We’ve discussed attentional blink several times on CogDaily. It’s a fascinating phenomenon: if you see a series of images flashing by rapidly, you can normally pick out one of the images (for example, a banana in series of pictures of familiar objects). But if a second such image (another piece of fruit, like an apple) appears shortly after the first, you’ll probably miss it. The one exception in many cases is faces. This video illustrates the point:
You probably spotted both the piece of fruit and the face in the second sequence, but you may have missed the watch in the first sequence.
The first sequence is a classic attentional blink: If the second image they are looking for flashes between 200 and 400 milliseconds after the first, people don’t seem to notice it. It’s as if there’s a literal blink in our attention to the pictures. While faces have some “immunity” to attentional blink, that immunity can be overcome if both of the pictures being searched for are faces. If you’re asked to indicate whether one or two faces were shown in a sequence of pictures, you’re still likely to miss the second one: attentional blink still occurs.
Recently researchers have found an exception to this exception: If the second face being shown is expressing fear, and the first face is neutral or happy, then attentional blink is again diminished. So fearful faces appear to be more immune to attentional blink than other faces. But this research has been questioned because fearful faces might give other cues to observers — they may be darker or have a different skin tone compared to non-fearful faces.
So a team led by Frances Maratos devised a simpler version of the task, using schematic faces instead of actual photos:
These were placed in a series of scrambled faces, like this:
Sometimes there were two non-scrambled faces in the sequence (which was about 20 images long), and sometimes there was just one. 23 students watched hundreds of these sequences, indicating each time whether one or two non-scrambled faces were seen, and what the emotion depicted in the last face was. When only one face was shown, viewers were about 80 percent accurate no matter what emotion was portrayed (though they were a little better for happy faces than for the others). But when two faces were shown, the results were different. Here are the key results:
The graph shows how accurate viewers were at identifying the second face at various different lengths of time following the first face. While there is an attentional blink for both angry and neutral faces, the blink is much larger for the neutral faces, and angry faces are identified better no matter how long the lag between the first and second faces.
So Maratos’ team confirmed the results of earlier studies on photos of faces, showing that some emotions are more susceptible to attentional blink than others. In addition, they demonstrated that the effect is due to the emotion being depicted, rather than some other artifact of the pictures, because each picture was built up of similar schematic components.
Maratos, F., Mogg, K., & Bradley, B. (2008). Identification of angry faces in the attentional blink Cognition & Emotion, 22 (7), 1340-1352 DOI: 10.1080/02699930701774218