Originally Posted by masque de Z
(This post is only long out of respect for the possibility that MrWookie will finally understand me)
MrWookie are you here to throw insults or discuss properly? Are you here as a moderator even? If i can discuss things with you without personal attacks why cant you do the same?
This post cannot be avoided to be that long. If you want to continue to attack me you must know exactly what on earth you are talking about. This is your chance to find out why you completely are missing my point.
I gave you an example of a class of pranksters where adding a few habitual pranksters in it or adding a single frequent prankster alters the probability you have a prank in there. It is not negligible. Do the math to see why its not negligible in the 3 cases i offered when you have 1 with 5/10, 10 with 0/10 , 5 with 1/10 and 4 with 2/10 vs 10 with 0/10, 5 1/10 and 5 2/10 vs 10 with 0/10, 5 with 2/10 and 5 with 3/10.
You accuse me of not knowing what i am talking about because i didnt do the math and did you maybe do the math? Can it be that i "did the math" in my mind because its simple enough for its purposes and also because the entire thing is complex enough to not have exact absolute numbers, only a relative estimation is possible and also because i didnt feel the need/point to use it in any action the teacher follows because it is still a small probability? The probability we have a bomb here doesnt rise to dangerous or significant levels because the kid is muslim. Other things do that. Its still a small probability even then.
Why dont we do it properly now?
Before we do it lets consider that a teacher doesnt know at all who Ahmed is or any other student. For some magical reason is given only their ethnic origins. Notice that in general this is not the correct way of approaching things that have greater context. I do not see people that way when i interact with them.
Do you accept more or less the following basic starting points?
1) A random student has probability to bring to school an engineering project that is Pe. A muslim student has also the same Pe. Why same? Because we assume kids are exhibiting same tendency to be creative and scientifically interested, smart etc across all ethnic groups say. We do not know what Pe is but imagine some small number per day. There is also a probability Pb that the student brought a bomb to school. Is it the same for both muslim and random student? We will examine that later. We can even imagine the avg student or the muslim student have a prank probability to bring to school a project that looks engineering or bomb but is a prank/hoax instead. Call that Ph. Do we have a reason to think Ph is different for a muslim than a random? I dont think so. One may argue if the prank is related to bombs its higher for a muslim possibly, because why would a non muslim make a prank that related to islamophobia more often than someone that has reasons to think it exists to his group or personally experienced it etc. But lets ignore that for now and imagine its the same for all groups. After all the teacher is concerned at this point about whether this is a bomb not the other alternatives that are manageable. Also all these probabilities on a per day basis are very small each.ie Pb,Pe,Ph<<10^-3 for all students, muslim or not.
2) A random muslim (before knowing anything else about them) in the general population is roughly 30 times more likely to be related to a terrorist plot than a random person. Where you meet them etc will greatly change that number. That comes from the fact that eg since 2013 we have had 30% participation in terrorist events of that 1% of the population. If it was 1 to 1 with a random person they would have participated in only 1% of the terrorism plots not in 30%. We can further refine that number by considering not old people or not middle age and older women say or preteen kids etc but lets say we didnt do that yet. That number can also be a changing function of time relating to global events etc. For example it was much lower 20-30 years ago, much closer to 1 actually in US. That number can be large for other groups too. Groups affiliated with anti government or neonazi campaigns etc are much more elevated even. For some groups it may even be much smaller than 1. I bet its much smaller than 1 in the group of top students in a class if we could know them in advance. It will be also elevated for kids that have been bullied in a class. If we compare a muslim kid or young teen with the avg for the entire adult population (muslim and others) it will be smaller than 1 there too. The probability a person is related to terrorism depends heavily on the age and many other details. It likely is a rising function until some age and then it declines again. It makes absolutely no sense for this reason to imagine anything about random people you meet. Its very wrong to do so. All kinds of information/context/action can rally or crash this number.
3) A very young kid from that muslim group say 10 or less is probably as likely to attempt a terrorist attack as any other kid in the population, ie some ridiculous unreal low number. So the very young muslim kids start at 1 vs random kids. All very young kids are equally unlikely. What this means is that Pb'/Pb=1 for all very young kids (Pb the bomb probability per day for random kids and Pb' for random mulsim kids). Aging and experience changes these things. This is because different groups live in different situations and face different struggles and abuses or exposure to ideologies, worldview perspectives etc. But it starts the same for all likely.
