Enneagram Type and MBTI Type Compared (Statistics)

dollparts said:

It's interesting how many INFP there is!

Click to expand...

Yeah... that is interesting... just look at all of those INTUITIVE INTROVERTS AND INTUITIVE EXTRAVERTS. Sensors here are so difficult to come by >_>

And wow, I had no idea we had so many type 5s and 4s :shocked: They make up at least 1/3 of the entire forum. Interesting that they're both withdrawn types, too... that's just... almost too crazy to believe... maybe because it is -.-'

I could keep going, but I think I've made my point. This "survey" is balls, guise. Totes balls and scrotum and balls. Dick, too.

metaforge said:

INFJs are the rarest type my ass! We're second only to INFP (I can't believe how many INFPs there are OMG!) :shocked:

Interesting statistics! Thanks for sharing.

Click to expand...

Actually, the one reason why they are overrepresented on here is because INFJs gravitate toward personality systems AND the Internet while other types (maybe ESFPs) are out there having fun away from the computer not worrying about typing people.

So, using this data (if this image is from the same study as the OP's), I found that the association between MBTI and Enneagram is as follows:

ISTJ: Type 1
E_FJ: Type 2
E_TJ: Type 3
_NF_: Type 4
I_T_: Type 5
IS_J: Type 6
E__P: Type 7
ENTJ: Type 8
ESFP: Type 9

I apologize if this information is redundant.

Updated data and analysis for 2014

This is collected from your profiles using the "advanced member search" feature. I only collected from those who list both types, a total of 21,105 users. For your reading convenience, especially those who do not care to read a detailed explanation, I place important notes in bold.

Because of the nature of the data available, I will treat a type and it's wings as separate groups. For example, when I mention type 5, I will not include 5w4 or 5w6 unless otherwise noted.

Please be aware of a confounding variable, that a large portion of the population is mistyped, both intentionally or unintentionally. The population also includes many who created an account yet never posted in the forum. For example, within the sample there are five ESFP 1s. Three of them never posted in the forum, one posted once, and the last seems to occassionally visit the forum. The four who posted once or never all went permanently inactive 48 hours after creating their accounts. Because of their lack of activity, it’s impossible to know whether they had truly explored the personality types before labelling themselves. This happens with other type combinations also, but this is one of the easiest to notice. 47% of forum members (including those who were not included in this data) have never posted. The studied population is merely a subset of the entire forum (about 29% of the members).

The first table is the raw data, collected 20 Aug 2014. The columns are color coded (highest number in green, lowest in red) to show which Enneagram is most common to each MBTI. The totals at the bottom (not side) are color coded by that single row, to show which MBTI type was reported the most (INFP>INFJ>INTP>INTJ). The total column (on the right side) is colored to show which Enneagram reported the most (5>4>6>2).



The second table is Enneagram percentages to each MBTI, the sum of the Enneagrams totalling 100% for each MBTI. The average representation of the type is show in the right column. The color coding is identical to that of the raw data, but numbers in the form of percentages may give a quicker understanding of the data. Observing each column and then comparing numbers, we find the strongest correlations: 43.92% of INTPs are 5s, 40% of ISTJs are 1s, and 31.98% of INTJs are 5s.



The third table is a flip of the second, MBTI percentages per Enneagram, totalling 100% for each Enneagram. Color coding is by row. We find that 50.22% of 4w5s are INFP, 49.81% of 4s are INFP, 41.78% of 7w8s are ENTP, 41.38% of 9s are INFP, and 40.57% of 1s are ISTJ. On average, INFPs take up 16.17% of each Enneagram, while ESFPs take only 0.98%. There is a large chunk missing from the ESxx types, but comparing the colors within the columns (although they’re coded per row), we see that, when compared to other MBTI types, ESTJ proportionally takes more 8w7 than 8, despite having more wingless 8s than 8w7s according to the raw data.
An additional example: If you look at ESFP 6w7 & 7w6, you’ll notice that 2.11% is less than 2.89%, yet 2.11% is closer to green. This is because proportionally, more 7w6s are ESFPs than 6w7s are ESFPs.



The fourth table considers the percent that each type combination occupies of Personality Cafe. Totals are colored independently of the data within the normal cells. Without adding those who reported a wing, Type 5 takes the most at 16.57%, while 8s take the least at only 2.21%. Considering wings, 8w9 takes the least at 0.37%. INFPs constitute 22% of the forum, while ESFJs are 0.77% and ESTJs 0.72%.



The fifth table clusters the raw data for the wings into the main Enneagram type.



The sixth table shows the percent of the forum that each clustered type combination takes. On the high end, 9.42% are INTP 5, and 9.22% are INFP 4. ISTJ 7, ISFJ 7, ESFJ 5, ESTP 1, and ESFP 1 are each 0.02% of the forum. ESTJ 4s are 0.01%. The lowest, ESFJ 8, is 0.00% when rounded, because only one person reported as such. However, every combination was represented in the forum, if the Enneagrams are clustered together.



The seventh table shows the predicted numbers that would occur if there was no correlation between MBTI and Enneagram. It multiplies the percentage that each individual type takes, and multiplies it by the type from the other system. For example, ENFPs are 8.202% of the forum, and 9w8s are 0.796%. If we multiply this, we find the predicted number of ENFP 9w8s. The color coding merely shows which combinations are predicted to fill more of the forum. The red outline indicates the top and bottom 2%.



The eighth (which is the last) table shows the difference of the prediction from the actual. INTP 5s are 800 more than predicted, INFP 4s 733 more, and INTJ 4s 280 less. There is a positive skew, so we can say that certain type combinations are more typical, instead of saying that certain combinations are especially odd. However, it is very likely that these are just the most common mistypes. The top and bottom 1% are outlined with red. Since I have personal experience being mistyped as INTP 5 when I'm actually ENTJ 1, I recommend that reconsider your type if it is outside of the predicted 98%, merely for the sake of statistical oddity.

Paradigm said:

It boggles my mind that there's so few self-typed 9s... That number should be so much higher. (Well, so should 1/2/3/..., AKA not-4/5, but wow.)

Click to expand...

Either that, or there are just few 9s who are into self-analysis.

I like this data and presentation a lot, if only it didn't have the catch in that it only reflects people who are self-typed and spend time on this forum.

MBTI Enthusiast said:

Either that, or there are just few 9s who are into self-analysis.

Click to expand...

It's not like the entirety (or even most) of this forum depends on deep self-analysis, nor do I believe in the idea that "X type is drawn here more than others" to explain such a vast disparity. It's far more likely a decent portion of those ~9000 self-typed 4/5s should be under a different label.

TBF the same goes for explaining why there's such a high number of certain MBTI types (INxx, mostly), but this isn't the MBTI part of PerC so may as well gloss over that.

Maybe it would be interesting to have some statistics over how many active PerC members have changed their type, or identifies as previously mistyped. And further what type they mistyped as, how many times they've mistyped, under what theory/theories etc. If one is to go down the statistics road, I mean. (I'm not sure how much more clarity more statistics would add to these statistics though, but I'd at least file it under interesting). :3

The relationship between the complexity of personality and the relative rigidness of typing as such (especially, perhaps, as depicted on a chart) will inevitably make mistyping happen. Especially when you add limited knowledge or just incorrect ideas about what the theories are and aren't supposed to account for, poor research, biases an so on and so forth.

Personally I have been mistyped (by myself and other members) within both theories. Within the enneagram it was indeed the type 5 and 4 I was mostly considering, and it took me a while to figure out I was neither. Now I am very confident I've defined myself "correctly" within the enneagram system. Within MBTI I was totally sure I was an INTP, although I now count myself as most probably being ENTP. My slight discomfort with my MBTI typing shows that I am undereducated and/or do not agree with the entirety of the MBTI system.

Anyway, these statistics are interesting. However, they don't say anything certain outside of the correlation of what these PerC members in question has typed themselves as on their profile at a given time. That in itself is interesting, and can probably spark some fun discussion and discoveries, but it should definitely be looked on with a sceptical eye as well.^^

intrasearching said:

I've been thinking and researching for years and it turns out I am an INFP enneagram 5. I believed myself to be INTP 5 for a year or so but it has since become very clear that I use Fi-Te. Having a significant other who knows me extremely well has helped in finding my true types. It seems I had some misconceptions about myself and with my S/O's help I found I am INFP enneagram 5w6 and tritype 541 sp.

Although I do have strong emotions many times, I am nearly always highly anxious about them, regarding my emotions with apprehension and fear of oblivion, so I withdraw, hide, analyze, and do tons of reading to ground myself in the literature on what and why I feel this way or that way. The most typical behavior for me is to be withdrawn in my bedroom, reading and essentially trying to soothe the fear of my emotions. This often manifests itself in a deep fear that I may lose my mind, and I then read exhaustively about the mechanisms behind various mental illnesses and how they present in the typical patient.

