BEST REVIEW I'VE SEEN YET!! Thank you so much for simplifying and making it SUPER easy, I was stressing because I have my exam in a couple hours and had no clue what to do.
For anyone reading this, don't forget that you prepared hard for this, and you'll do absolutely fantatistic in this test! Good luck, and let's all get that 5!
Hello! I just wanted to say thank you so much for making these videos. Because of you, I actually managed to score well on this exam, which is something I never thought I would be able to do. Thank you for all the effort you put in to these! :)
If you're running a hypothesis test, you should be writing out all your hypothesis, test statistic and standard error values, the Pvalue, and comparison with the significance level to justify your conclusion. You still need to write all of that out. The calculator just helps you calculate the specific numbers correctly.
Hi! I've been taught that, even if the sample is large (>=30), I should use a t-test. At least, this is what my book said. It also recommended that I only use a 1-sample z test for mean only when we know the population standard deviation (and not if the sample is sufficiently large). Is this correct? Thank you!
Yes that is technically correct. However as the sample size increases, the t distribution looks a lot like the normal distribution. So it's acceptable to get your answer. You have to remember that all of these distributions are approximations. So we are really discussing how "accurate" you're trying to get. A t distribution with 1000 samples is practically identical to a normal distribution. At some point, you're just splitting hairs on the calculation.
@@edbrains9742 oh man I'm so sorry to hear that. I hope the score comes out more favorable and that they scale the score easier given the different format. At least it is done and you can put it behind you.
Yes. You can use the pooled proportion to check for normality as long as your null hypothesis is that the proportions are the same (and this, their variance is the same)
Here's an example. If I have a bag of 5 red marbles, and 5 blue marbles. With replacement: if I draw two marbles with replacement, that means I draw 1 marble. Then put it back, and draw another marble. The probability of getting two red marbles is 1/2*1/2 = 1/4. If I draw without replacement, then I don't put the first marble back. The probability of drawing two red marbles is 5/10 * 4/9 because I already removed one red marble when I draw the second one. Without replacement means each draw is not independent.
