Monday, June 13, 2011

Statistics what do you think

There are also even more problems involved in using statistics for psychology research. Psychology does not draw enough basic distinctions between separate entities, or consider the nature of different sorts of entities that would delimit the research in such a way to make the outcomes of studies understandable. The point of what I am writing is to make two such distinctions and show how they can improve statistical research.

The first distinction is between dynamics and objects. These are two basic differences that are also important to grammar, in the sense of verbs and nouns. My proposal is that verbs are not effectively described by adjectives in themselves, while nouns can be used as signals when described by adjectives in the context of those verbs to make meaningful discoveries. Here is an example:

We know that the weather in England is warmer than the weather in Labrador, Canada. These two places are on the same latitude of the globe. Here, England and Labrador, Canada are nouns. The temperature is an adjective. The latitude is a system of measurement that roughly corresponds to the construct of temperature (it is colder at the poles than the equator).

The dynamic in this example is the rotating jet streams, which sends warmer weather down to England, and colder weather to Labrador. On the other side of the globe, it sends warmer weather to Alaska, and colder weather to Russia, colder weather to California, and warmer weather to Australia. The problem is that if one wanted to describe this dynamic (verb) with an adjective construct (the weather), it would by itself be ambiguous, confusing and meaningless to us when using abstract constructs whose relationship we are unsure of to begin with (as is the case with psychology).

What we do find, however, is that there are quantitative signals or clues embedded in the study of measurements and nouns, that signal the presence of a dynamic.

In the last example, there are two relevant quantitative features. One is the measurement of latitude, which proposes a system by which an entire object can be measured (the globe), and the measurement of heat, which roughly corresponds with the measurement of latitude.

These features are at first subjective measurements, but there is a measure of objectivity in these measurements when we see the presence of organized paradoxes; in the past example, we see how places on the same latitude (measurement) have different temperatures, and how this theme repeats along the distinction of continents and oceans, with Alaska being at the same latitude but colder than Asia, and England being warmer than Labrador.

Organized paradoxes (results which contradict the understanding of past correlations, such as latitude and temperature) are signals or hints of dynamics, which regulate adjectival constructs (temperature), and are therefore not well described by them. In a sense, dynamics include or contain the range of an adjectival construct within the boundaries imposed by the object (in this example, the cold air from the poles and the warm air from the equator, within the boundary of the earth's temperatures).

Objects may also be measured in terms of ranges. For example, the temperature of England varies in the summer and the winter. Accordingly, England could be confused for a dynamic. However, the relatively limited range of England, Antarctica or Venezuela, as compared to the range of the dynamic of rotational weather patterns, should classify it quantitatively as an object. This is because the range of any certain place does not include all possible temperatures of the whole set (the globe). This shows another shortcoming of modern statistics; extreme responses (very cold or very hot) are blurred by averaging. The extreme measurements of a construct are essential in determining its locale on a measurement scale, and even determining the merit of any certain choice of measurement. The confusion caused by choosing arbitrary constructs, and being unable to distinguish between meaningless results and the possibility of having chosen a meaningless construct, is resolved by crosschecking the extreme results of objects according to a certain measurement with the disparities of averages in similar locations (places on the measurement scale), in the context of a dynamic, and the dynamic's predictable interference of similar objects based on their similar placements on the locale measurement.

To clarify, the measurement of latitude makes England the same as Labrador, and Northeast Asia the same as Alaska. We can tell that our constructs of measurement are meaningful because there are two paradoxes according to the averages (Asia is in the same place and colder, Labrador is in the same place and colder). The dynamic (jet winds) can be differentiated from the locales, because its range of extremes includes the range of extremes of locales, including, importantly, both the terms of hot and cold. Furthermore, there ought to be reflective paradoxes in the southern hemisphere (which can be understood according to a measurement of latitude), by which two locales on the same latitude differ in average temperatures, but in the opposite direction (the west is warmer than the east). This is because the dynamic in question is rotational, and the presence of inverse paradoxes may or may not be an abstract principle of dynamics, or a specific case of a certain dynamic.

