.
DISTRACTION & DROWSINESS – Are consumers more or less distracted by what they are looking at? Are they more or less awake during the test (e.g., watching a movie and ads, playing a game, listening to a speech)?
STRESS – Are consumers stressed in the store, or by using multiple media channels (multitasking)? This measure is combination of emotional and cognitive factors, and is highly related to whether people are stressed during certain times or generally during different phases.
Taken together, there is much to be said about the use of NeuroMetrics relative to other measures. When NeuroMetrics is combined with eye-tracking, it has an unparalleled access to consumers’ emotional and cognitive responses. In both lab tests and in store tests, these measures allow us to determine responses to specific events and items with extreme precision, and this helps us become much better at measuring, understanding and predicting consumer behaviours. Indeed, there are no real alternative to NeuroMetrics.
Do you want to understand even more about this exciting topic? Why not read Dr. Ramsøy’s textbook: “Introduction to Neuromarketing & Consumer Neuroscience” which is available through Amazon.