3

As I understand it, objective probability is based on a frequency of something occurring over an infinite number of observations. A typical example would be: "the probability of getting heads is 50% because if I flipped a coin often enough, I would get heads 50% of the time".

Subjective probability is your belief in the probability of a single event. An example might be: "the probability of it raining tomorrow is 40%; this is just speaking about a single observation, as you can't repeat the day over and over".

Is my understanding correct? Are these two examples correct?

Dave
  • 1,641
  • 2
  • 14
  • 27
  • 4
    There are many different interpretations of "objective probability," some of which are entirely epistemic in nature! The fundamental problem with the one you cite is that there is no such thing, and can never be any such thing, as an "infinite number" of observations: the coin eventually wears down and (therefore) changes character along the way. Similarly, your characterization of subjective probability captures the subjective aspect but contains nothing to distinguish such a quantitative belief, no matter how irrational or contrafactual it might be, from "probability." – whuber Mar 02 '18 at 22:53
  • Hmm is there maybe a different name for the distinction I'm alluding to? Is it "long run frequency" vs. something else? – Dave Mar 03 '18 at 00:38
  • 1
    Check https://stats.stackexchange.com/questions/173056/how-exactly-do-bayesians-define-or-interpret-probability/237772#237772 – Tim Mar 03 '18 at 19:29

3 Answers3

1

This is what I found on the web without half-trying: Subjective probability is a probability derived from an individual's personal judgment about whether a specific outcome is likely to occur.

A simple, if not especially good example of Subjective Probability

In a scenario in which a person is asked to predict the percentage chance of whether a flipped coin will land with heads or tails up, his initial response may be the mathematically true 50%. If 10 coin flips occur, all resulting in the coin landing tails up, the person may change his percentage chance to a number other than 50%, such as saying the chance of it landing tails up is 75%. Even though he knows the new 75% tails prediction is mathematically inaccurate, the individual’s personal experience of the previous 10 coin flips has created a situation in which he may choose to use subjective probability.

EDIT: As mentioned above, there is a good link for this provided by @Tim How exactly do Bayesians define (or interpret?) probability?. The view that is most commonly associated with Bayesian statistics is subjectivist view, also known as personalistic probability.

The objectivist view, A.K.A. frequentist view is that with infinite data, probability exists as an objective phenomenon with the Black Swan argument often cited as a demonstration of the non-computability of the probability of the consequential rare events using scientific methods (owing to the very nature of small probabilities).

Another example "Not all probability is objectively strictly quantifiable. Premise: “Some X are F and x is X”. Conclusion: “x is F”. The objectivist can only say “The probability (given the premises) the conclusion is true is greater than 0 but less than or equal to 1.” This is because the logical “some” implies “at least some and perhaps all.”

The subjectivist is free to say, for example, “The probability (given the premises and my beliefs) the conclusion is true is 42.8%.” But he does so only by some mysterious introspection which, in effect, adds to or subtracts from the fixed premises. Of course, most subjectivists in practice would agree with the objectivist."

Carl
  • 11,532
  • 7
  • 45
  • 102
  • 1
    The concept of subjective probability is much richer and more nuanced than this. One basic concern is how to connect "personal judgment" to reality--but the definition you quote is incapable of that. As such, it's not relevant to most statistical problems. (That site, IMHO, is a poor source for any statistical concepts.) – whuber Mar 03 '18 at 20:47
  • @whuber OK, I call your rich response and raise you a black swan. – Carl Mar 03 '18 at 20:54
1

The example I always use comes from Ted Dunning (at least, he was the first person I can attribute this to). Typically it's performed live with a real coin and $ N \ge 2$ people.

I have a coin in my hand, and I flip it without revealing it. I ask person 1 what's the probability of heads. They respond 50%. I then show person 2 the result of the flip, without person 1 seeing, and ask what's the probability of heads to person 2. The respond either 100% or 0%. Thus, we have two individuals giving very different probabilities about the same event. Thus probability is subjective.

Cam.Davidson.Pilon
  • 11,476
  • 5
  • 47
  • 75
  • 1
    On the contrary, similar examples are often used to demonstrate that probability is a function of an experimental setup along with a "degree of focus" on how the setup is characterized. Objective probability theories (of several types) have been created from this point of view. – whuber Mar 03 '18 at 21:37
0

this was helpfull for me ,

check this

Please mark my answer if it is usefull for you. I need oints bc i can not comment : (

Enes Kuz
  • 34
  • 3
  • 3
    It seems you're suggesting that subjective probability is non-quantitative. I think many would disagree with that position (e.g. see subjective Bayes for a counterexample). – user20160 Mar 03 '18 at 01:00