Saturday, March 26, 2011

Phenomenology in Science

Phenomenology of Information


The term phenomenology in science is used to describe a body of knowledge which relates empirical observations of phenomena to each other, in a way which is consistent with fundamental theory, but is not directly derived from theory. For example, we find the following definition in the Concise Dictionary of Physics:
Phenomenological Theory. A theory which expresses mathematically the results of observed phenomena without paying detailed attention to their fundamental significance.[1]
The name is derived from phenomenon (from Greek φαινόμενoν, pl. φαινόμενα - phenomena and -λογία - -logia, translated as "study of" or "research") which is any occurrence that is observable.

Contents

Phenomenology in physical sciences

There are cases in physics when it is not possible to derive a theory for describing observed results from the known first principles (such as Newton's laws of motion or Maxwell's equations of electromagnetism). There may be several reasons for this. For example, the underlying theory is not yet discovered, or the mathematics to describe the observations is too complex. In these cases sometimes simple algebraic expressions may be used to model the observations or experimental results. The algebraic model is then used to make predictions about the results of other observations or experiments. If the predictions made by the algebraic model are sufficiently accurate, they are often adopted by the scientific community despite the fact that the algebraic expressions themselves cannot be (or have not yet been) derived from the fundamental theory of that domain of knowledge.

The boundaries between theory and phenomenology, and between phenomenology and experiment, are fuzzy. Some philosophers of science, and in particular Nancy Cartwright argue that any fundamental laws of Nature are merely phenomenological generalizations.[2]

Examples in physics

The examples below are in chronological order.
  • Second law of thermodynamics: Prior to the development of statistical mechanics by Ludwig Boltzmann (1896), this law was phenomenological. For instance, spontaneous net flow of heat from a lower temperature to a higher temperature had never been observed and this fact served as the basis of the second law.
  •  
  • Rutherford model also known as planetary model (1911) describes the structure of an atom based on the experimental results. It has a number of essential modern features, including a relatively high central charge concentrated into a very small volume in comparison to the rest of the atom. It resembles the planetary system, a known physical object larger by several orders of magnitude. It was superseded in 1913 by the Bohr model, which used some of the early quantum mechanical results to give locational structure to the behavior of the orbiting electrons, confining them to certain circular (and later elliptical) orbits.
  •  
  • Landau theory of second order phase transitions (1936).
  •  
  • Bloch equations (1946).
  •  
  • Ginzburg-Landau theory of superconductivity (1950).
  •  
  • Modified Newtonian dynamics (1983)

Phenomenology in social statistics

In the science of statistics, the collection of quantifiable data from people involves a phenomenological step.

Namely, in order to obtain that data, survey questions must be designed to collect measurable responses which are categorized in a logically sound and practical way, such that the form in which the questions are asked does not bias the results. If this is not done, data distortions due to question-wording effects (response error) occur, and the data obtained may have no validity at all, because observations are counted up which do not have the same meaning (it would be like "adding up apples and pears") [3] A prerequisite of a good survey is that all respondents are really able to give a definite and unambiguous answer to the questions, and that they understand what is asked of them in the same way. One could, for example, ask farmers, "How much risk do you run on your farm?" with a scale of response options ranging, for example, from "a lot of risk" to "no risk".

But this yields quantitatively meaningless data which is not objective, since the interpretations of "how much risk" by farmers could focus, for example, on the number, size, frequency, severity or consequence of risks, and each farmer will have his own idiosyncratic idea about that. All farmers may suffer, for example, from a lack of rainfall, but some will personally consider it a large risk, others a low risk and some not a risk at all. Furthermore, in actually asking the questions of respondents and subsequently coding the responses to numerical values, a technique must be found to ensure that no misinterpretation occurs of a type that would lead to errors. In other words, in designing the survey instrument, the researcher must somehow find a satisfactory "bridge" of meaning between the logical and practical requirements of the survey statistician, a statistical classification scheme, the awareness of respondents and the processors of the raw data. Finding this "bridge" involves an abstraction process which necessarily goes beyond logical inference, theory and experiment and involves an element of "art", because it must establish an appropriate connection between the language used, the intersubjective interactions between the surveyor and the respondent, and how respondents and those who process the data construct the meaning of what is being asked of them. For this cognitive process, it is impossible to provide a standard procedure which will always work, only "rules of thumb"; it requires a "practical" human insight [4].

See also

References

  1. ^ Thewlis, J. (Ed.) (1973). Concise Dictionary of Physics. Oxford: Pergamon Press, p. 248.
  2. ^ Cartwright, Nancy, intro., How the Laws of Physics Lie, 1984, Oxford U.
  3. ^ see, e.g., Nicolaas J. Molenaar, "Non-Experimental Research on the Effects on the Wording of Questions in Survey Interviews". Quality & Quantity, 16, no 2 (1982) 69-90 and Norman M. Bradburn and Seymour Sudman, Response Effects in Surveys : A Review and Synthesis. Chicago, Aldene Pub. Co., 1974.
  4. ^ See Stanley Payne, The Art of Asking Questions. Princeton: Pinceton University Press, 1980

1 comment: