Random Sampling:
For random sampling, I would have to make sure that all of the items that I am selecting from are identical in shape, size, and texture (if I am sampling from actual objects). This eliminates a bias approach to the sampling by having different variables that may attract attention more than other items in the same combination.

**Basically, when I go to feel around for a random sample-without looking, all of the items should feel the same.**

For example, I want to mix 100 metal-free elastic hair bands-all of the same thickness and size-just different colors. I am going to put all of these hair bands in a black container with a cover that has a slit to put my hand through and draw random hair ties without seeing them. -This would successfully give me a random sample setting.

Systematic Sampling:
This form of sampling includes taking samples at regular intervals, not excluding sampling work from every input rather than just the work of one input.

** Simply put, we want to pull an individual sample out of an entire population at a regular interval, and make sure that the samples pulled are not from one single worker-but samples from each person/thing working on the item population.**

For example, I am studying an assembly line where there are five people working on the specified job. I know that for every five items that come through the line, each person has worked on one, and they all do equal work throughout the process. SO, to systematically sample this assembly line's success rate, I will take 5 consecutive samples once every hour-this will give me one sample from each individual, and at the end of my sample study I will have an equal number of samples from each of the five workers.

Cluster Sampling:
Cluster sampling consists of looking at entire population of items and dividing the population into equal parts. We would then take one those parts of the population and find our sample variable.

**We are basically taking a set of information from one part of the population and assuming it is the same throughout, thus multiplying the set of information by the remaining parts of the population.**

For example, (I remember this from biology...lol). We want to figure out about how many squirrels are in the 40 acre woods. SO, we would most likely divide the 40 acre woods into 1 square acre plots. Next, we would count how many squirrels we could find in one of the square acres. After finding that number, we'd assume that the density of squirrels is the same for the whole 40 acres and multiply the squirrel count for one square acre by 40. Like this:

1 square acre = 75 squirrels
75 squirrels x 40 acres = 3000 squirrels in 40 acres

Stratified Sampling:
This means that we are looking at a whole population of items, but there are various factors that may differ between one part of the population to another, so we separate the various characteristics into "strata" and then take our random samples.

**Basically we want to separate like items into groups and then sample from the individual groups randomly**

For example, we have 500 tennis balls; the colors are mixed however, there are 5 different sizes-100 of each size. SO, we separate the sizes into individual containers (still considering the colors are equally mixed within each size), and draw random samples from each size's individual container.

Convenience Sampling:
Is a easy way of obtaining information, however it can be biased and is not as accurate as other forms of sampling. We would most likely be sampling the most accessible part of the population-and that may not be the part of the population that will be giving us the most accurate information.

**Simply put, we are obtaining a sample from a population that may not accurately reflect the actual results.**

For example, we are told to find out how many world leaders (of all global countries) have positive, negative, or indifferent feelings toward the United States. We may only be able to access information on the more well-known leaders-leaving out lesser known global leaders that would greatly affect our statistics. We might only be able to get information on countries that are negative or indifferent-shedding a more negative light on the U.S.'s affect on the rest of the world.