Anyone want to help make socionics scientific?

[Socionics Is A Cult]

Overall you can test for whether or not the concrete claims made by Socionics are generally true, which is basically the same thing. Seeing as it’s a philosophical interpretation of reality anyway.

Which concrete claims are those?

[sbbds]

Which concrete claims are those?

Claims made e.g. in type descriptions, in IE descriptions, and any “studies” made by socionists previously where they derived ideas like health groups.

[Socionics Is A Cult]

Claims made e.g. in type descriptions, in IE descriptions, and any “studies” made by socionists previously where they derived ideas like health groups.

ah. I thought it was odd how you distinguished "concrete claims" as though they were distinct from "claims".

There's a danger of not disproving claims because they seem so obviously false as to not merit disproving, and of not proving claims because they seem self-evident.

[sbbds]

ah. I thought it was odd how you distinguished "concrete claims" as though they were distinct from "claims".

There's a danger of not disproving claims because they seem so obviously false as to not merit disproving, and of not proving claims because they seem self-evident.

That’s really how most of the types and facets of Model A are defined though. I mean how else would you test for them.

[Socionics Is A Cult]

That’s really how most of the types and facets of Model A are defined though. I mean how else would you test for them.

In the short to medium-term, there is probably little that can be done. But any evidence from the world of psychology that shows that distinct types exist rather than collections of traits would be a step in the right direction. And any evidence that would indicate that Socionics intertype relations are true would also be be a good sign (for example, you might expect pairs of people with low Conscientiousness to have especially beneficial relationships).

[Lao Tzunami]

@Subteigh, a lot of people bring up the Barnum Effect when I tell them about socionics: the possibility that socionics is like astrology, in that there seems to be different personality types people really believe in, but they are all so vague that they fit anyone, and are meaningless. Because of this, I've tested this out on a few people. If I can decide their type, and I don't mind burning my credibility with them, I will read them their conflictor profile with conviction, so I'm actually biasing the results in the wrong direction. Everyone I've done this to has flat out told me I am wrong and it does not describe them. By contrast, when I read someone a profile of what I actually think they are, about half the people say it fits them really well and the other half say a lot of it fits them but there are details they disagree with.

If socionics is not describing real difference between people, how do you explain that?

[Socionics Is A Cult]

@Subteigh, a lot of people bring up the Barnum Effect when I tell them about socionics: the possibility that socionics is like astrology, in that there seems to be different personality types people really believe in, but they are all so vague that they fit anyone, and are meaningless. Because of this, I've tested this out on a few people. If I can decide their type, and I don't mind burning my credibility with them, I will read them their conflictor profile with conviction, so I'm actually biasing the results in the wrong direction. Everyone I've done this to has flat out told me I am wrong and it does not describe them. By contrast, when I read someone a profile of what I actually think they are, about half the people say it fits them really well and the other half say a lot of it fits them but there are details they disagree with.

If socionics is not describing real difference between people, how do you explain that?

I don't dispute that it is possible to be meaningfully descriptive about an individual's personality.

The issues are, as Singu has covered many times in this thread:
1) is this merely descriptive, rather than explanatory?
2) do distinct types exist, or do people merely fit on a bell curve?
3) do the 16 Socionics types accurately represent the whole of human personality?

[sbbds]

Meh you deduced 2 and 3 yourself @Subteigh , no taking gold stars home on Singu’s behalf.

[Socionics Is A Cult]

Meh you deduced 2 and 3 yourself @Subteigh , no taking gold stars home on Singu’s behalf.

Well, if I didn't steal them from him, I stole them from someone else. Although they may have stolen them from me.

[Lao Tzunami]

I don't dispute that it is possible to be meaningfully descriptive about an individual's personality.

The issues are, as Singu has covered many times in this thread:
1) is this merely descriptive, rather than explanatory?
2) do distinct types exist, or do people merely fit on a bell curve?
3) do the 16 Socionics types accurately represent the whole of human personality?

And as I've said many times, the only way to find out is to do empirical testing within the scientific method.

