Imagine that you have just invented a statistical test...

Imagine that you have just invented a statistical test called the Mode Test to determine whether the mode of a population is some value (e.g., 100). The statistic $(M)$ is calculated as $$ \mathrm{M}=\text { Sample mode/Sample range } $$ Describe how you could obtain the sampling distribution of M. (Note: This is a purely fictitious statistic.)

Imagine that you have just invented a statistical test called the Mode Test to determine whether the mode of a population is some value (e.g., 100). The statistic $(M)$ is calculated as
$$
\mathrm{M}=\text { Sample mode/Sample range }
$$

Describe how you could obtain the sampling distribution of M. (Note: This is a purely fictitious statistic.)

Key Concepts

Sampling Distribution

A sampling distribution is the probability distribution of a given statistic based on a random sample. It describes how the statistic (which could be a mean, median, mode, ratio, etc.) varies from sample to sample when samples of the same size are repeatedly drawn from the same population. Understanding the sampling distribution is critical for making statistical inferences about the population parameter.

Monte Carlo Simulation

Monte Carlo simulation is a technique that uses repeated random sampling to compute the properties of a statistic. By generating a large number of simulated samples from the population and calculating the statistic for each sample, one can empirically approximate the sampling distribution. This method is particularly useful when an analytical derivation of the distribution is complex or infeasible.

Resampling Methods

Resampling methods, such as bootstrapping, are techniques for estimating properties of an estimator (such as its variance or bias) by repeatedly drawing samples from the observed data. Bootstrapping involves sampling with replacement from the available dataset to create many pseudo-samples and then computing the statistic for each one, thereby approximating its sampling distribution without relying on strict parametric assumptions.

Transcript

00:01 All right, so we're playing with a simulation applet, which is pretty fun to play us to get a concept of what's happening with our statistics, with our distributions and samples and all that good stuff.

00:15 So we need to get 100 random samples of size 5, and we're going to look at, or describe the distribution of the sample mean based on what we see.

00:25 And then according to theory, what is the distribution of this sample mean? and we're going to do that for size 10 and size 30.

00:36 So is that.

00:37 And then we're going to compare a, b, and c.

00:41 And how are they same, how are the different? but we're given that the population we're going to create is bell -shaped, which means it's normal.

00:51 So the population itself is going to be normal.

00:53 So that means regardless of the size, the sample will be approximately, the sample distribution will be approximately normal.

01:01 Because that's what the central limit theorem says.

01:04 Central limit theorem.

01:07 Kind of a big deal.

01:09 Central limit theorem.

01:18 So hopefully at this point, this is not, this video is not the first time you've heard of this, but that's kind of a thing.

01:28 Kind of a big deal.

01:29 So we can do this.

01:30 So here's the applet.

01:33 We want to do bell -shaped distribution.

01:38 And so here's the distribution of the population right here.