Study Results (1/??)
I'll post some of the results now! Firstly, I'll write about how I processed the data for those interested in details. (Please inform me if you see anything wrong or want to suggest a different method. I'm just an average person with an undergrad degree...)
If you just want to see the analysis, skip the data cleaning and test assumptions part! As a disclaimer, I don't know if this is all for now or if I can bring myself to work more on it soon 😭
Special thanks to @xavieslittlestar and @m00nchildwrites for sharing this on bluesky and twitter
Data Cleaning
I realized that some of you guys didn't do the test and wrote the numbers by yourself 😊 And it took me hours to clean the data
As expected, there were some typos, and I corrected them before deciding which participants to exclude. For instance, some wrote 29/20, and since the sum of the subfactor scores is equal to your main factor score, I was able to confirm that it should be 19/20 by doing main_factor_score—other_subfactors = target_subfactor.
After dealing with typos, all participants who entered random values like 0000 and values that didn't align with the main factor scores were excluded.
In the end, the sample size was reduced from ~2600 to 1929.
Test Assumptions
The normality and homogeneity of variance were checked. Most of them fulfilled the homogeneity of variance(Levene's Test), but normality was problematic. I assessed normality based on Kolmogorov-Smirnov(KS), kurtosis, skewness, and Q-Q plot. There were some outliers, but most of them were a natural part of the data, so I didn't change anything about them. Although KS values were lower than .05 , I assumed normality if kurtosis and skewness seemed average (-2 to +2 for kurtosis, -1 to +1 for skewness). For others, I tried trimming, LOG, SQRT, Z, and LN transformations, but still couldn't normalize... Winsorizing would probably work, but there are too many participants(yes, I'm lazy). So, I'm gonna be relatively liberal about normality.
Results
Please note that all of them have small effect sizes (0.2 Cohen's d or 0.1 η2). So regardless of the p values, the actual association between your personality based on the big five and character choices in lads might be small in degree. Also, this is an observational study, which means you shouldn't try to infer a causation or a strict relation (so use it just for your enjoyment, not for fandom chaos~). But I'll put my personal comment under them because I need to enjoy myself too lol
Also, I realised that some of those are changing or disappearing when I compare poc(people of color) and non-poc within themselves. Therefore, the effects might be moderated/mediated by cultural differences as well.
===Trust====
Sylus mains had the lowest mean score in terms of trust(M=11.2, SD=3.42). Yet, they significantly differed only from Xavier mains who had the highest mean(M=12.1, SD=3.28).
Personal comment: I don't know what to make of it. But let me put those here as well.
==Conscientiousness and Agreeableness==
I'll write this together because I have a theory about these results.
Caleb mains' scores were the lowest in terms of conscientiousness (M=74.3, SD=13.7) and agreeableness(M=84.4, SD=13.8). For conscientiousness, they significantly differed from Zayne and Sylus mains. And for agreeableness, they significantly differed from Zayne and Xavier mains. [Agreeableness showed unequal variances, so I used Kruskall-Wallis test]
Personal comment: Honestly, I'm surprised. (Caleb and his girlies not beating red flag/villain allegations)
If anything, I would expect Sylus mains to have lower agreeableness and conscientiousness (first impressions don't disappear).
Then I thought, what if it's related to Caleb being MC's adopted brother? Listen before shooting me. To feel at ease with this setting, you should have higher (but healthy) dissociative tendencies to detach yourself from its real-life associations. And guess what— low agreeableness and low conscientiousness are related to high disassociation [refer to:https://www.sciencedirect.com/science/article/pii/S1130527417300373]. Secondly, having a lower need to conform(related to agreeableness and conscientiousness) with others might be making it easier for some people to choose Caleb as their main love interest.
(But as I said, these results are liable to change. For instance, for the non-poc sample lowest mean values of conscientiousness belong to Rafayel mains)
===Emotionality===
Caleb(M=14.7, SD=3.49) mains had the lowest emotionality score, and the difference between them and Zayne(M=15.6, SD=3.04), Rafayel(M=15.5, SD=3.11), and Sylus(M=15.5, SD=3.33) mains was significant
Personal comment:......Babes lolololol. Is Caleb's robot alter taking over his mains?? Did they get a Toring chip as well? It's so funny omg
===Self-Discipline===
Rafayel(M=10.1, SD=2.96) and Caleb(M=10.1, SD=2.97) mains had similarly low scores of self-discipline, and they differed from Zayne(M=10.8, SD=2.87) and Sylus(M=10.8, SD=2.79) significantly.
Personal comment: As a Rafayel main, I can confirm this.
=Second Love Interest=
Most of Zayne mains chose Sylus as their second love interest. The other most preferred second love interest was Caleb, but since this test was done just before Caleb's new myth, the results might be affected by our excitement for it. Also, Zayne mains showed low preference for Rafayel.
A note about less significant/non-significant observations:
As a whole, we have low extraversion and average conscientiousness, as well as high openness to experience, neuroticism, and agreeableness
For subcategories, we have the highest mean for imagination and the lowest mean for gregariousness. (perfect combination to date fictional men)
Sylus and Zayne mains had similar mean scores in many main factors and subfactors of the Big Five test. Likewise, Rafayel and Caleb mains' scores were close to each other, and they contrasted with Sylus and Zayne(not always significantly tho). Xavier mains generally had balanced scores.
None of the groups differed significantly in terms of openness to experience, neuroticism, extroversion, and these factors' subcategories.
To check if the data gives standard results, I checked socioeconomic status(SES) and its relation to neuroticism since previous studies indicate higher neuroticism for low SES groups, and it was the same for this dataset as well. Anxiety, depression, and neuroticism showed a slightly decreasing trend as the SES increased.
Some studies suggest that people of color have lower trust scores, but in this study, poc had higher trust than non-poc. However, it might be due to the differences in exposure to discrimination between poc who live in the US and those who are natives of other countries.
Tag list: @dadddybangtan @gingers-random-junk @cloudyasteria @eoe-1379 @kwtdrn @punksausages @satorusfrontallobesilverhair @xanxann01 @dandellien @auraficial @fictionalmenlover5 @bundle-of-sunlight @starryfilled @cupcakefactory @rayamalaya @mandapanda16 @nouerzzz @jonggunkitten @fckkntired @dreamienebula @tiffyelefano @bbnique @maybeyougotmewrong @starrychxn @situationsheep @maimaily @irlsammy @meowumis @piranha-teeth @svnflowery
The results are interesting! I did the survey as a Zayne main with Sylus second. The results of my big 5 test were:
