<p>Can someone give me a quick and easy way to tell exactly which test to use (e.g. matched pairs, one sample z, two sample t). I sometimes have trouble with this--so THANKS IN ADVANCE! :)</p>
<p>matched pairs is using 2 sets of data as one for the same population or sample, usually when each person recieves two treatments.</p>
<p>Here would be an example of a match pairs: Weight loss program is suppose to make people lose weight, they give you the before and after weights FOR EACH PERSON, and they ask you if the weight loss is significant. Then you subtract the final weight from the initial weight, and do a one sample t test on the mean weight loss using the mean and standard deviation from the difference (Post-Pre).</p>
<p>a one sample z test is usualy is usually for proportions, and would be something like: a company claims 50% of their workers are tall, and a random sample of 100 workers was taken and 75 were tall, all you need to solve it is the current standard proportion, the new proportion, and how many people were used to make that proportion</p>
<p>a two sample t test is when you have two means and you are computing the difference, usually you are asked whether there is a significant difference, or is the mean of this is significantly greater than the mean of that. basically the same as match paried except the data is from two samples so you.</p>
<p>but yea im actually really bad at explaining it apparently but basically:</p>
<p>matched pairs = usually used for means with pre and post results subtracted from each other</p>
<p>one sample z = used when when a new proportion is being compared against an old</p>
<p>two sample t test= comparing two means(from independant samples) with the population standard deviation is unknown</p>
<p>hope that helped try to get better explanations, x D good luck!</p>