Agh. x P

Click to expand...

Now, I suspect INFP 5w4--2w1--9w1 Sp/So...(most definitely Sp-Dom, anyway). I don't do that. Instead, I go on the internet and vent to strangers.

Corrections to 2014 Statistics

Dalton said:

The eighth (which is the last) table shows the difference of the prediction from the actual. INTP 5s are 800 more than predicted, INFP 4s 733 more, and INTJ 4s 280 less. There is a positive skew, so we can say that certain type combinations are more typical, instead of saying that certain combinations are especially odd. However, it is very likely that these are just the most common mistypes. The top and bottom 1% are outlined with red. Since I have personal experience being mistyped as INTP 5 when I'm actually ENTJ 1, I recommend that reconsider your type if it is outside of the predicted 98%, merely for the sake of statistical oddity.

Click to expand...

I would like to issue a replacement for this table. While the previous table (shown abo) works through raw numbers, this new table expresses that data as the actual number of (insert type combo) divided by the predicted number. The resulting number shows that the actual number of people typing as that combo is [number revealed in table] times the amount of the prediction. Therefore, these new numbers are comparable. The previous table had numbers which were NOT comparable and could even be ignored.



Again, this uses the same data which was available in August 2014. It is merely interpreted. I color-coded it so that <0.125 is red, 1 is yellow, and >8 is green.

This next table is the same thing, but condensed.



I will attempt to translate this into English.

ENTJ 8s are 10.0 times more likely than predicted (based upon the percent that ISTJs and 1s respectively take of the forum). ISTJ 1s are 8.3 times more likely than predicted. ENTP 7s are 5.1 times as likely.

ISTJ 7s are 0.05, in other words 1/20th of the predicted value! (the visible chart rounds down to 0.0). ISTJ 4s are 0.06, INFP 8s are 0.07, ESFJ 5s are 0.09... You can read the rest for yourselves, but it shows that there are more unlikely type combos than there are likely combos.

Again, remember that the main limitation of this data is that it is self-reported. With no standardized understanding of what constitutes the types, there is bound to be a lot of mistyping. Compared to the ENTJ 8s and ISTJ 1s, the INTx 5s really don't set off the balance of the predictions.

Nevertheless, having such a large sample (n = 21,105) is a terrific advantage. The totals along the bottom don't really amount to anything that makes sense, so tell me if you have an idea of what went wrong.

Pangelicus said:

Click to expand...

Interesting. I did the same thing just this morning before accidentally finding your post, which I am now quoting here in the sticky thread so as to not unnecessarily revive your dead thread.

These are my results using the data which I updated last year.

2: J, F, E, S.
3: E, T, J, N.
4: IF, N, Fi, P.


5: I, T, N, P.
6: J, S, T, I.
7: EP, Ti, N.


8: T, E, Te/Se but not Si.
9: F, P, I.
1: J, T, S, I.

Not very different from your conclusions! By the way, iirc, that data comes from PerC, but a outdated by a few years? I guess this proves that the general population of PerC types themselves in the same ways, regardless of whether those ways are "the right way".

I've spent the entire day goofing around with these. I'm bored and I hate it.

The following table is simply a different way to look at the data which was shown in the previous comment that I made.

I did wonder about analysing the functions, but wasn't bored enough

Here is another one I did with slightly different data, which I found somewhere else. The main idea is for clarity of presentation, so that people can understand it and remember it.

(I can send you the Open Office spreadsheet if you want it)

Dalton, your way of presenting the relative frequency is very useful, and - you have the figures for types AND wings! Could you post the data here, as a numerical tab grid, or mail it to me?

It will certainly be useful to rejig your types/wings grid to match my MBTI sequence (starting with ISTJ on the left), as I played around with it for a long time, looking for a sequence that most naturally displays the patterns.

Pangelicus said:

Dalton, your way of presenting the relative frequency is very useful, and - you have the figures for types AND wings! Could you post the data here, as a numerical tab grid, or mail it to me?

It will certainly be useful to rejig your types/wings grid to match my MBTI sequence (starting with ISTJ on the left), as I played around with it for a long time, looking for a sequence that most naturally displays the patterns.

Click to expand...

Dalton said:

This is collected from your profiles using the "advanced member search" feature. I only collected from those who list both types, a total of 21,105 users. For your reading convenience, especially those who do not care to read a detailed explanation, I place important notes in bold.

Click to expand...