Interpreting the sample mean, $\bar y$ - Cross Validated most recent 30 from stats.stackexchange.com 2022-01-21T04:54:41Z https://stats.stackexchange.com/feeds/question/88253 https://creativecommons.org/licenses/by-sa/4.0/rdf https://stats.stackexchange.com/q/88253 3 Interpreting the sample mean, $\bar y$ user 31466 https://stats.stackexchange.com/users/31466 2014-02-28T11:32:58Z 2014-02-28T17:16:07Z <p>I have a hypothetical population that contains the values $2,3,4,5,6,7,8,9$. I have to draw a sample of size $4$ from that given population. So I have $\binom{8}{4}=70$ ways to draw the sample. I wrote down all possible samples and computed their sample mean. I got $70$ as the sample mean ($\bar y$). Then I made a frequency table of $\bar y$, showed $\bar y$ is an unbiased estimate of population mean, $\mu_Y$ and drew a histogram of $\bar y$.</p> <p>The histogram shows that the distribution of $\bar y$ is a bell shaped curve. So at the end of this work, my interpretation of the whole task is:</p> <blockquote> <p><em>Though the parent population follows uniform distribution, the sampling distribution of the sample mean, $\bar y$, tends to Normal distribution.</em></p> </blockquote> <p><strong>Is that all there is to interpreting the whole task? I want to interpret most appropriately so that I can get the whole picture.</strong></p> https://stats.stackexchange.com/questions/88253/-/88254#88254 0 Answer by Vaaal for Interpreting the sample mean, $\bar y$ Vaaal https://stats.stackexchange.com/users/28509 2014-02-28T12:02:25Z 2014-02-28T12:02:25Z <p>As told you in the comment, it is not clear what you are asking. You have got a normal distribution because, as the Central Limit Theorem says, "given certain condition, the arithmetic mean of a sufficiently large number of iterates of independent random variables [...] will be approximately normally distributed". </p> <p>Now that you have got the distribution of $\bar y$, you can make some statistical inference based on it. For example, you can take a sample from a population that MAY or MAY NOT be the same as the initial population. Calculate the mean of this new sample, let's call it $\bar y_j$ and compare the "value" of $\bar y_j$ with the histogram of $\bar y$, you can actually have the feeling of how likely $\bar y_j$ is drawn from the same population as $\bar y.$</p> https://stats.stackexchange.com/questions/88253/-/88299#88299 2 Answer by gung - Reinstate Monica for Interpreting the sample mean, $\bar y$ gung - Reinstate Monica https://stats.stackexchange.com/users/7290 2014-02-28T17:16:07Z 2014-02-28T17:16:07Z <p>Yes, I suspect that is the gist of it. I have provided similar exercises to students. </p> <p>Another thing you can do is see how the sampling distribution of the sample mean gets narrower as the number of data in your sample increases (although that does not seem to have been a part of your assignment). For example, if you sampled only $2$ data from that population, the highest sample mean you could get is $8.5$, and the lowest is $2.5$. Consider how the range of possible sample means changes as a function of sample size:<br> $$2\quad 3\quad 4\quad 5\quad 6\quad 7\quad 8\quad 9 \\ \underbrace{\ 2.5\quad\quad\quad\quad\quad\quad\quad\;\ 8.5} \\ \underbrace{3\quad\quad\quad\quad\quad\quad\quad 8} \\ \underbrace{3.5\quad\quad\quad\quad\;\ 7.5} \\ \underbrace{4\quad\quad\quad\quad 7} \\ \underbrace{4.5\quad\ \ \ 6.5} \\ \underbrace{5\quad 6} \\ \underbrace{5.5}_{\mu_Y}$$ When your sample is smaller, there are a wider range of possible estimates that you can get. As your sample gets larger, the range narrows. In the end, when you have the entire population, you necessarily get the population mean. (In fancy terms, the sample mean is a <a href="http://en.wikipedia.org/wiki/Consistent_estimator" rel="nofollow noreferrer">consistent estimator</a> of the population mean.) </p> <p>However, it isn't simply that the <em>possible</em> range narrows, but you are increasingly likely to get an estimate that is closer to the true population mean. You can see this in histograms that display all possible sample means as a function of sample size: </p> <p><img src="https://i.stack.imgur.com/dQhiZ.png" alt="enter image description here"></p>