There are more accurate models " it seems that there is a non-zero probability of finding particles with negative sizes in the sample"
As you know that electron has to be treated as a wave and all answers of its being will be given by its wavefunction and Schrodinger's equation.
for more details, see the following:
https://books.google.com.my/books?id=nYjCjNJ3dSQC&pg=PA105&lpg=PA105&dq=non-zero+probability+of+finding&source=bl&ots=cDNXpO8aF4&sig=FTFUMQL2MZeG2f4ZeAXAbYfe5tg&hl=en&sa=X&redir_esc=y#v=onepage&q=non-zero%20probability%20of%20finding&f=false
Ali -
My guess is that the statement that concerns you is one that comes from someone examining the red curve that you fit to your histogram. I think the comment means that the bottom left corner of the graph should be the origin, and the (x,y) values for the red curve should always be positive, unless you can consider negative values. Part of the red curve you have extends to the left of the origin along the x-axis, where particle size would be negative. I am not commenting on the subject matter, just that I think your red-curve fitted distribution is being questioned. There are points on the curve that would be read as representing negative sizes.
Cheers - Jim
A normal distribution was fit to the data. A lognormal distribution would be a better fit. Nevertheless, the statement contains the common error of assigning physical attributes to the assumed distribution. The data have physical attributes, The distribution does not, A distribution is a mathematical convenience. Indeed, it is irresponsible to reify any assumed distribution and, worse, a poorly fit distribution. There is no meaning to the quoted statement. As stated there are more accurate models. A more accurate model is, still, a model. A model is not data.
Khalid Al-Hussaini, Joseph L Alvarez, James R Knaub..thanks for all
As Joseph L Alvarez stated it is just an artefact of approximation of the data distribution via normal distribution. It is common that after that you could have non-zero probability e.g. of negative concentration of hormones in blood etc. The solution is a simple - use better approximation of your data (as lognormal) or just avoid to interpret something what is just artefact of data modelling (selection of type of data distribution is kind of the data modelling).
- Badges
- Science topic
- Similar topics
- Elementary Particle Physics
- Particle