How age, gender and sexuality relate to your preference for ATLA or LOK
Intro
In January 2016 @avatarsymbolism ran a survey yielding 951 responses (summary of results and list of all questions available through this post). The summary of results is just that – a basic summary of all responses. But contained within the data is a story, one of trends and behaviour and complex interactions. While the last 7 months of my tumblr has been dedicated entirely to worldbuilding, I’d like to introduce another type of post because I find behavioural data analysis therapeutic, thoroughly interesting and fun.The way to read the graph is that the percentages (y-axis) represent the thing in the legend by the thing on the x-axis. For example, looking at “ATLA/LOK by Age and Gender” we can see that just over 65% of females aged < 18 prefer ATLA over LOK. This’ll be true for pretty much all the charts I make unless otherwise stated.Importantly, I’m using percentages. We will be in constant danger of coming to false conclusions. Generally speaking, we’re looking for obvious trends and reversals in trends. There are methods that provide statistical evidence for/against any conclusions we might come to, but that takes more time than its worth for me. In any case, I’ll do my best to indicate when I think the charts could potentially be misleading (from a purely statistical standpoint).
ATLA or LOK?
This is the starting point in our story. This is the information that a summary of results will show you but it's not the end of the story.
ATLA/LOK by Gender
It’s at this point that I must tell a sad story. Of the 951 responses, 852 were used to make this chart. Each time I introduce a new variable (ie. question asked in the survey), some responses are by necessity excluded. Two reasons: a. the question went unanswered; or b. the answer was given by text input and I wasn’t able to put it into a category. I manually went through the column of all answers to “What is your gender?” and either 1. put it into a category which, to me, seemed most appropriate; 2. left it blank; or 3. put it into the Other category, which isn’t used because it’s too small. For example, I labelled “Bigender (Female/non-binary)” as female, while “bigender” went into Other. This process, where I in all my ignorance applied my own labels, repeated itself for questions of sexuality and race. The harsh reality is that this sort of analysis is a numbers game, and I need to consolidate people into categories. I apologise to anyone whose response I inaccurately categorised.
ATLA/LOK by Age and Gender
This is one of the reasons I love this sort of stuff. While the ‘by Gender’ chart told us females and males have opposite preference (not terribly surprising), this chart clearly shows that as age increases the preference for LOK increases. Additionally, we can see that this is independent of gender effects.My reasoning is that older people relate more to LOK. What’s fun here is that I could be totally wrong. The numbers can tell us something is happening, but it’s still up to us to work out why we see what we see.
Age, sexuality and gender
Bisexuals and homosexuals break the female trend and, overall, prefer LOK. As a future post will demonstrate, this is indeed linked to Korrasami. What’s interesting is that 0-18/Bisexual/Females break the trend of other female bisexuals and prefer ATLA. My reasoning is that the relatability of ATLA’s younger characters is more of a draw than the older Korrasami pair. But if this is the case, why don’t we see the same pattern in 0-18/Homosexual/Females?
Please note that 31, 49 and 19 people represent the categories homosexual female <18, 19-24 and 25+, respectively. We can certainly say that, overall, homosexual females prefer LOK, but I don’t think we can make a conclusion about the age effects in homosexual females.
Please note that Unsure 25+ contains 6 people, so it doesn’t really represent anything. Additionally, both Unsure/Female and Asexual/Female had less than 68 and 77 responses respectively, so take these with a grain of salt too.If you’re wondering why males aren’t subdivided, or why ages for only female heterosexuals and homosexuals are displayed, it’s because there aren’t enough numbers in the other groups to make separation by age useful. It’s worth mentioning that of the 951 responses recorded, this particular chart uses 781. In particular, people who identified as panromantic or demisexual were not included because the numbers were simply too low.
Full disclosure: I don’t have permission from @avatarsymbolism to post my analyses. I queried them about using their publicly released data and received what I consider curt replies. I release my analysis on the basis that:
A. 951 people took a really long survey and deserve to see what’s capable with the data collected, even after 1 and a half years
B. I think there are people that are interested in this kind of ‘communal introspection’
C. @avatarsymbolism didn’t engage or reason with me despite my best efforts to have dialogue (if you’re reading this, I still encourage you to talk to me)
