Quite a while ago now, I looked at the first 100 image search results returned by Google for the phrase “disabled person”. It had some interesting results and I fully intended to do a similar thing for “Autistic Person” and then life got in the way and both my blogs went dormant for a while. Now that I’m back, I thought this was as good a place to pick up as any.
This was a little more complicated than my previous casual research of Google Images, as these were some of the front page results I got when I google “autistic person”:
Now – armchair diagnosing is quite common when it comes to autism but even so, I’m quite certain that I have never heard anyone say that they think 50 Cent is autistic. Clicking on that image and checking where it came from reveals that it is from a news story about 50 Cent insulting an autistic person. Mark Zuckerberg and Bill Gates, on the other hand, have a considerable number of people both arguing for and against the possibility that they might on the spectrum. I didn’t have the same consideration when I searched for “Disabled person” because there weren’t armchair diagnosed people in the first 100 images.
When it came to deciding which people to include in my informal research of google images, I ended up a bit uncertain. After all, even if it were armchair diagnosing, individuals such as Bill Gates are often upheld as being figure-heads for autism. Surely this should be counted when thinking about representation?
In the end I decided to run two tables to track my data – one including armchair diagnosed celebrities (to see if there was a change in data when they were included) and one not, and to use the stock image representations on website at face value as those of autistic people.
Another difficulty was the topic of gender – as I discussed in the last post, I was having to make assumptions about whether the person in the image was female or male based on how they presented in the image. This was a massive assumption, I could be wrong, and it’s not always possible to find out which gender someone identifies as. So this one I was just going to have to take the hit on potentially getting it wrong and make changes if/when corrected.
So what did I find? Well – in the first 100 photographs returned on Google images of autistic people, the individuals are predominantly white, predominantly male and predominantly adults.
Now whilst the first of these two statements may have been met with some head nodding that showed little in the way of surprise – the latter may have prompted at least a pause. We hear over and over that the stereotype and media go-to for autism is a child, yet adults were by far the largest age demographic amongst the Google images. Though, given the recent media alteration towards autism = white, adult male who is gifted in STEM subject, perhaps this shouldn’t have been a surprise.
There were 9 images which were of people who were armchair diagnosed as autistic – all 9 were male and 8 of them were adults with 1 elderly. The difference when these 9 were removed and replaced with the next 9 images in the search results was minimal – though the skew towards male representation was marginally reduced from 87.50% to 84.30%.
When we add age to the mix adult men make up 44.17% of the images, followed in number by male children with 25.83% and then adolescent boys with 13.33%. Again, you can see that removing the 9 images as mentioned above reduced the skew towards male slightly.
And then looking at everything together:
Images of white, autistic men made up 37.50% of images; images of white autistic male children made up 24.17% of images; and white autistic male adolescents made up 10.00% of images. Removing the 9 images discussed above changed these percentages to 31.40%, 25.62% and 10.74% respectively but did not change the top three results.
In the first 100 non-cartoon and non-repeating images returned by Google images, 86.67% of the people with autism were white, and when the 9 images were removed this lowered slightly to 85.12%.
So in a very unscientific and horrifically flawed piece of time-wasting research, it would seem that the biggest representation for autism (at least on Google image returns) is white, adult men and representation of women and people from BME groups are very under-represented. Which pretty much matches up to current climate.
*NB: I am aware of how unscientific this is and that there are massive flaws in my methodology. This was never meant to be a serious piece of research but more a fun way for me to play around with excel and look at autism representation.