Documentation for the Nearest Neighbors Personality-Type Matching Survey

Premise and background

So one of the first steps is deciding close it terms of what. Self report questions, presumably, but which questions? The base dataset that I have to use is from the SWCPQ, which is a solid foundation because it has an extremely large sample size, self reported MBTI and Enneagram and includes self reports from a collection of 500 different items. Over time the NNPMS will need to develop its own dataset to include more different type systems than just the two in the SWCPQ dataset, but having this preexisting dataset we can repurpose will allow us to hit the ground running. Right now if you take the NNPMS, you can see that other identifies such as Four Temperaments, Hogwarts House, and Zodiac sign are being collected from users. These are not currently part of the test results, but will be added once the sample size gets big enough.

Selecting which items to be in the survey

So, to start off with the SQCPQ has a set of 36 items that have been chosen by pretty detailed stastical analysis to be the best set to use for predicting what fictional characters people identify with. This is probably pretty analgous to identifying with personality types so, reusing all these 36 seems reasonable.

However, it turns out that a lot of what people base their evaluation of fictional characters on falls outside the area covered by the classic personality type systems. For example, when matching to fictional characters good-vs-evil type descriptions are very important and the Myers-Briggs types were all designed to all be positive and flattering so don't capture well such concepts. Its not neccesarily a problem, still it seems like since users will be taking this test to explore personality types the items should be weighted at least a little bit towards ones that measure personality types.

Starting with Myers-Briggs, the table below shows for each of the four dichotomities, which items show the largest difference between the users who identified as one or the other.

For each of the 4 dichotomies, 5 items were select (marked with an ✅). Some items that performed well were skipped for content considerations however (⛔). For example, extrovert<--->introvert is not used because it literally repeats the criterion

Next, for the Enneagram, items that discriminated between people who identified as one type versus the 8 remaning types combined were calculated.

Again, 5 items for each type were marked for selection. Note that many of the Ennegram types have redundant items because many types appear to exist on opposite poles of an axis, for example Type 9 and Type 8 appear to be very clear mirror images of each other. Type 4 and Type 5 to a lesser extent.

Condensing the set of items from the SWCPQ basic algorythm, predictors of the MBTI dichotomies and predictors of Enneagram type brings us up to 74 items to include in the quiz. .

Selected?	Item	User rating	Sample size
🔘	107 : ['optimistic', 'pessimistic']	3.6	2136
🔁	72 : ['extrovert', 'introvert']	3.6	1952
🔁	27 : ['impulsive', 'cautious']	3.6	942
✅	495 : ['likes change', 'resists change']	3.6	1122
🔁	82 : ['hesitant', 'decisive']	3.6	980
✅	500 : ['overthinker', 'underthinker']	3.6	1016
✅	298 : ['secretive', 'open-book']	3.6	990
✅	187 : ['idealist', 'realist']	3.6	931
🔁	25 : ['forgiving', 'vengeful']	3.6	973
🔘	35 : ['emotional', 'logical']	3.6	923
🔘	62 : ['confident', 'insecure']	3.6	2071
✅	178 : ['patient', 'impatient']	3.6	2129
🔁	154 : ['judgemental', 'accepting']	3.5	981
🔘	49 : ['scheduled', 'spontaneous']	3.5	988
🔁	10 : ['trusting', 'suspicious']	3.5	930
🔁	127 : ['self-conscious', 'self-assured']	3.5	957
🔘	130 : ['assertive', 'passive']	3.5	986
🔘	470 : ['leader', 'follower']	3.5	1016
✅	171 : ['open-minded', 'close-minded']	3.5	2087
🔁	54 : ['open', 'guarded']	3.5	984
✅	52 : ['focused on the present', 'focused on the future']	3.5	989
🔘	74 : ['calm', 'anxious']	3.5	2039
✅	163 : ['independent', 'codependent']	3.5	2171
🔁	316 : ['realistic', 'fantastical']	3.5	1346
🔘	132 : ['imaginative', 'practical']	3.5	967
⛔	415 : ['thinker', 'feeler']	3.5	1032
✅	349 : ['thinker', 'doer']	3.5	1264
🔘	15 : ['orderly', 'chaotic']	3.5	975
🔘	141 : ['thick-skinned', 'sensitive']	3.5	981
🔘	212 : ['social', 'reclusive']	3.5	932
✅	17 : ['competitive', 'cooperative']	3.5	953
🔁	218 : ['persistent', 'quitter']	3.5	933
🔁	1 : ['playful', 'serious']	3.5	2143
🔘	2 : ['shy', 'bold']	3.5	1994
🔘	60 : ['quiet', 'loud']	3.5	1978
✅	340 : ['ambitious', 'realistic']	3.5	1298
🔁	299 : ['perceptive', 'unobservant']	3.4	981
🔁	47 : ['spontaneous', 'deliberate']	3.4	960
🔘	134 : ['obedient', 'rebellious']	3.4	917
🔁	378 : ['giving', 'receiving']	3.4	1320
✅	147 : ['morning lark', 'night owl']	3.4	956
🔘	136 : ['unambitious', 'driven']	3.4	950
✅	269 : ['self-destructive', 'self-improving']	3.4	1018
🔘	75 : ['disorganized', 'self-disciplined']	3.4	1004
🔘	444 : ['harsh', 'gentle']	3.4	981
🔘	34 : ['chatty', 'reserved']	3.4	2071
🔘	344 : ['lover', 'fighter']	3.4	1270
✅	292 : ['dispassionate', 'romantic']	3.4	972
✅	341 : ['stuck-in-the-past', 'forward-thinking']	3.4	1460
✅	18 : ['tense', 'relaxed']	3.4	2073
✅	357 : ['neutral', 'opinionated']	3.4	1306
🔘	29 : ['creative', 'conventional']	3.4	886
🔘	439 : ['outgoing', 'withdrawn']	3.4	1038
✅	493 : ['energetic', 'mellow']	3.4	1010
✅	164 : ['family-first', 'work-first']	3.4	981
🔘	28 : ['loyal', 'traitorous']	3.4	923
🔁	38 : ['arrogant', 'humble']	3.4	2084
✅	294 : ['intense', 'lighthearted']	3.4	907
🔘	121 : ['sarcastic', 'genuine']	3.4	2077
🔁	42 : ['mischievous', 'well behaved']	3.4	1010
✅	225 : ['alert', 'oblivious']	3.4	916
✅	46 : ['straightforward', 'cryptic']	3.4	946
🔁	360 : ['on-time', 'tardy']	3.4	1358
🔁	352 : ['stubborn', 'accommodating']	3.4	1322
✅	455 : ['outdoorsy', 'indoorsy']	3.4	1069
🔁	345 : ['overachiever', 'underachiever']	3.4	1379
🔘	12 : ['artistic', 'scientific']	3.4	913
🔘	150 : ['neat', 'messy']	3.4	2093
🔁	36 : ['moody', 'stable']	3.4	912
🔁	128 : ['vulnerable', 'armoured']	3.4	898
🔘	5 : ['charming', 'awkward']	3.4	946
🔁	32 : ['diligent', 'lazy']	3.3	939
🔘	199 : ['workaholic', 'slacker']	3.3	987
🔘	186 : ['juvenile', 'mature']	3.3	940
🔁	422 : ['grumpy', 'cheery']	3.3	979
🔁	94 : ['coordinated', 'clumsy']	3.3	2078
🔁	367 : ['unemotional', 'emotional']	3.3	1335
✅	156 : ['gossiping', 'confidential']	3.3	988
✅	490 : ['intuitive', 'analytical']	3.3	1064
🔘	477 : ['focused', 'absentminded']	3.3	1073
🔘	494 : ['hopeful', 'fearful']	3.3	1078
🔁	173 : ['dramatic', 'no-nonsense']	3.3	977
✅	30 : ['curious', 'apathetic']	3.3	904
🔘	397 : ['touchy-feely', 'distant']	3.3	1563

