cores/objects/models-hypotheses-notes.html
Hilborn, Ray and Marc Mangel. 1997. The ecological detective: confronting models with data. Princeton University Press. Princeton, NJ.
pg 24.
theory - "implies considerable evidence in support of a formulated general principle, explaining the operation of certain phenomena"
hypothesis - "implies an inadequacy of support of an explanation that is tentatively inferred, often as a basis for further experimentation"
model - " is an archetype, "a stylized representation or a generalized description used in analyzing or explaining something." Thus, the models are tools for the evaluation of hypotheses (our best understanding of how the world works), but they are not hypotheses."
pg 26
"We use models to combine what we know with our best guesses about what we do not know"
pg 27
"since a model is not a hypothesis we must admit from the outset that there is no "fully correct" model."
pg 30
"Simpler models often provide insight that is more valuable and influential in guiding thought than accurate numerical fits."
pg 11
confront multiple models with data and assess the likelihood that that model describes the system
One model might have a 50% goodness of fit and another might have 75% goodness of fit.
Neither is correct.
challenge (which takes skill) is to design sets of models that can be tested with experiments that will be able to distinguish between the descriptions
pg 12
referring to Platt (1964) "strong inference" consists of the following steps
quoting list from Hilborn and Mangel 1997
- devising alternative hypotheses
- devising a crucial experiment (or several of them) with alternative possible outcomes, each of which will, as nearly as possible, exclude one or more of the hypotheses
- carrying out the experiment so as to get a clean result
- recycling the procedure, making sub hypotheses or sequential hypotheses to refine the possibilities that remain, and so on
Platt, John R. 1964. Strong Inference: Certain systematic methods of scientific thinking may produce much more rapid progress than others. Science. (16 Oct 1964) 146 (3642):347-353.
same list as above, but calls "4" "1'"
"Strong inference, and the logical tree it generates, are to inductive reasoning what the syllogism is to deductive reasoning"
"How many of us write down our alternatives and crucial experiments every day, focusing on the exclusion of a hypothesis?"
"the use of the second great intellectual invention, the "method of multiple hypotheses""
"when there are multiple hypotheses which are not anyone's personal property"
when you hear about any scientific explanation, "what experiment could disprove your hypothesis?"
Jorgensen, Sven Erik (2002). Integration of ecosystem theories: A pattern. Third edition. Kluwer Academic Publishers. Dordrecht/Boston/London. 420 pages. QH 540.5 .J67 2002.
pg 5 "in complex systems, one can give posterior explanations of events, which one cannot predict with complete certainty"
pg 7 -Heisenberg dilemma extends to ecology " a science of parts cannot explain the multiscale reality of wholes"
pg 8 - analysis (as stressed in current science) pulls out statements that are true and stand on their own, synthesis is trying to put these together into a whole
pg 10 - four approaches to studying ecosystems; empirical studies, comparative, experimental manipulation, and modeling
"it is necessary to use all four approaches" , <!-- Waldo study has only two of these -->
pg 18 - "We cannot capture the complexity as such with all its details, but we can that ecosystems are complex and we can set up a realistic strategy for how to obtain sufficient knowledge about the system, "
ecosystems are complex adaptive systems
pg 29 - the environment includes the history
pg 40 - uncertainty principle indicates that it is impossible to make enough observations about a system to describe it
quoting Bohr "It is not possible to make one unambiguous picture (model) of reality, as uncertainty limits our knowledge"
and it fits with Kierkegaardien view, the multi-aspect view. "reality can only be fully comprehended only from a different view in terms of disparate conceptual schemes."
pg 41 - ecosystems are irreducible systems, can't make observations and then reduce that to a simpler set
"indirect effects are often more dominant than the direct ones"
need models to deal with irreducible systems
pg 47 "A model can be considered as a synthesis of elements of knowledge about a system."
"We use models, therefore, to reveal holistic properties." models have to be on a smaller scale, either faster or smaller
pg 51 calibration - "find the best accordance between the computed and observed data by the variation of some selected parameters"
verification - "a test of the internal logic of the model", does it react as expected?
validation - "an objective test on how well the model outputs fit the data", state variables may be weighted differently
pg 52 - steps for modeling
a more complex model may have more uncertainty, selecting the right level of complexity is an important consideration
choice of level of complexity is a matter of balance
pg 66 there is an optimum in either knowledge vs complexity or effectiveness (judged from validation) vs. articulation (a measure of connectivity and complexity)
pg 72 using scientific tools to test hypotheses - leads to double doubt
when the model is correct and hypothesis is correct you can get further confirmation by testing the model against the "theoretical pattern"
this text attempts to build a tentative pattern for a theoretical framework