[Exodus]

Socionics may only show that certain behaviours can be grouped together with other behaviours that are considered similar. That in itself does not mean that the eight Information Elements exist in reality.

What Socionics conjectures about intertype relations is distinct from the types themselves, which could be tested, but first, the types would have to be well-defined and there would have to be a rigorous method for determining an individual's type.

If you had a reasonably good enough questionnaire you could verify (statistically) some correlations that are implied by socionics, without having a 100% way of determining type. This is simply the limitation of the behavioral psychology approach (as opposed to the other neurological approach). Actually, you could apply it using a particular socionist's typings, to show that they are seeing something real. You could then use whatever correlations you find to make a questionnaire. IR is harder because you'd need more people to get the same confidence.

Showing "that the eight Information Elements exist in reality" is a different ball game.

[Lao Tzunami]

1) is [socionics] merely descriptive, rather than explanatory?..

Except this. Obviously socionics is proposing explainations. We need to verify if those explanations are true, but socionics is an explanitory model. Anyone who doesn't understand this has not done their basic research and should stop advocating their ignorant opinions. The reason this is a valid criticism of the big 5 is because it is a brute force factor analysis model. SOCIONICS IS NOT A FACTOR ANALYSIS MODEL. It has other issues, but not lack of explanation.

[sbbds]

There is an explanation, but it’s philosophical mainly and the parts that aren’t philosophical need to be verified and related to mainstream science.

[sbbds]

If you had a reasonably good enough questionnaire you could verify (statistically) some correlations that are implied by socionics, without having a 100% way of determining type. This is simply the limitation of the behavioral psychology approach (as opposed to the other neurological approach). Actually, you could apply it using a particular socionist's typings, to show that they are seeing something real. You could then use whatever correlations you find to make a questionnaire. IR is harder because you'd need more people to get the same confidence.

Showing "that the eight Information Elements exist in reality" is a different ball game.

Sorry for the Se here; this isn’t directed at you but generally to the thread:

I don’t really want to think about this unless we’re actually planning on doing it. I think we should find out some numbers like on what’s an acceptable sample size and success rate for this kind of study in behavioural psychology, and other limitations. I want to know what the scale of the project would need to be like.

[sbbds]

Unless you guys don’t want to actually do this or at least set up a thorough coherent plan. In which case I’m not helping anymore.

[sbbds]

No need to sorry. You are in a right direction.

I don't know how to plan a good DoE for this forum, because when I am trying to wonder about it, i 'see' many possibilities about what we can do for it. But when you asked about success rate, that's something that I personally should limit all that possibilities and try to make a small-sure step rather than big-unsure step.

The only things that I am sure it can reduce the risk of failure when doing this project is, we should do vote in this forum (like what we do in superlatives this year) for pick 16 members (one for each type) of this forum, whom willing to join into this project to do typing for any materials that this project might offered to them. So each type should have a representative that being agreed by most members (no need to more than 50% or any number. Just majority). The next things like what material and experiments should be do later, is still unsure for me, especially because you need a number to scale the success rate. But I will keep my mind open to make DoE voluntarily here. Hope I found something later. But, yeah, this is what I am sure as a first step.

...

First of all, I think people who don’t care about this are just not participating, which is most of the forum. It’s obviously been open to everybody this entire time already.

Second of all, it’s wasting people’s time to try to involve them in something if you’re not sure if it will succeed nor even have anything more than an overly vague idea of how to make it happen so far.

[sbbds]

Well, that's my problem too. To avoiding bias, it really needs all 16 types to 'sit' together when in a project. No matter whether they will care or not.

out of topic ...

Like in soccer, there should be 11 in-field players, 5 bench players, and one manager before the kick off, no matter whether they will win a match or lose. Even all 11 in-field players, not all of them are have attacking mentality. They must have one goalkeeper that can touch ball with hands when any players can't, some defenders that stay in back when everyone is in front to anticipating counter attack, some midfielders that distributing ball from back to front, and few attackers that focused on goal. Bench players, maybe they seems not care with success, but when there are tired or injured in-field player, they are being substitutes to keep their team on a right track. And a manager, that never kick or touch any ball in game, but overseeing all aspect when they are playing soccer, and be responsible for a success of a team.

