<p>You can use any alpha level, the most commonly accepted is .05. This is your significance level. If your p-value is below your alpha level, then it is statsitically significant and there is strong enough evidence to suggest the rejection of the null hypothesis in favor of the alternative hypothesis.</p>
<p>So if its not significant, then you dont reject?</p>
<p>Correct. If the p-value does not meet the alpha level given (or, if there is not one given then you would choose .05 and explain how it is the commonly accepted level), then you would say that there is not strong enough evidence to reject the null hypothesis in favor of the alternative hypothesis.</p>
<p>so, pvalue<avalue (.05), then it is significant and you reject?</p>
<p>It doesn't have to be 0.05. It can be 0.1, 0.01, or any other number.</p>
<p>Well it can be any number, but .05 is the most common alpha level chosen.</p>
<p>Glucose,</p>
<p>Correct, if p-alue < a-value, then there is strong statistical evidence to suggest a rejection of the null hypothesis in favor of the alternative hypothesis.</p>
<p>Ok let me see if I got this right
p-alue < a-value = significant = reject Ho
p-value > a value = not sig = accept Ho</p>
<p>Yup that is how it goes.</p>
<p>Ya....I always use .05 because it seems most logical for me.</p>
<p>No you don't accept Ho you just say there is not strong enough evidence to reject it. I think.</p>
<p>Well ya...but to me it basically means reject. You can only tell that there is substantial evidence, but in statistics you can never "accept" or "reject" something. Also you can't say why something occured (causation), but just that there is a pattern of something happening.</p>
<p>i am scared about the stat test :(. God I am taking both calculus ab and stat and i think i will get a 2 on both.</p>
<p>I know what you mean. We took a practice test, and I had no idea! I felt so confident about this test being easy also.</p>
<p>we've used the following</p>
<p>p-value < alpha - statistically significant evidence to reject Ho.
p-value > alpha - not sufficient evidence to reject Ho; the model seems appropriate.</p>
<p>The P-value is just the area to the outside of your critical values. So if p>significance level, that just means that your critical value lies inside your confidence interval.</p>
<p>Another question:
Do you ever use the t-chart for sample proportions, confidence level for proportions, etc? Or is it only for means?</p>
<p>Z goes with proportions and T goes with means.</p>
<p>Silver, that's wrong.</p>
<p>The difference between z and t scores is based on the sample size. For n>40 you use z scores. For stuff with small sample sizes, you use t scores. When you use t scores, you also do not have the population standard deviation. You only have the standard deviation of the sample. This is the difference between z and t scores. At large sample levels, you can assume that the mean of the sample and population is equal. Then, you use z scores.</p>
<p>You can use either test for proportions or means.</p>
<p>neb my stats class has learned something different.
for Z tests, the standard deviation of the population MUST be known.
otherwise, you do a T test. simple as that.</p>
<p>for large values of N, if you don't know the standard deviation of the population, you still use T, but there is a p-value for df(degrees of freedom) = 1000, which is effective at high values of N. it's pretty much equal to the z-score table, so guess there isn't much difference.</p>
<p>however silver did have a point.
for proportion tests, i dont think i've seen a proportion T test. ever.
but I may be wrong. I will consult my grapher soon (left it in school somewhere, spring break ending tonight)</p>
<p>I learned: t-test=n<30 and sigma unknown
z-test=n>30 and sigma known</p>