The table below shows the correlation between romantic parners.

Selected?	Item	Couple-Similarity	Sample size
✅	24 : ['dominant', 'submissive']	-0.48	31982
✅	116 : ['pronatalist', 'child free']	0.47	8034
✅	216 : ['liberal', 'conservative']	0.46	31276
🔘	4 : ['masculine', 'feminine']	-0.45	60191
✅	220 : ['patriotic', 'unpatriotic']	0.45	8320
✅	48 : ['libertarian', 'socialist']	0.4	7860
✅	176 : ['androgynous', 'gendered']	0.4	8072
✅	348 : ['freak', 'normie']	0.4	8539
✅	83 : ['devout', 'heathen']	0.39	8049
✅	219 : ['hedonist', 'monastic']	0.39	6640
✅	122 : ['human', 'animalistic']	0.38	8579
✅	16 : ['normal', 'weird']	0.36	28737
✅	364 : ['devoted', 'unfaithful']	0.36	32205
✅	314 : ['tattle-tale', 'f***-the-police']	0.35	31622
✅	203 : ['cool', 'dorky']	0.34	8846
✅	6 : ['lewd', 'tasteful']	0.31	8476
🔘	180 : ['creepy', 'disarming']	0.29	55969
✅	368 : ['rap', 'rock']	0.29	8440
	45 : ['kinky', 'vanilla']	0.28	28340
✅	198 : ['traditional', 'unorthodox']	0.28	34708
✅	370 : ['cat person', 'dog person']	0.28	28410
✅	93 : ['hipster', 'basic']	0.28	8547
	33 : ['lustful', 'chaste']	0.27	27506
✅	157 : ['official', 'backdoor']	0.27	7814
🔘	139 : ['alpha', 'beta']	-0.26	30451
✅	155 : ['average', 'deviant']	0.26	8249
✅	222 : ['wholesome', 'salacious']	0.26	8563
🔘	267 : ['stinky', 'fresh']	0.25	31591
✅	102 : ['mundane', 'extraordinary']	0.25	8641
✅	117 : ['sad', 'happy']	0.25	8550
✅	207 : ['rural', 'urban']	0.25	8669
🔘	156 : ['gossiping', 'confidential']	0.24	63721
🔘	179 : ['poisonous', 'nurturing']	0.24	34085
	224 : ['zany', 'regular']	0.24	30538
	113 : ['cosmopolitan', 'provincial']	0.24	7507