So do Socionics. The founders made this to explain how each types are important for this society. So, if now members whom interesting to join this project or this thread is still some types and not all, it doesn't mean it's enough. No. It isn't. 16 members that represents one of each types, are more important than 50 members that represent only some types. Should I need to explain why? I don't think so. You might know why, right?


For people whom valuing time and effort, yes, it's wasting. But not all people are feeling wasted when facing failure.

out of topic...

for example, the famous youtube channel, failarmy. When capturing a fail is considering as a success.
Even Space X are planning a fail maneuver before landing its Falcon 9 to understanding the possibilities of future water landing.

I can work for success or for fail, depends on what purpose Socionics will go. So if the objective for this thread is for a success, then please follow the restriction to reduce risk of "unsure for will succeed". If this thread are being headed to test a failure, then go with a big step without caring other types, and I don't care if later in future this project might be successful but useless.

I apologize for what I said.

How will you know if each of the 16 people we’ve chosen to represent each of the types has been typed with scientific accuracy?

[sbbds]

I don't say that all the 16 types that we've chosen has been typed with scientific accuracy. Because actually none of us never seeing any scientific approach to Socionics. It just to reduce the risk of failure because of bias.

For example, if I am EII and I typed Mr. X, as 100% LSE, then all EII might have tendencies towards my own typing too because that's what our way of thinking, feel, sense, or instinctively respond to how Mr. X inform us when in typing. So there might be another EII that can say, "He might not LSE 100%, but just 70%, because there is 30% some of their face looks more like LIE rather than LSE." or another EII can say, "Okay, LSE, but just 20%. I can see his introvert side more than extrovert, so he might SLI 80% too." and many various possibilities in EII.

But if we bring SLE to see that Mr. X, I believe that SLE have much different framework of information processing that might reveal something that EII can't understand from Mr. X, so it will be possible to said that "No, I believe that Mr. X is 0% LSE, he is not LSE. He is lying." and type another instead with another possibilities of types rather than LSE. Maybe LIE, SLI, SEE, anything that EII never can see or understand before.

That's why doing vote is a small first step to make sure that this project will being run with all possible type frameworks (16 types, 32 if subtype matters too), before next we go to bring all this 16 volunteers into the next step, about scientifically typing. So, what I said was is still not scientifically accurate, but it a small step to make sure this project will be held with a low risk of bias.

Right, we want to make sure a variety of people are collecting the information from people if possible. That’s a fair point Andreas.

[sbbds]

I’m feeling anxious and am starting to lose interest about this already @ajsindri . Anyway let me know by PM or something because I don’t always get pinged. Oh and if we’re doing this, you’d need to recruit people to do it with you in the Americas or English-speaking Europe first and foremost. I can’t effectively help you from where I am now in Asia.

- Find out what usage rights are involved with this and how to navigate any red tape. Consult Gulenko or whoever owns “schools” for it.
- Find out the real scale and what you’d need to carry out the experiments and support claims, what you’d need to make it valid enough for publishing in an actual psych or scientific journal ... ideally speaking. Consult people in the actual real field. See what they have to say.

Then if it still seems within your means:
- Maybe get like 4 people per Socionics type (2 per gender) at least to start to do mini tests on, and develop your criteria based on those. Do your math magic and see what you find. Also get different people to help you administer the typing criteria or testing.
- Consult the pros in Eastern Europe and show them what you’ve got, if they can help or whatever, update them
- Recalibrate your methods, test again, analysis

Then you scope out labs / the means to actually carry out a full scale study with hundreds of people etc., including whatever else that entails.

That’s the only way you with your means now can probably hope to make the whole theory of Socionics scientific all at once.

Another way you could do it would be to just choose one or two unique easily falsifiable claims made by Socionics and go through the same process above to get your foot in the door faster, but the disadvantage of this is that we wouldn’t know if the entirety of Socionics is valid or not then, just a small part of it.