4) At some point as they age the 1 becomes order 30 and higher (say the avg age of a terrorist participant is 25 or whatever it is and that number may be 50 or 100 for the younger males 18 to 35 etc and only 30 for the population avg). Is it possible to assume for teens its also elevated and somewhere in between 1 and 30? Do we know that number exactly? Of course not. What is a good guess? Hard to say. I say the root of 30 is a crude nonlinear fit on things that increase with age reaching a maximum at some point with a very slow start from the beginning. That implies something near 5. So do you agree the avg muslim teen is ~5 times more likely than the average random teen to be involved in a terrorism related plot involving bombs? What does that mean ? That say Pb'/Pb~5 where Pb' is the chance a random muslim teen (14y) will bring to school a bomb a given day and Pb is the chance a random general student would. I can agree if one protests whether its 5 or 10 or 3 or 4. We do not really know that number. But we do know it cannot be 1 anymore (only compared to other teens here) because we are on the way to 30. Social pressure exists already. It doesnt magically start at say 18y age at value 30. It has some more complex distribution that is not perfectly known so that over the entire population the avg is 30 or whatever it is is exactly at the time today (very hard to know, anything other than its larger than 1 by a lot).
5) Keep in mind that the relative chance of that ethnic group to be in a terrorist related plot increases with age but the probability you are involved in a terrorist plot in any ethnic group is also a sensitive function of age. So obviously the probability a young teen is a terrorist is vastly lower than the probability an adult is, same for muslims. If you were given to choose who tried to attack, an adult or a young teen, you would choose the adult because they are many times/orders magnitude more likely. In fact in this country the statistics are not enough yet to have that number for teens (bombs since 2013 say). We can look of course worldwide for such examples of teens attacking with various methods involving explosive devices. We can also look in other cases that teens are not violent but are starting to get influenced by terrorist groups etc.
6) Can we go now to school and see a device and before examining it up close imagine its a bomb risk with some small probability? The only problem now we face must be to protect the school, determine better the probability etc. What if we stopped and did something i have suggested in this thread to not do as a teacher at this point when the probability looks small anyway. We considered the ethnic origin of the student involved in the accusation of possible bomb. Does the chance to be a bomb increase if the kid is a muslim? Keep in mind we do not know if the kid is muslim just by looking at the name or appearance etc. Its an unreliable way in a modern country with many ethnic groups. But lets say that someone forced us to do it because we had to. We examine nothing else like what clothes the kid wears or past history or if they look scared or confident or nerdy etc.
The reason you didnt see an explicit calculation from me is because these numbers and distributions above are not accurately known. And because its not the correct thing to do in this situation because other issues have higher priority. We only discuss it here because you made it an issue to be discussed on the basis of a racism/bigotry accusation.
What is our problem now? If you agree with the above points we need to try to answer the following question (or if not, modify them to a level we can agree, otherwise you lose the right to insult me on bad math and other crap because we do not have a common starting point to do so and you cannot possibly know if its bad or not then). Is the probability to have a bomb affected by the knowledge the teen is a muslim of age 14?
The probability something is a bomb is affected by how it looks and a more detailed examination. If you do that here it goes to 0 fast. But the discussion was about the first moment one is faced with such possibility of bomb before knowing anything else with the alternatives being bomb, prank/hoax, engineering project. I do not know how the English teacher thought here. I tend to think she was mostly irritated at the prank possibility which by the way is the wrong priority actually as establishing risk is 0 is more important.
All i am arguing is that the mix bomb, prank, engineering project is different for a muslim than a random kid. The chance its a bomb is still very small in that case too and accusing the kid its a bomb is a false accusation the vast majority of the time.
If you have a problem with what i say you have it because you argue that points 4 and 5 are not correct. So argue that or any other points 1-6. You do not argue with me because i failed a Bayesian calculation. The real "Bayesian" here is very complex problem. I only crudely treated it because nomatter the error involved (in knowing how things change with age etc) i remain confident there is a relative elevation above 1 for that group. For bullied kids it would also be elevated and likely higher too.
How does it become a ratio of 30 for the general population if the teens do not also have an elevated relative probability? The true Bayesian calculation here is far more complex than most people imagine it. It is actually along the lines of this;
What is the chance the object is a bomb given that the kid is 14 and a muslim. How does it compare with a random teen. That number would be different if the kid was muslim and of age 22 or a muslim child age 10. Its close to impossible to have exact number for this, only offer legitimate qualitative arguments for how they evolve. Furthermore the number completely changes when you start adding other properties to the person or the situation. I see no reason to think of such numbers while you act even if they are there.
