<p>Hello,</p>
<p>Does anyone have a minimal list of all of the tests, their assumptions, and if you pool/don't pool?</p>
<p>what is pool?</p>
<p>This would actually be pretty helpful. Anyone have one?</p>
<p>Pool meaning when you “pool” the data. 2-Pr p Z Test, we pool the n values and successes to find an estimated p hat.</p>
<p>Don’t worry about pooling on the AP test. You don’t need to know about it, and often time if you use it, you may get marked down for a less-accurate answer. That’s what my teacher told us on friday.</p>
<p>Never pool in AP. Ever.</p>
<p>Actually you do pool, but only for two-proportion z-test. However, it’s automatically pooled (whether or not you wanted pooled) if you input the data in your graphing calculator.</p>
<p>@Jerry. Really? I always thought it asked you if you want to pool or not. Or maybe i’m just confusing myself with another test on the calculator.</p>
<p>@Jerry: That’s true, but since that’s done automatically it’s easiest to avoid pooling whenever possible for the exam.</p>
<p>If anyone else is interested in the concept of pooling, there’s a good article on AP Central: [AP</a> Central - Why Do We Pool for the Two-Proportion <i>z</i>-Test?](<a href=“Supporting Students from Day One to Exam Day – AP Central | College Board”>Supporting Students from Day One to Exam Day – AP Central | College Board)</p>
<p>And since you all wanted a convenient list of conditions for each test, here ya go.</p>
<p>One proportion Z-test
Assumption:
-individuals are independent
-sample is sufficiently large
Conditions
-SRS
-population>10n
-np>=10,nq>=10</p>
<p>Two proportion z-test
Assumption:
-groups are independent
-data in each group is independent
-both groups are large
Condition:
-SRS
-population>10n for each group
-np>=10,nq>=10, each group</p>
<p>One Sample T-test
Assumption:
-individual independent
-population is normal
Conditions:
-n>30;Central Limit Theorem
-SRS
-population>10n</p>
<p>Two-Sample T-Test
Assumption:
-groups are independent
-individuals are independent
-both populations are normal
Condition:
-both group are normal;CLT
-SRS
-population>10n for each group</p>
<p>Matched Pair T-Test
Assumption:
-data are matched
-individuals are independent
-population of differences is normal
Conditions:
-difference is normal;CLT
-SRS
-all>10n</p>
<p>Goodness of Fit Chi-Square Test
Assumption:
-data are counts
-data are independent
-sample is large
Condition:
-SRS
-all>10n
-all expected counts>=5</p>
<p>Homogeneity Chi-Square Test
Assumption:
-data are counts
-data are independent
-sample is large
Condition:
-SRS
-all>10n
-all expected counts>=5</p>
<p>Independence Chi-Square Test
Assumption:
-data are counts
-data are independent
-sample is large
Condition:
-SRS
-all>10n
-all expected counts>=5</p>
<p>Regression Test
Assumption:
-Form of linearship is linear
-errors are independent
-variability of a errors is constant
-errors have a normal
Conditions:
-Scatterplot looks approximately linear
-no apparent pattern in residuals plot
-residuals plot has consistent spread
-histogram of residuals is approximately normal</p>
<p>(errors being data)</p>
<p>^Very helpful, thanks a lot for posting :)</p>
<p>Don’t forget random sample for all of them.</p>
<p>SRS = random sample. It’s actually possible to have a random sample that is not an SRS, but from what I understand the AP test graders really don’t care much about the distinction.</p>
<p>^Yeah looking at all the scoring guidelines on old frqs, they always accept it if it said “random sample” in the problem as being an SRS</p>
<p>gosoccer2006 also posted a link for Inference Procedures:</p>
<p>[AP</a> Statistics Inference Procedures](<a href=“http://www.docstoc.com/docs/36836754/AP-Statistics-Inference-Procedures]AP”>http://www.docstoc.com/docs/36836754/AP-Statistics-Inference-Procedures)</p>