So anyway you’re welcome, Merry Christmas.

@squark @Adam Strange and other STEM oldfags please review this and I welcome any feedback ty

[sbbds]

Actually as I have mentioned I work in science/tech too but mostly just the dealing with people side or light research, plus I’m completely a noob. My bosses have no reason to trust me.

[Socionics Is A Cult]

If you had a reasonably good enough questionnaire you could verify (statistically) some correlations that are implied by socionics, without having a 100% way of determining type. This is simply the limitation of the behavioral psychology approach (as opposed to the other neurological approach). Actually, you could apply it using a particular socionist's typings, to show that they are seeing something real. You could then use whatever correlations you find to make a questionnaire. IR is harder because you'd need more people to get the same confidence.

Showing "that the eight Information Elements exist in reality" is a different ball game.

Sure. I think if Socionics could be shown to have practical utility (e.g. by showing that its predictions about intertype relations could be shown to have some merit), that would at least show it is somewhat meaningful and perhaps superior to other typologies. But perhaps it doesn't need that.

[Socionics Is A Cult]

Except this. Obviously socionics is proposing explainations. We need to verify if those explanations are true, but socionics is an explanitory model. Anyone who doesn't understand this has not done their basic research and should stop advocating their ignorant opinions. The reason this is a valid criticism of the big 5 is because it is a brute force factor analysis model. SOCIONICS IS NOT A FACTOR ANALYSIS MODEL. It has other issues, but not lack of explanation.

I don't understand how Socionics in its current form is explanatory, rather than people saying they are more "this" than "that".

[Lao Tzunami]

I don't understand how Socionics in its current form is explanatory, rather than people saying they are more "this" than "that".

Have you read Psychological Types by Jung and The World as Will and Representation by Schopenhauer? If you haven't, please go read those first, and then we can talk more about this.

The World as Will and Representation sets up the Kantian basis for socionics, but the most pertinent section is Book 1 Section 12:
Audio: https://www.youtube.com/watch?v=Vm0Iwxlkrzo
Text: https://en.wikisource.org/wiki/The_W...Book#.C2.A7_12

[Lao Tzunami]

@sbbds In the US, only a specific inventory can be copyrighted, so any Russian resource can be used as long as it is paraphrased. We need to finish the math work before the data can be analyzed. If you want to help with that, do you know linear algebra, group theory, statistics or any kind of programming?

[sbbds]

@sbbds In the US, only a specific inventory can be copyrighted, so any Russian resource can be used as long as it is paraphrased. We need to finish the math work before the data can be analyzed. If you want to help with that, do you know linear algebra, group theory, statistics or any kind of programming?

Ok so you are going to do that in the US. I am not good at figuring that kind of thing out so I’ll pass unless you really need it. Stats is most familiar. Are you saying you’ve already looked into the first points I listed up there?

[Lao Tzunami]

- Find out the real scale and what you’d need to carry out the experiments and support claims, what you’d need to make it valid enough for publishing in an actual psych or scientific journal ... ideally speaking. Consult people in the actual real field. See what they have to say.

I am actively working on this step.

[sbbds]

I am actively working on this step.

How?

If you want help on this PM me or whatever. I wouldn’t pitch in with math work unless that’s been concluded first though.

[Socionics Is A Cult]

Have you read Psychological Types by Jung and The World as Will and Representation by Schopenhauer? If you haven't, please go read those first, and then we can talk more about this.

The World as Will and Representation sets up the Kantian basis for socionics, but the most pertinent section is Book 1 Section 12:
Audio: https://www.youtube.com/watch?v=Vm0Iwxlkrzo
Text: https://en.wikisource.org/wiki/The_W...Book#.C2.A7_12

I don't see how that is relevant.

[Lao Tzunami]

How?.. I wouldn’t pitch in with math work unless that’s been concluded first though.

I'm working with @thehotelambush and a few other friends on the group structure of socionics. I have a working model of the reinin dichotomy analysis. I just finished a psychometric statistics class and have already talked with the psychology faculty at my university. I have plans to make a beginner socionics YouTube tutorial with another friend, which will be a good time to consolidate different definitions about the same socionic objects, which needs to be done before testing them.

