It is possible to understand personality differences between people beyond those assumed by socionics. Here's how.
First, find people who are insistent on resisting the influence of their personal experiences on their person. Of course experience shapes everyone, but it so happens that some people allow traumatic events to mold them more so than others. To deduce the riddle of personality, focus on people who resist being altered by tragic events. This will allow you to see the "core" about which the personality interprets themselves and their being.
Once you have found these people, ask how they make decisions using socionics as a base. Type them first -- the MBTI can help in this regard. Some people have a great deal of anxiety about which category they fall into -- these are bad subjects because they can't give you a confident answer. Instead, find people who are confident in their responses.
Once you are in agreement with the subject about their type, it's time to challenge assumptions. This board is a great place to find such assumptions, but so are the MBTI boards scattered across the web (this boards are also much bigger than this one). If you know the rules, you can do socionics analysis using MBTI data, because the two systems model exactly the same phenomena. On the matter of MBTI, there are two variants, the official variant (which Keirsey also uses) and the Jungian-compliant variant. The Jungian compliant variant produces a 1:1 correlation with socionics; the official variant, in contrast, requires a J/P switch in the case of an introverted type.
One assumption that people tend to make is that type corresponds with prowess. For example, you might think that an ESFJ would be particularly bad at programming. Yet, there are such people out there even though programming is a heavily NiTe skill. Skill does not require originality -- in fact, the only study ever done on correlation between MBTI diagnosis and programming aptitude found that sensors actually more skill at programming than intuitives, and the aptitudes difference between T and F were discovered to be statistically insignificant. Although being T and N may be useful in helping you to conceive of a new program, it is apparently irrelevant. Data such as this guides one in the direction of thinking that there must be a further difference level possible between people of the same type with respect to the aptitude question.
One must use socionics as the base for expanding one's understanding of personality, because socionics is beyond question. Why? Because our experience show it to be true -- socionics is proven through subjective experience. The relations hold and the people do identify with the types, therefore they must be real. Just as they must be real, so must the elementology of Augusta be valid, to the extent that it is understood.
Let us consider the implications of Augusta's elementology. What Augusta asserts is no less than this: there existist a supreme categorical system that man can observe and understand. This system is defined by three dichotomies of being: externals/internals, statics/dynamics, and bodies/fields. All concepts are divisible between these poles.
Augusta gives broad but non-overlapping descriptions of these elements. Indeed, the way she describes them, having strength in a certain element implies a level of skill, the ability to observe immediate things and their immediate relationships. There is a problem in this regard however, which expresses itself in terms of the uneven skill aptitudes between people of the same type. Although a function may indeed be described under Augusta's terms, the assumptions of skill implied in ascription of recognizable concepts to it do not hold universally, or even at all. Experience shows that they do, in fact, contradict each other. This does not mean Augusta's definition is wrong, only that skill using functions in certain ways is not an implication of type diagnosis under the rules Augusta stipulates. (meaning, competence in one measure of intellect at the notable expense of another).
Skills are related to energy, particularly the effective use of energy to change the environment around oneself. On the one hand, we must contend with Augusta's elementology; on the other, we must contend with the self-contradictory nature of its definition. Given that Augusta's elements are ubiquitous, we can safely assert that they are everywhere and define everything -- even energy. At first glance, the definition of Te as "work" encompasses energy; however, we are allowed a certain level of leeway due to the nature of the functions which process the elements -- no one ever said the functions were beholden to the elements, because the elements are by definition what the functions process. We observe the elements only because we have functions to process them; without them, we would be wholly ignorant and blind to them. We can take this thesis a step further and observe that all reality perceived by a man exists as a simulation created by the information processing functions. However we also know that we know of the functions only because of disparities of choice between two alternatives, therefore the observation of disparate choices of energy expenditure, if these are themselves desparing of decision making criteria upheld as the indicators of type, are indicative of a further personality dimension on basis of the rules by which we reckon the functions in the first place. The response to conflict at either level, as it were, is uneven given the disparate preferences. However we can still categorize either of these preferences, and their alternatives, under Augusta's categorization system, and that alone indicates that this second personality dimension can be reckoned with in manner striking similar to that used in the case of the dimension proposed by Augusta/MBTI -- it is assumed, of course, that in any system where the elements are processed, their relationships WILL hold, thus there is a differentiation process and relations set inherent to this new dimension as well. Because we have identified a new processing level, we are thus allowed to observe the existence of reality simulation within the simulator: we end up with Ne that is good at dealing with Ni conditions, Te conditions, Si conditions, etc. all as independent modules of Ne that do, in fact, have their own differentiation process not unlike that faced by Ne itself relative to the other functions.
Yet for the universality of these submodules, can we really say they are subordinate? Observing the persistence of fixation on strong elements of environmental response in spite of the function used at any moment (and only one such function can be used, per Gulenko's ordering rules), it is clear that the strength of the submodules is shared between functions: Ni's Fi submodules is just as strong as Si's Fi submodule compared to all the other submodules in either function. Therefore is hardly incorrect to propose that the differentiation systems of information metabolism and environmental response are independent of each other, under which reasoning the environment response structure is a wholly different typological system which just happens to be employed by the information metabolizing system. But in what way?
Having performed the necessary divestiture of immediate condition observation duties to the environmentally responding structure, what is left for the information metabolizer? The apparent answer is the duty of relating environmental data and observing patterns in it, things the data have in common. We would thus speak of the Ne potential of a certain set of Ni dates, or of the potential of a person to have a certain sensation given surrounding sensory conditions. Far from being displaced, information metabolism is redefined as the function of conceptualization, the relations between its functions being a factor of varying degrees of self-competence in the understanding of these conceptualizations. To understand why conceptualization competence varies, we must look a little deeper into the nature of the functions themselves.
One way to look at the functions is to see functions inside function: see the elements of elements by asking how elements are themselves composed of elements. They are by no means fundamental, being divisible by themselves, and by this further division we can learn more about them. There is reason to believe that a structure may exist for processing these "subelements", because we do not all see the actual realities created by the functions alike.