Please enable JavaScript.

Coggle requires JavaScript to display documents.

Pt 5 - To what extend do you agree with the claim “all models are wrong,…

- - - - 1) It is built on mixture of physical marks and mental images
        
        Physical marks are visuals and diagrams that illustrate the whole function of scientific modelling.
        
        Mental images are mental process and analytical thinking flow that the scientist has taken in acquiring the scientific model
        
        The presence of mental images can open up to deviate models from real life and make it wrong as elements such as cognitive bias which unconciously opens up to make them expect what they see is correct (Kahneman D., 2011)
        
        Mental manipulation of data are prone to mistakes and can be even appear clearly in more difficult models and tasks (Shah, P., 2005)
  - - - Simplification and Idealizations can be performed to simplify and for analysis, but some errors could be made too
      - Errors must be argued and proven on why with these errors can still pinpoint nearly accurate representations of the model
  - - - 1) Models are Incomplete
        
        Models are just abstractions of reality
        
        Just to ensure to capture all the critical elements of real world to make it applicable
        
        If the insignificant features are omitted, it should not greatly disturb the model
        However, there will be trouble in interpreting data if the main features are left out
      - 2) Models are merely forecast of future events like how past events happened
        
        Models are built on past recorded data. Instead, they are utilised for future outcomes
        
        If there are changes in the relationships that exist, quality of model will decrease
      - 3) Problems in Data
        
        Mathematical models mostly rely on data for it to work properly
        
        Data may be incomplete and pure estimates
        
        Methodology in obtaining data may also affect the accuracy of data which will lead to wrong relationships in the model
      - 4) Apparent Precision of Forecast may be Misleading
        
        The output does not mean that it is already accurate or precise
        
        Confidence bands are used to confirm the range of values that the output may deviate from the output with a certain probabilty
        
        Simple example: Role of error bars which included Uncertainty propagation in graph plotting to show possible values that may deviate from the original point
        
        Simple modelling may have exact, simple and accurate methods to determine this range.
        However, very complicated models does not have simple analytical techniques to get this range
  - - - Overfitting of data refers to how the data is specifically taylored too precise that the fit becomes too close or even exactly how the models are meant to be shaped.
      - It's dangerous as it hinders the model to form new generalisation on new data. If a new set of data could not be generalised, it cannot do predictions and make classification which is intended for
      - By overfitting, the training data for the model could be significantly wrong and when a new set of data with different discrepancies come, it negates new data because it does not resemble the training data that is presumed to be correct
- - - - Is it inevitable for knowedge to have weaknesses as it develops more advanced?
  - - - How the delivery of knowledge should be justified by clarifying its weaknesses?
  - - - Models with justified limitations create Strong and Reliable methods
  - - - Can a model still be accurate despite errors introduced?
  - - - Does Different purposes influence the function of the
        model
  - - - Can limitations negatively affect the usefulness of the model
- - - - High Speed Objects - Special Relativity Theory
      - 2) Very Massive Objects - General Relativity
        
        Even this theory has holes in it for discovery
        (Emspak Jesse,2023)
        
        Einstein’s theory does not fit into quantum mechanics, because the three other fundamental forces do fit into the equations
        
        Nicholas Yiunes, a physicist at Montana State University states that the bar is getting higher - Means that some aspects of general relativiity might be improved in the future
        
        Despite that, the theory has accomplished to examine space like any other physical quantity which Newtonian mechanics could not
        
        It has predicted black holes and succesfully proves time dilation theory
      - Very Small Objects - Quantum Theory
    - - Etc: Calculations provide estimate speed and acceleration in Newton’s 2nd Law of Motion which is useful for predicting speeds before performing harder calculations which may involves other aspects such as drag, air resistance and friction
        
        Provides the first step begore moving to complicated calculations
        
        One of the most accurate and simple models in physics
        (Xin-she Yang, 2017).
        
        A question can be raised: Is it practical to introduce complicated partial diferentiation and high level calculus to high school students who just learn physics?
        That’s why technically in reality it does not represent the real motion, which is “wrong” in its essence, but it can benefit new students by simplification to physics and get approximate results
  - - - Usefulness: It was the basis and foundation for understanding Quantum Theory Which led to applications such as thermal imaging, astrophysics for star temperature calculation and understanding molecular structure and energy in Chemistry elements
      - Its limitations:
        
        1) It only describes radiation as discrete packets of quantamised energy, not the dual wave-particle behaviour of the atom. Later, scientists in the likes of Louis de Broglie and Einstein suggested that electrons and other discrete matter also have wave properties too.
        
        Only electromagnetic radiation was recognised for quantization which just maps continuously infinite values into smaller discrete values. Other physical quantities like angular momentum, matter and electrical charge are not quantized. The concept must be generalised to address this problem.
  - - - Basic concept for ecology in six decades
      - Simple and useful
      - Applies to all ecosystems; it can be said as universal
        
        EXAMPLE: At a lake, there's the phytoplankton, dissolved organic matter (basal species) and zoo plankton
        The forest has forest litter and vegetation, herbivorous insects, birds and mammals.
        
        Can be compared and studied
    - - 1) More than 1 species can occupy the trophic level and form food webs which happens more realistically
      - 2) Does not consider climate and weather to affect the trophic pyramid where energy change is varied
      - 3) Difficult to specify and assign an exact trophic level to a species due to complexity in biology