Biased and unbiased estimators from sampling distributions examples

Biased and unbiased estimators from sampling distributions examples

[pratikpwagh]

Studying for CFA Level I and this video is one of many KA videos that are making my study a whole lot easier. Thank you!

[tapoutluke]

thank you sal khan, you've been very helpful.

[dewinmoonl]

1 minute in, already can tell this is good

[Dekike]

Thanks for this wonderful video!!! Can I ask you something? I am doing my PhD and I wonder if, with my data, I can say that my sample is biased or not. Following your example in the video, imagine that I have a population of 4000 balls, and I know the values for my balls. Then, using a novel technique, I obtained 300 out of those 4000 balls. And my question is, is it my novel technique biased in the type of balls it gets? In my real case, what I did is to get the 4000 balls with one technique (A) and 300 balls with other technique (B). The problem is that I can not repeat the process of obtaining balls with those two techniques since it would require thousands of dollars. In my case, the data distribution is sharply non-linear for both techniques, with the majority of the values close to 0 and then, the further from 0, less frequency of values. What I did so far was to run a non-parametric test (Wilcoxon Test) to check if A and B belong to the same population. I think that the Wilcoxon Test accepts unbalanced samples. However, I colleague say that I can not talk about "bias" here, since to prove bias from a statistic way I would need repeated measures, like in your example, where you took 5 balls 50 times. In my case, I took 300 "balls" once. Wilcoxon test results say that the values from method A are significantly higher (P<0.001) than for the method B. However, as I told you, what I am interested in is on show evidence about that the method A gets a biased sample. Could you tell me if I can talk about bias in my case although I do not have repeated measures? Or, is there a way to do the analysis to be able to talking about bias? For instance, Monte-Carlo permutations?

I would appreciate your

Thanks in advance!!

[hiel]

Thank you for this great video. Just one point of clarification for me (I'm learning statistics for the first time in a college 200 level-class). In the second example, are you using "population parameter" (where you drew the line on the 5) interchangeably with mean or median? Is population parameter, in this example, used as a general term to describe a center point in the population distribution that's not a mean or median? If so, what is a population parameter in the example. I'm asking so I have a better understanding of why I should care that the sample distribution is a bias/unbiased estimator in regards to the "5."

[plttji]

Thank you for the video, can you help me how to prove that is unbiased in this question? Question: Compare the average height of employees in Google with the average height in the United States, do you think it is an unbiased estimate? If not, how to prove it is not mathced?

[giacomocaio]

Hey there! Why I can't find this video on Khan Academy app?

[AhmadFauzan-zn]

I have a question. how do we tell the sample median is biased or unbiased if we don't know the population median? is it possible?

[claudiamesaaparicio]

bravo! great video
but i have some doubt in other exercise, can you help me? i really appreciate it....

[jeavon]

Unbiased due to the variety of times he calculated the result.
Statistic B

[bigcat]

That first distribution looks pretty flat, you'd expect a more normal distribution.

[innosanto]

What if I don't know the median of the population. Howdo I find if sampling is good?

[spencersparks]

yes 6th comment even tho it was published in 2017

[BravesDahn]

Looking up video on estimators and I find my name in the problem🙃

[kiwi]

sal khan give me some of your fortune😺

[michelkoenig]

i want someone to cut the video like that he only says ball and balls