The last issue is how one can distinguish the measurement of latitudes from the construct of a dynamic, since the construct of latitude is also inclusive of the range of temperatures. If we begin with the construct of global jet streams, we can tell it is a dynamic because it is inclusive of the range of temperatures. Thus, there are actually two questions; how can one determine that a dynamic is not a measurement of locales, and that a measurement of locales is not a dynamic?

This requires a consideration of the characteristics of a measurement. They should be both qualitative and quantitative. A duck is not a good standard of measurement, because it includes no obvious feature by which we could compare one duck to another duck. A group of ducks could be a measurement, because we could compare how many ducks a group of ducks has compared to another group of ducks.

The utility of this measurement is still questionable, however. Any measurement that is both quantitative or qualitative can still be considered a sterile or arbitrary measurement (insofar as it leads to conclusions that cannot be appreciated).

For a construct of measurement to be useful, it has to be selected in view of its strategic utility. This requires two things: First, that it is an inclusive measurement, such that it does lend potential meaning to a broad context. Second, it has to generate correlations with a second construct that are reflective at the extremes when compared to the original measurement construct (latitude), such as temperature, which shows extreme (cold) weather at the poles.

Thus a useful construct of measurement is a construct combined with another construct that produces a full range of data, with the data having similar measurements at its extremes.

This allows a researcher to be certain that his study will be meaningful in describing a full range of a given contextual construct, and to be able to hypothesize the locales and relative placements of object constructs according to what measures we have for them (that England is not close to the equator, based on its extreme and average temperatures, or based on its latitude). Researchers can also deduce expected responses according to locale and temperature, and inconsistent data can be tested to see whether it is in fact an organized paradox (that the data is inconsistent in an organized way on both the high and low ends of the measurement). This solves the problem of trading reliability for validity and vice versa, since there are now testable features that can be cross checked to determine the utility of a construct.

The issue of mistaking dynamics for the measurements themselves can be solved by inverting this procedure. Latitude correlates with extremes and averages in temperature, and locales have their own identifying averages, and identifying extremes. A dynamic, like a jet stream, should contrast and deviate with local measurements in a more complex and unpredictable way (why it is not a good choice to use dynamics as a starting point for research). However, there should be extreme temperatures in related locales that are shown when studying a dynamic in the correct context of organized locales and temperature measurements. Studying dynamics themselves may not be useful, as they are too complex, but determining what is too complex to be studied, by measuring both averages and extremes, and selecting constructs according to inclusive and reflexive polarity criteria, allows the researcher to make this determination. The researcher can figure out what data is too complex and what data can be organized, and organize it.

To summarize, the distinctions of nouns, adjectives and verbs can be useful to statistics and how statistics can identify each based on: a) variation consistency b) extreme scores c) variation and extreme scores

Nouns should have a variation that is included within an adjective of a defined system. There can be many nouns within an adjective construct. Different nouns should have different variation schemes in terms of a qualitative measurement (location/heat or cold). They should be consistent. The two polar averages (hot or cold) should define the locale possibilities.

Ajdectives should include all variations; it is the defined system for a given qualitative measurement (location or heat/cold). An adjective construct should include all extremes. A pair of adjective constructs should include all measurements and should be able to list the averages in an organized, linear way from either perspective.

Verbs do not have a consistent variation. They usually do not include all extremes.

To summarize how to deduce the presence of a system or of useful constructs:

A system should have paradoxical measurements based on nouns acting inconsistently, or against expectations on the same quality constructs within an adjective construct. However, they should be consistently inconsistent in a symmetrical way across all quadrants of two combined adjective constructs.

A noun can be acknowledged to be acting inconsistently based on odd averages of half a measure (an unusual hot average compared to other locales with similar cold averages). It can also be acknowledged if its score on one of the two combined adjectival constructs is known with certainty. There is one interior and one exterior signal to tell whether a noun is out of place: Internally: half of its internal average score when the qualitative adjective measure is dissected into its polar opposites. Externally: A priori knowledge of the exact location of the noun on one of two adjectival measurements that are defining relations

No comments:

Post a Comment