/ Thanks for being interested in helping. We probably won't be done with the math phase for at least another year. When we finish, I'll let you know.

[Socionics Is A Cult]

I have the feeling in this thread that we each know what others mean when they refer to a Type or an Information Element, or at least, for all practical purposes (at least in the short term, not so much the long term where there is an ultimate desire to test the claims of Socionics), we have visualizations that are synonymous, and yet these visualizations are rather nebulous and prone to shifting in definition depending on the context, and having a tendency to being circular in terms of proof.

[Lao Tzunami]

Have you read Psychological Types by Jung and The World as Will and Representation by Schopenhauer? If you haven't, please go read those first, and then we can talk more about this.

I don't see how that is relevant.

I have the feeling in this thread that we each know what others mean when they refer to a Type or an Information Element, or at least, for all practical purposes (at least in the short term, not so much the long term where there is an ultimate desire to test the claims of Socionics), we have visualizations that are synonymous, and yet these visualizations are rather nebulous and prone to shifting in definition depending on the context, and having a tendency to being circular in terms of proof.

Reading the seminal texts is vital because they are the basis of socionics. Augusta's Model accepts Jung's 8 types as correct, and then abstracts them into information, which is projected onto the 8 base types, creating 16 types of information metabolism and 16 intertype relations. I think socionics has done a great job explaining the rationale behind its unique advancement to Jungian typology. But if you have issue with the 8 information elements, you have to show that either the information abstractions, or Jung's original 8 base functions are incorrect. To do that, you have to read and understand Psychological Types, and that may also mean reading Jung's inspirations, especially Schopenhauer's World as Will and Representation, depending on the depth you want to understand these concepts at.


Especially if you haven't read Psychological Types, of course this doesn't make sense to you, because you're not educated. I have done the research, but I don't have the time to teach you specifically, nor is it my responsibility. Reading these books is a lot of work, and I'm sorry there are not better educational resources available, but having to do your own independent research comes with the territory of pioneering theories.


As far as I can tell, the reasoning is solid, and besides, it still needs to be demonstrated empirically. So in either case, the solution is fair scientific tests. Any social construction or archetypal forms should be observable in the data. We can't argue it out, the only way to settle this is to do the tests.

[sbbds]

@ajsindri What did your school’s psych faculty say? Did they say they’d let you use their lab?

I meant concluding that side of things, like figuring out everything you’d need to carry the experiments out.

I’d be willing to help towards that, or anything after you guys have secured all that.

[Lao Tzunami]

@ajsindri What did your school’s psych faculty say? Did they say they’d let you use their lab?

I meant concluding that side of things, like figuring out everything you’d need to carry the experiments out.

I’d be willing to help towards that, or anything after you guys have secured all that.

They are only interested in helping me analyze data, not the theory. They said most graduate PhDs focus on only one or two factors. If you did a proper test of Model A, not with dichotomy abstractions, you would have 8 functions x 8 information element = 64 combinations, which is way too much for any person to do on their own. That's one reason it would be so good to have a way to synthesize many people's individual projects into the most compelling application of socionics.

[sbbds]

They are only interested in helping me analyze data, not the theory. They said most graduate PhDs focus on only one or two factors. If you did a proper test of Model A, not with dichotomy abstractions, you would have 8 functions x 8 information element = 64 combinations, which is way too much for any person to do on their own. That's one reason it would be so good to have a way to synthesize many people's individual projects into the most compelling application of socionics.

.......

Ok so they just told you what I’ve been telling you this entire time from my first post, #6 in this thread:

If you want to give Socionics a start in making it scientific, you need to take one of its small, specific concrete claims, such as “people who act in x way (implying they’re of a certain type(s)) will also later act in y way more than the control”, and find a way to reliably measure it. That’s how Big 5 / OCEAN built a name for itself. It took many years though.

@ajsindri

”That's one reason it would be so good to have a way to synthesize many people's individual projects into the most compelling application of socionics.”

