Meet-up Group Image Analysis

First of all, I would like to thank everyone who took part in this exercise. It has been very useful to my researches.

I asked 15 friends, who are members of the Edinburgh Photography Meet-up Group, if they would judge 32 of my recent images against two different criteria.

The objective for the exercise is to understand more about my own photographic thinking and evaluation processes by asking a group of other photographers to assess a set of my recent images against:

  • their levels of “interest” in the image content, defined as intrigue or narrative potential.
  • their personal preferences for each image.

I chose a variety of my images, taken on a visit to Iceland but of varying subject matter and mood which ranged from the spectacular to the banal. The group were asked to ignore print quality (which was more variable than I would have wished) and to focus only on content and composition.

I first looked at how the Meet-up group average level of “Interest” compared with my own assessment of “My Interest” in each image. The following results covering 32 images:

On this chart, any difference less of ~20 or less is not significant. Therefore, on 6 images my own interest level is significantly higher than that of the group and on just 1 image my interest is significantly less than that of the group.

I reviewed these images and concluded that the ones where I showed more interest than the group were all ones where I had had significant emotional reactions to the location. These particular scenes had evoked memories or narratives for me at the time of taking. Obviously, this could not be true for anyone else just looking at the resulting images for the first time.

The one image where my own interest was low relative to that of the group was one I took because of the dramatic lighting and not because of any emotional or narrative content for me. This was an image I would classify as “spectacular” rather than meaningful. It was therefore weak against my own criteria for narrative content. However, 8 members of the group scored this image high for “Interest”. They obviously saw more content in the image than I had.

I then wondered about how the individuals in the “group” had scored for interest on each image. Was the averaging of their scores hiding their real levels of interest or not? Here is a chart of the individual scores for each image.

My concern was correct. By averaging the group’s individual scores, I was actually losing the variety in the opinions of the group members. This chart shows that for each and every image there was a range of at least 4 points (on a scale of 5) between the highest and lowest scores. So, levels of interest varied widely. Consequently, averaging group scores is not a satisfactory form of analysis.

I then decided to look how each individual group member had scored each image but focusing only on those images where I had given the strongest positive or negative scores. These were images #4, 5, 13, 14, 16 and 3.

For these images 5 people agreed with my scoring for 2 or more images and one person agreed for all 6 images. Consequently, my images were proving to have some degree of intrigue and or narrative potential for around 1/3rd of the participants.

I then moved to looking at the scores for image “preference”. That is, on a scale of 1 to 5 how does each image match my own standards for preference?

This first chart shows my personal scoring for each image against my view of both Interest and Preference.

I remembered what I had observed looking at group interest versus my own interest.

Thinking about the information hiding effect I had observed by averaging the group interest scores, I wondered whether my own Preference scores were hiding subtleties. Was my Preference score actually made up of 2 or more separate strands that I had not yet identified?

After reviewing my images again, I concluded there were two key strands and possibly more, that I was unconsciously using when scoring for Preference. They were:

  1. Would the image be appreciated by other photographers?
    • In other words, not my own preference but my ideas of others’ preferences. I am seeking the approval of others!
  2. From my own perspective, had I succeeded, in turning the banal into a piece of art with a significant narrative potential?
    • Now, this is far more about myself, my real photographic preferences and goals.

I now re-scored for my own preferences in terms of these two, now separate, criteria which I have labelled “Others Preference” and “Banal Art”.

Immediately, there is a significant difference between these results and those where I had looked only at my “preference” as a single, broad category.

I chose to review only those images where I scored it as a 5 for Banal Art. These are images where I am particularly happy that they meet my artistic preferences for Banal Art.

Result – There are 6 images that meet my criteria and yet I believed these are not images that would be generally appreciated by other photographers.  They are images #2, 4, 8, 9 13 & 16.

Now to compare my “Banal Art” results with the group’s results for preference.

Again, the group scores varied greatly and so averaging was not appropriate. Therefore, I looked for comparisons against individual group members.

Here 6 people agreed with my “Banal Art” preferences for at least 4 of the images. Consequently, when I am truly succeeding against my own goals (rather than aiming to please others) ~1/3rd of the group are also showing a preference for these works.

So, the appreciation for my “Banal Art” images is greater than I had suspected.

Finally, I compared my highest scores for “Others’ Preferences” with the individual scores given by the group for the same images. Unsurprisingly, as this is where I am judging my images for their appeal to others, there is a strong correlation between my scores and the group scores.

I looked at all the images that I scored as 4 or 5. That was 7 images, numbers 1, 6, 15, 19, 26, 31 & 32.

Result – for all 7 images, 9 to 12 members of the group also scored them highly. So that means I do have a good sense for the types of images that will appeal to other photographers, despite that creating that type of image is not my primary goal.

Conclusions – The objective for the exercise was to understand more about my own photographic thinking and evaluation processes.

My most significant learning has been the separating of my own “Preference” into the two distinct categories of “Banal Art” and “Other’s Preferences”. I now have a means to test my images as to whether they are meeting my real internal goals or just being created in the hope of gaining the approval from others.

My identification of the word “banal”, as applied to my images, has also opened up a new and rich area for academic research into the work of other photographers.

I have also discovered that, amongst photographers at least, images that I consider to be banal art do create a satisfying although certainly not universal, level of interest and intrigue.

Finally, I have learned that using averages when the same image can evoke such wide-ranging views hides rather than revealing information.

Thanks, once again, to all the participants. It was your help in creating the raw data that has guided me towards questioning my own thinking processes and so to finding these results.

1 thought on “Meet-up Group Image Analysis”

Leave a Reply

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