Image Poll #1
Image Poll #1 – Aesthetic Preference and Preference Modification Experiment
This two step experiment aimed to determine the factors influencing respondents to revise their initial choices once the creator of an image was disclosed. Respondents participated in a straightforward aesthetic preference test. They were presented with a set of four images, with no prior information except that these images were crafted by relatively renowned individuals. The following is a transcript of a poll shared on WhatsApp and Facebook through the author, Karyn Easton’s social media network.
Step #1 – A Simple Aesthetic Preference Test
Step one – In the first instance participants were invited to choose a single image from the images above based on limited information. This was to determine each respondents choice. Step two involved revealing key information about the images above and then inviting respondents to revise their choices.
WhatsApp and Facebook Post 1
Below is a transcript of the Facebook Image poll and the images just as they appeared on the WhatsApp and Facebook posts. The following poll ran along with the accompanying text among some of MA students in a WhatsApp group. As the respondent numbers were low this exact poll was repeated on Facebook.
“Today I’m working on some primary research for my research paper. Below are four images by relatively famous artists – I’d like to buy a print of one of them for my new extension as they are all pretty well known. Please can you help me choose which one would you go for? Apologies in advance of the image quality, but colour and composition are probably more important to me than the finer details – Thanks” 😊
The results were as follows:
At this stage the respondents were completely unaware of who or what created each image. Out of a total of 47 Respondents the breakdown of image popularity was as follows:
- Image 2 was the most popular, with 23 votes (49%) – created by a pig called Pigcasso
- Image 1 received 12 votes (25%) – created by Jackson Pollock a human
- Image 4 secured 7 votes (15%) – created by a GAN or text to image computer generation
- Image 3 was the least popular, receiving 5 votes (11%) – created by Ai-Da the humanoid robot that creates art
Step #2 – A preference Modification Test
The second stage in the this experiment was to establish whether or not additional information would persuade respondents to revise their choices in light of the new information.
In this second step of the experiment the creators of each image were announced and based on that information the respondents were asked if they would like to switch their choice of image to another image. This was to see whether or not the actual creator of the image would have a bearing upon image preference.
Respondents were given the following information, (although since posting the information it has been subsequently updated above with the latest figures). Based on the new information they received regarding the creator of each piece of art, respondents were asked if they would like to change their image preference.
Facebook Post 2
Below is the second part of the experiment to see whether additional information would encourage respondents to revise their choices. Of particular note was whether or not there would be a revision away from artificial intelligence towards human generated art particularly in light of the known negative bias towards AI generated art. The post text was as follows:
“So what was it all about? I’m about to reveal who the artists were, but what I really want to know is that with the extra information I’m about to give you all would this sway your preference. What would your favourite image be now? Knowing what I’m about to tell you. If you choice remains the same please comment with same and if you choice of image has changed please say what number you moved from and your new favourite number now.
Here goes….
Out of 40 Respondents
Image 2 stole the show with a whopping 20 votes a cool 50% of the vote. This image was drawn by an artistic pig called – wait for it ‘Pigcasso’. According to the internet her owners have sold over $1 million dollars worth of paintings with a single painting fetching around $26,000 each!
Image 1 came second with 11 votes and 27.5% share of the vote. The artist who created this image was Jackson Pollock who up until 2011 had the highest price ever paid for a painting at a staggering $140 million dollars.
Image 4 came in 3rd with 6 votes giving it 15% of the vote and this was created by a computer with early form of text to image GAN (General Adversarial Network). A computer generated Ai image if you like. This sold at Christie’s in New York for $432,500.00
Image 3 came last with only 3 votes. This image was created by AI-DA a humanoid art creating AI Robot. Her works have sold in excess of $1 million dollars according to the internet.
What I’d like to know is knowing this information have you changed your mind and I’d love to hear your thoughts? Which image do you prefer now? Or do you still feel the same?”
The Results
This time only 19 people replied. This could be because Facebook’s algorithms may not have shown the new poll, or perhaps people didn’t want to reply, or simply people were unaware of the poll because they don’t frequently use Facebook. However, below are a list of the comments from the respondents. Only one person wanted to revise their choice because they liked the idea of an art creating pig. Everyone else kept their original choices.
Comments by Participants Image Poll #1
- Still feel the same prefer number 2 good artist 😂 x
- Still 4 for me
- Still feel the same – shows that with all the AI intelligence you still can’t beat a talented pig! Still no. 2
- Still 2 for me. It’s organic on lots of levels!
- I’d say the same, 2 ( he says with a grunt ) 🐷🐷 xx
- Still 2 bless the pig.
- I’m happy I went for the Jackson Pollock, shame it’s not hanging on my wall £££££££😂
- I think some people have more money than sense🤣. Still wouldn’t have any of them in my house …for long…I’d sell them and buy some decent 🎨 art! 😁 I’ll stick to my choice
- Same choice x No.2 as before
- Fascinating, thank you 😊 x Karyn Easton still keep it the same 💗
- I’d stick with my choice, I liked it because of the expressive marks, energy and use of colour, which I think AI, cannot replicate… I personally find AI images very soulless, the human element is missing. They’re a bit like the makeup in 70’s movies. The piggy’s effort isn’t bad, thier DNA is very similar to ours 🧬.