^^^^ What the hell do you mean by this?

[sbbds]

@ajsindri So assuming they said “yes” to analyzing data, they’re willing to offer some resources in the form of slapping their faculty insignia on it and a chance for it to be published and enter peer review. That’s something at least. Do give as many details on this as possible because it will help.

[Singu]

All of this stuff has been done before already.

If you actually study and research psychology, then you'll realize that psychologists have already gone through and tested most of these kinds of theories, including psychoanalysis and typologies, and found that they had little to no predictive power, let alone explanatory power.

For example in the book Social Foundations of Thought & Action: A Social Cognitive Theory by Albert Bandura (the most cited psychologist alive today), it goes through all of them comprehensively:


Ch 1. MODELS OF HUMAN NATURE AND CAUSALITY
Social Foundations of Thought & Action, Albert Bandura

Many theories have been proposed over the years to explain human behavior. The basic conceptions of human nature that they embrace and the causal processes that they posit require careful examination for several reasons.

PSYCHODYNAMIC THEORY

Human behavior is commonly viewed as motivated from within by various needs, drives, impulses, and instincts. In psychodynamic theory, for example, human behavior is the manifestation of the dynamic interplay of inner forces, most of which operate below the level of consciousness (Freud 1917, 1933). Since the proponents of this school of thought consider the principal causes of behavior to be drives within an individual, that is where they look for the explanations of why people behave as they do. Although this theory has gained widespread acceptance and is deeply entrenched in the public view of human behavior, it has not gone unchallenged.

Theories of this sort are criticized on both conceptual and empirical grounds. The inner determinants are often inferred from the very behavior they supposedly caused, creating interpretive circularities in which the description becomes the causal explanation. A hostile impulse, for example is deduced from a person's irascible behavior, which is then attributed to the action of an underlying hostile impulse.

Similarly, the existence of achievement motives is deduced from achievement behavior; dependency motives from dependent behavior; curiosity motives from inquisitive behavior; power motives from domineering behavior, and so on. There is no limit to the number of drives one can find by inferring them from behavior. Indeed, different theories propose diverse lists of motivators, some containing a few all-purpose drives, others encompassing an assortment of specific drives. If causal propositions concerning drives are to be empirically testable, then drives must be specified by the antecedent conditions that activate them and govern their strength, rather than being inferred from the behavior they supposedly produce.

The conceptual structure of theories that invoke drives or impulses as the principal motivators of behavior has been further criticized for disregarding the complex and changeable patterning of human action. An internal motivator cannot adequately account for marked shifts in a given behavior under differing situational circumstances. When varying social conditions produce predictable changes in behavior, the postulated cause cannot reside mainly in a drive in the organism, nor can the cause be less complex than its diverse effects.

Psychodynamic theory assumes a thorough psychic determinism, but it does not as a rule, postulate definite relationships between the unconscious inner life and human thought and action. In fact, the inner dynamics are said to produce any variety of effects, even opposite forms of behavior. Such formulations are, therefore, not easily testable nor refutable by empirical evidence. While the conceptual adequacy of psychodynamic drive theories could be debated at length, their empirical limitations cannot be ignored indefinitely. They provide ready interpretations of behavior that has already happened, but, as we shall see shortly, they are deficient in predicting future behavior. Almost any theory can explain things after the fact. Findings from research conducted is from other perspectives have underscored the need to shift the focus of causal analysis from internal dynamics to reciprocal causation between personal and environmental factors. Behavior patterns commonly attributed to unconscious inner causes can be in stated, eliminated, and reinstated by varying appropriate social influences and by altering people's ways of thinking. Such findings indicate that the major determinants of behavior arise from transactional dynamics, rather than flow unidirectionally from inner dynamics of unconscious mental functions.

