Eye tracking research; PaperSense 1. The effect of different media and human presence in pictures.

This is a non-academic publication, based on my personal reflections on a highly interesting research project that I have been partly involved in.

PaperSense – a research project  investigating the differences between how we perceive pictures when they are presented on a screen, or on paper, and how the presense of a person in the pictures influences what we look at.

The papersense project covered many aspects of information and marketing in various media.
The pictures used for the test were provided by IKEA and are copyright Inter IKEA Systems B.V.

PaperSense - pict 08. With people.
The pictures are presented further down. Each picture is a combination of the original picture, a diagram, the version as presented on screen, and lastly the paper version. Original pictures copyright Inter IKEA Systems B.V.

I was involved in a part of the project, where;
– we investigated the influence of the actual media on which test pictures were presented, i.e.; on a computer screen or printed on paper.
– at the same time we made a comparison with a person present and not present in the pictures.

To better understand the underlying processes, eye tracking was used on test persons to measure eye movement and visual attention when viewing different pictures. To complete the investigation, questions were presented to further collect data about the pictures presented and their effect.

To read more about eye tracking, please see page What is eye-tracking.

The pictures were always presented in pairs; one picture without people, and a version with people present.
The pictures were chosen to represent different levels of complexity, a room setting was used to represent a complex picture, and a product picture to represent a simpler composition.
The pictures were supplied by IKEA and some appeared in the catalogue around 2009 – 2010.
Two test groups were used. Each group consisting of 30 women between the ages of 25-35 One group viewed all the pictures on a computer screen. The other group viewed the same pictures printed on paper. Each picture was presented for 5 seconds.

The results.
The collected data from the eye tracking was analysed and the amount of attention spent on each picture was evaluated. This is presented as heat maps, which reveals how much attention the eyes spend at different areas of a picture. It is assumed that a high amount of interest (yellow and red on the heat map) tells that a person might be interested in what they focus on, but it might also indicate that the person can not figure out what is being reviewed.

The results showed that when the pictures were presented on paper, measurements indicated a significant increase in attention, implying that the test persons took in much more information than when the same pictures were presented on a computer screen.
Differences in attention for pictures presented on paper was sometimes more than plus 50% compared to screen presentation.
When people were present in the picture, the attention was sometimes more than plus 75% more compared to the pictures without people.

In this test, it was obvious that we seem to take in more information from pictures when they are presented on paper. The test, however, does not say anything abut the reasons for this somewhat overwhelming fact.

Eye tracking as such is very interesting in relation to picture composition. I think very much is revealed when we see the heat map clouds that show where the eyes often focus, while the areas with low attention can be equally interesting.
The presence of people in the pictures is also exciting. Much more information was taken in when people were present in the pictures, with most focus on the faces, less on the surroundings.
The test persons also answered a questionnaire which showed that the pictures with people, were regarded as warmer than without.

One reaction I got when showing the results of the tests, was that there should be no people at all in pictures when one wants to sell furniture. On the other hand it might very well be that the presence of people is what makes a person actually see a particular picture in a flow (we very seldom see pictures isolated, single handed) and possibly see the products that we hope to make them realise are needed in their daily life.

Two of the pictures and some reflections.

I see some important lessons from the tests;
– pictures presented on paper possibly transmit far more information to the viewer.
– pictures containing people attract much more attention from the viewer.
– people in pictures create a warmer atmosphere.

About the pictures: Each picture is a combination of the original picture, a diagram of the attention values and the superimposed heatmaps.
Please note; the heat maps are always relative, meaning that they show where the eyes looked the most in a particular picture, indicated by red colour.
But you need to look and evaluate the calculated attention to see the actual total amount. The red in one picture may indicate 100, and in an other picture 1500.
PaperSense combined picture explained

A medium complex picture, with people.
PaperSense - pict 08. With people.

The first version has people, children, present in the picture. The heat maps on both the screen version (big, left) and the paper version (big, right) clearly show that the faces of the children attract most attention in the picture. The calculated attention value is 666 for the screen version, where the paper version shows a value of 992, which is plus 49%.
The clouds of the heat map from the paper version show that attention has been paid to other areas of the picture as well.

The second version of the picture is without people.
PaperSense - pict 08. No people.

The heat maps are similar to the previous version of the picture, which had children in it. This is possibly due to the composition which has a lot of red colour in the centre, where the children were in the former picture. The sofa is the natural centre in the composition.
In this version the eyes have moved a little more around the scene, especially true for the screen version.
It still seems obvious that the red coloured sofa has gotten most of the test persons attention, which possibly is the aim of this picture.
The calculated amount of attention has fallen down to 35% in the picture shown on screen. A rather low figure.
The picture shown on paper has now reduced to 65% compared to the version with the children. A viewer would possibly move on to the following pictures without paying much attention to these pictures. In this case, one can conclude that the picture without people on screen is rather uninteresting.

This picture is a more product like picture, but still somewhat complex. Partly with people.
PaperSense - pict 11. With people.

The first version has people present in the picture, although not very visible. The heat maps on both the screen version (big, left) and the paper version (big, right) shows that the hands of the person attracts the most attention in the picture (and even a little on the shoes).
The calculated attention value is 254 for the screen version, where the paper version shows a value of 699, which is plus 275%, a rather big difference.

The calculated attention values however are low compared to the previous picture, the red sofa with the children.

The second version of the picture has no hints of people in it.
PaperSense - pict 11. No people.

The heat maps are similar to the previous version of the picture. The composition is supposed to draw attention to the open drawers and their function, which is confirmed by the heat maps.
It is a little strange to see that the calculated amount of attention now is almost the same in the picture shown on screen, regardless of the presence of the person.
The picture shown on paper, is now reduced to 74% attention compared to the version with the partly visible person.
I also find it interesting to compare this picture, to the picture with the red sofa without people, where the interest rate is almost on the same low level.

One obvious conclusion is that when the person in the picture does not reveal their face, the attention to the picture is lower. The concealed person is now an attribute amongst others in the picture.

This was the first article about the papersense project.
More pictures and comparisons are to come.

Leave a Reply

Your email address will not be published. Required fields are marked *