The only Bayesian process here is the one done when you assign a probability boost given the knowledge the subject belongs to a group and has a certain age ignoring all else. You look at the distribution of that group with age in the Pe,Pg,Pb numbers.
The "Bayesian" or the relative conditional probability here goes from eg the knowlegde its 30 for the avg population in a random setting to the distribution with age for that group and to the value at 5 or whatever bigger than 1 it turns out to be for that age and setting. Its a "multidimensional" distribution because it changes not only with age but also with the setting ie school vs market vs transportation/buildings etc)
Since this number is significant>>1 we can imagine it. We do not even have to mention it anywhere other than our mind. The overall probability for a bomb remains small and making it 5 times larger than a small number will still leave it small. This is why it must never be a concern that the child is a muslim at this level. The concern is to examine the device better and eliminate the risk in objective terms. This why any teacher that treats the situation is never supposed to be influenced by what is the ethnic background of the kid. Kids do not build bombs as often as adults do. They are negligible risks. In negligible risks it makes no sense to wonder about how something influences a small number. That only matters if the numbers are very big. Only the device itself can make that number look much bigger. So look at the device or if this is seen as risky check other details such as the kid's behavior, past history, character, emotional state and the overall context, size of the object, details seen from far away etc. All these things together take the number to 0 very fast in this case.
What can be some example of these P numbers for teens ~14?
Lets me guess them more carefully for the first time on a per day basis if the action is not encouraged by school (as part of a project so completely spontaneous)?
Pe~10^-5, Ph~10^-6, Pb~10^-10.
Those numbers are intending to give you a general idea based on common experience and news. The exact numbers are obviously not available. They only help here to give you an order of magnitude appreciation for what i am saying.
So what am i really saying here? That maybe for a muslim teen this goes to
Pe'~10^-5. Ph'~10^-6. Pb'~5*10^-10.
See what i mean? On the day the random kid brings to school something imagine it has only 1 in 10^5 to be a bomb and 5 in 10^5 to be a bomb if the kid is muslim. It might have been 1 in 10^4 for some bullied kid or some hate group kid etc. All these numbers have huge uncertainties.It still has of course eg ~90% chance to be a legitimate thing and ~10% to be a prank or a failed legitimate thing that looks like prank by coincidence, naivete or poor choice. Those are some guesses.
If you have a clue the object looks like a box and has a timer in it then of course we may change these numbers to 1 in 10^3 and 5 in 10^3 or something. They are still small numbers and the teacher must never think in terms of ethnic origin etc at this level. Closer examination of this object will take it to 1 in 10^5 or 10^6 levels and soon enough to 0.
We must only think in terms of such numbers if the population is say students eg age 18-22 where then the above numbers change remarkably and the ratio from 5 may go to 30 or more even but also the overall bomb chance due to age situation and only if thats all we can do and it can affect any protection taken etc. In a crisis its not important at all at first level to think anything other than the properties of the object and the situation. You do not calculate numbers in a crisis as things happen.
You may object to me assigning a muslim teenager a 5 to 1 ratio. I have no problem with it. You know how i use it if i were a teacher? Understanding that others may exaggerate this number too much, including his classmates, i anticipate a random muslim teen to be abused or mocked by others while at school or to have seen things happen to his family.
So when i interact with him i will not be more careful than with others students to avoid an aggression risk. I wont be afraid or cautious like the prejudiced idiots. I will be more careful not to contribute to the perceived bias/phobia in others but do it in a way that appears genuine and not intentional (like white guilt) to go on the other side. That way the student doesnt feel they are specially treated with more care than others but treated only slightly better even in some little things and is offered chances to show their skills and interact in outside projects and overall feel encouraged because there are people who believe in their potential. So i may go the extra distance with such student to offer them a chance to see another side of bias on the plus side and do it in such a way that its never at the expense of other students to create any animosities. I would want to meet his parents and talk to them and suggest books, science related activities or hobbies etc. I will give hints that i care and expect from them great things. I will investigate if they are bullied too. This is how you fight racism. You do not do it by ignoring underlying realities of our world. You do it by not seeing these realities for more than they deserve given their uncertainty and by living your life in such a way so that the statistics/realities will stop showing a difference one day for all groups indicating society finally treats in life all similarly for such animosities to not exist.
You do the same for other groups that are in higher risk also like bullied kids or kids in hate groups etc trying to balance things a bit. You try to prevent harm and offer the kids and society at large (because it has a role too) the chance they need to never become part of the real reason the ratio in adults is ~30 or whatever for that group. This is how you win. You do not win by being wrongly angry at me or the truth.