The explanatory power of a psychological theory is gauged in several ways. First, theories must demonstrate predictive theories power. Second, the methods the theories yield must be capable of effecting significant changes in human affect, thought, and action. Weaknesses in theories become readily apparent when they are put to work and can be judged by the results they produce. One can predict and change events without knowing the basis for the successes. So third, theories must identify the determinants of human behavior and the intervening mechanisms by which they produce their effects. But explanations that have no predictive value will be pseudo-explanations. The adequacy of explanation is, therefore, judged largely in terms of predictive accuracy. Psychodynamic formulations have been found wanting on all these counts.

CLASSIFYING PEOPLE AND INDIVIDUALIZING TRAIT DIMENSIONS

Bem and Allen (1974) have advanced the view that some people are highly consistent in some behaviors, but evidence of cross-situational generality is obscured when data from consistent and variable responders are combined and researchers, rather than the respondents, select which traits are relevant and which types of behavior represent them. From this perspective, a psychological theory which seeks to predict actions from traits must settle for the modest goal of predicting only some of the actions of only consistent people, provided one can identify beforehand who is likely to be consistent in what realm of behavior.

Bem and Allen use people's judgments of themselves as either consistent or changeable for the traits in question as the identifier. To demonstrate that self-reported consistency foretells uniformity, students were measured for their friendly and conscientious behavior and rated by their parents, by a peer, and by themselves for friendliness and conscientiousness on a questionnaire describing many different situations. The ratings for each of the two traits were summed for each judge across the situations described in a global score, and then the degree of agreement between judges was computed. Students who viewed themselves as consistent were rated by others with higher agreement than those who judged themselves highly variable in behavior.

Correlating verbal reports of behavior averaged across specific situations does little to illuminate the central issue of whether trait measures predict how people will actually behave under different conditions. Scores pooled across situations may embody high, moderate or low behavioral variability. In testing for behavioral generality, one must measure directly how individuals vary in their behavior under different circumstances, rather than how, on the average, they stand in relation to others, or how well judges agree among themselves in their over-all impressions of the individuals selected to study.

In the few behaviors that Bem and Allen actually measured, the self-described unchangeables were found to be more consistent for talkativeness but not for conscientious actions, thus revealing an inconsistency in the predictor of behavioral consistency.

In a more comprehensive study examining different measures of self-reported consistency and many personality dimensions, Chaplin and Goldberg (1983) found that self-reported consistency is uniformly unpredictive. Not only do different indexes of consistency disagree but, however it is measured, the self-reported consistent types are no more uniform in their behavior than the changeable types on any personality dimension.

That sorting people into consistency types gains little predictive power, as far as behavior is concerned, has been further confirmed by Peake and Lutsky (1981). Others agree more closely in their over-all impression of persons who see themselves as changeable than for those who characterize themselves as variable, but both groups show little uniformity in their actual trait behavior in different settings. The different social impressions probably arise because others often cannot observe how those they are rating act in various milieus and must either guess how they are likely to behave or rely on what they tell them. Presenting one-self as a highly consistent person may thus foster social impressions of consistency, but it does not improve the predictability of trait measures.

The source of this erroneous impression about behavioral generality has itself become the subject of study. Mischel and Peake (1982) have found that people's perceptions of others' self-consistency is related to how uniform the latters' behavior is in key features of the trait over time in similar situations but is unrelated to how they actually behave in different situations. People thus misread cross-situational generality in behavior from temporal stability in similar situations. Observers, whose information is limited about how others conduct themselves in diverse situations, would be especially prone to mistake behaving similarly in the same setting over time as indicative of behaving similarly in different settings. If the eyes do not behold a wide range of transactional situations, then behavior will appear consistent in the eyes of those beholders.

Studies of situational generality of behavior devote much attention to what trait behaviors should be assessed but give little consideration to the kinds of environments that should be sampled. It is perhaps not entirely surprising that approaches attributing behavior to traits would neglect the properties of the social environment. The most informative methodology for studying cross-situational generality would be to record how much people vary their behavior across situations which differ measurably in the functional value of the behavior being examined in those settings. Situations chosen for study should be scaled and selected in terms of the incentives and sanctions they customarily provide for the particular behavior, rather than chosen arbitrarily. Such studies would undoubtedly reveal that all people behave discriminatively most of the time, being more prone to express a given form of behavior when it is advantageous to do so than when it serves no useful purpose or brings detrimental results. It is only by including a range of environmental dispositions that the transsituational fixedness, or nonfixedness, of behavior can be adequately evaluated.