- Tbh I was torn between 1 and 2. I went for 2 because is was less messy than one. Still go with my original choice x
- I would stay with no. 3
- I will stick with 4, but I don’t really care for the scale, framing, lack of concept or absence of title in any of them. Process is important in art, and digital AI has to be the most remarkable statement of our current world. The most cutting edge.
- Same, 2, but I didn’t like any of them. 2 was the best of a bad bunch but I did like the colour palette compared to the others.
- I’ll stay with 2.
- Glad I chose 1 and will stick with it. It seemed the most complex in concept and had the most appealing colours.
- The same. Just prefer the colours.
- Visually prefer 4, but I want to swap to the pig!! Karyn Easton I’m not artistic at all. But I love the pig because it’s silly and still holds its own amongst the others! I don’t really ‘get’ modern art, but I’m more intrigued if it was created by a human expressing something than AI. It suddenly felt soulless- even if I preferred the image. So I guess there’s an intrinsic value to the art even if it’s not my preferred taste. At those price tags though I’m gonna see what my cat can do with some paint?! 😉X
Results From Image Poll #1
Following are the conclusions drawn from Image Poll #1 Based on the following data and responses outlined above:
Results:
- Image 2 was the most popular, with 23 out of 47 votes (49%).
- Image 1 received 12 votes (25%).
- Image 4 secured 7 votes (15%).
- Image 3 was the least popular, receiving 5 votes (11%).
Insights:
- Inherent Appeal: Without knowledge of the creators, Image 2 (artwork by Pigcasso) attracted the highest number of votes. This suggests that the visual appeal of the artwork itself played a significant role in its popularity. This could imply that the style and composition of the artwork resonated strongly with the respondents.
- Neutral Ground: The fact that Image 2 remained the most popular even with the context of the pig artist, reinforces the notion that art can be appreciated without the biases associated with art creators.
- Art for Art’s Sake: The lack of prior knowledge of the creators underscores the idea that art can be appreciated on its own merits, divorced from the artist’s reputation or backstory.
- Unbiased Preference: The preferences expressed by respondents both with and without prior knowledge of the artist, emphasises a level of unbiased judgment, suggesting that the artwork itself was the primary determinant of their choice.
- Diverse Appeal: Different images resonated with different participants, showcasing the diverse tastes within the group. This highlights the multifaceted nature of art appreciation and the subjective nature of preferences.
- Abstract vs. Recognisable Art: The popularity of the abstract artwork (Image 2) compared to the portrait (Image 3) might imply that abstract art has a broader appeal, especially when considered without the context of the artist.
- Facilitating a Change of Mind: All respondents except one refused to change their minds over their chosen image even when the creator of each image was revealed. This suggests some level of emotional investiture in the respondent’s choices that makes them remain with their original preference.
In essence, even when the respondents knew the creators of the artworks, Image 2 remained the most popular. This reinforces the notion that the visual qualities and aesthetic appeal of the artwork itself had a significant impact on their choices, highlighting the intrinsic power of art to resonate with the respondents regardless of its origin.
Further Discussion
The subsequent question arises: What factors, if any, might compel a respondent to reconsider their initial choice, given the apparent limited influence of known negative bias towards AI-generated art in this experiment? To address this question, the next study namely Image Poll #2, delves into the potential impact of an artist’s criminal background or negative notoriety on respondent preferences. An experiment was designed to specifically investigate this aspect, the details of which, along with the results, can be found in the section dedicated to the second experiment (Image Poll #2), focused on aesthetic preference and preference modification.
Recent Posts
Categories
3d 21 Day Challenge Abstract Art ai Animation Artists Augmented Reality Books Collaborations Culture Disasters Drawings Evaluations Exhibitions Experimentation Film Production Final Major Project Final Outcomes Food for Thought Galleries Ideas Lectures Little Wins Low Residency Materials Mentions My Learning my work opportunities Patterns Photography Problem Solving Reflection Research Research Paper Journey Risk Taking threats Timelapse Tutorials Unit 2 Blogs Unit 3 Blogs Videos Workshops Zentangles Zoom Meetings
Tags
3d 21 Day Challenge Abstract Art ai Animation Artists Augmented Reality Books Collaborations Culture Disasters Drawings Evaluations Exhibitions Experimentation Film Production Final Major Project Final Outcomes Food for Thought Galleries Ideas Lectures Little Wins Low Residency Materials Mentions My Learning my work opportunities Patterns Photography Problem Solving Reflection Research Research Paper Journey Risk Taking threats Timelapse Tutorials Unit 2 Blogs Unit 3 Blogs Videos Workshops Zentangles Zoom Meetings