<p>Can someone give me a list of assumptions/conditions for Stats that one would need to know or list on the FR?</p>
<p>This is just cold recall, so if anyone sees things worth correcting by all means go ahead:
Z-test: SRS + population is normal OR sample size is large
T-test: SRS + population is normal OR sample size is large OR sample size is moderate without skewing or outliers
2-sample Z: Same as above except as such for both listings, with independant SRSs
2-sample T: Same as above except as such for both listings, with independent SRSs
Proportion test: SRS + sample is less than 10% of population + np and n(p-1) are both greater than 10
2-sample proportion test: Same as above for both, independant SRS
Matched T/Matched Z: Same as one-sample for each.
Chi-square: Each observation is greater than five (might want to check on that one, though).</p>
<p>Chi Square - Expected numbers should be greater than 5.</p>
<p>Isn't it at least 80% have to be above 5?</p>
<p>yes, only 80% of the counts have to be above 5</p>
<p>yeah, no more than 1/5 can be > 5.</p>
<p>i hate our text, it's so convoluted.</p>
<p>This is a very good refresher. I get all the tests and their respective assumptions mixed up.</p>
<p>thank you!!! how well do you think you have to know chi-square </p>
<p>(other than the basic [(o-e)^2/e])</p>
<p>I would say you need to know how to implement chi-square, the assumptions, and when to run a test. Also, be sure to understand how to define your hypotheses.</p>
<p>what will a problem look like when you need to use 1-binompdf....and also, can someone explain the P A B U ^ things...I have NO idea.</p>
<p>I'm pretty sure that these Test Assumptions are all right:</p>
<p>Z Test: SRS; normal or n> 30<br>
T Test: SRS; normal, outliers (boxplot)
2- Sample Z Test: SRS; normal or n>30
2- Sample T Test: SRS; normal, outliers (boxplot)
1- Prop Z: SRS; small part of pop; np > 10 n(1-p)> 10
2- Prop Z: SRS; small part of pop; n1p1 > 5 n1(1-p1)> 5 n2p2>5 n2(1-p2)>5
Z Interval: SRS; normal or n>30
T Interval: SRS; normal, outliers (boxplot)
2- Sample Z Int: SRS; normal
2- Sample T Int: SRS; normal, outliers (boxplot)
1- Prop Z Int: SRS; small part of pop; np > 10 n(1-p)> 10
2- Prop Z Int: SRS; small part of pop; n1p1 > 5 n1(1-p1)> 5 n2p2>5 n2(1-p2)>5
Chi2 Test: SRS; no more than 20% of expected cells are <5; expected cells >1</p>
<p>linear regression t-test anyone?</p>
<p>am I allowed to input that in my calculator?</p>
<p>what will a problem look like when you need to use 1-binompdf....and also, can someone explain the P A B U ^ things...I have NO idea.</p>