Psychological knowledge is better advanced by exploring the sources of variability of behavior than by searching for subtypes of people who behave invariantly, regardless of circumstances. It would be a misleading truncated theory that called on persons to choose which of their actions are predictable but viewed persons as unpredictable in areas of functioning in which they very sensibly vary their actions to suit the changing circumstances.

Progress in gaining predictive knowledge requires research that systematically varies factor that contribute to behavioral variability as well as examines correlations among behaviors in naturally occurring settings. The number of persons who act invariantly would fluctuate depending upon the behavior selected for study, the extent to which the situations sampled differ in their likely consequences for the given conduct, how much variability is tolerated in the criterion of consistency, and whether one measures verbal reports of behavior or the behavior itself. Behaviors that are highly functional in diverse settings, as, for example, acting intelligently, would be more consistent than behaviors that have different effects under dissimilar circumstances. It would be difficult to find adolescents who are consistently aggressive toward parents, teaches, peers, and police officers, because the consequences for the same conduct vary markedly (Bandura & Walters, 1959). Even in the case of a widely acceptable behavior such as friendliness, the ranks of the consistent responders would shrink simply by including some situations in which friendliness is an unlikely response, as, for example, when individuals are being exploited or discriminated against. Only those who are grossly undiscerning or who have a poor sense of reality would remain steadfastly amiable.

[Singu]

tl;dr, you can't find something new by just reading some ancient texts of Jung. Nor is it anything revolutionary to try to "test" them. It has all been done before already.

[sbbds]

@Singu

”Psychological knowledge is better advanced by exploring the sources of variability of behavior than by searching for subtypes of people who behave invariantly, regardless of circumstances.”

That’s not what Socionics is. It doesn’t ever posit that people have invariant behaviours determined by their typing at all times.

“The most informative methodology for studying cross-situational generality would be to record how much people vary their behavior across situations which differ measurably in the functional value of the behavior being examined in those settings.”

That’s exactly what sindri is trying to do here.

[Socionics Is A Cult]

Reading the seminal texts is vital because they are the basis of socionics. Augusta's Model accepts Jung's 8 types as correct, and then abstracts them into information, which is projected onto the 8 base types, creating 16 types of information metabolism and 16 intertype relations. I think socionics has done a great job explaining the rationale behind its unique advancement to Jungian typology. But if you have issue with the 8 information elements, you have to show that either the information abstractions, or Jung's original 8 base functions are incorrect. To do that, you have to read and understand Psychological Types, and that may also mean reading Jung's inspirations, especially Schopenhauer's World as Will and Representation, depending on the depth you want to understand these concepts at.


Especially if you haven't read Psychological Types, of course this doesn't make sense to you, because you're not educated. I have done the research, but I don't have the time to teach you specifically, nor is it my responsibility. Reading these books is a lot of work, and I'm sorry there are not better educational resources available, but having to do your own independent research comes with the territory of pioneering theories.


As far as I can tell, the reasoning is solid, and besides, it still needs to be demonstrated empirically. So in either case, the solution is fair scientific tests. Any social construction or archetypal forms should be observable in the data. We can't argue it out, the only way to settle this is to do the tests.

I have read both those works, and re-read the bit you referenced of Schopenhauer. Jung and Schopenhauer are not known for their emphasis on making falsifiable predictions.

[Socionics Is A Cult]

They are only interested in helping me analyze data, not the theory. They said most graduate PhDs focus on only one or two factors. If you did a proper test of Model A, not with dichotomy abstractions, you would have 8 functions x 8 information element = 64 combinations, which is way too much for any person to do on their own. That's one reason it would be so good to have a way to synthesize many people's individual projects into the most compelling application of socionics.

Aren't these combinations arbitrary? It is not seriously disputed that personality can be defined in dichotomies.