You are viewing the chat in desktop mode. Click here to switch to mobile view.
X
MTH 220 May 20 Open Questions
powered byJotCast
Prev Page View Full Chat Next Page
Kasey
8:26
oh yeah got it. lol
AvatarGreg Miller
8:26
Well, you don't pick a signficance level (alpha) unless you've done a hypothesis test.. you know.. test Ho/Ha...
Kasey
8:25
what do you mean "if you've done a hypothesis test before"?
AvatarGreg Miller
8:25
Yes... if you picked an alpha.. that predestines what the confidence coefficient will be...
Kasey
8:25
Gotcha but if we determined a significance level before hand we need to.
AvatarGreg Miller
that just means I am covering all the different alpha's/confidence coefficients people MIGHT have picked
that doesn't mean you should have done them all..
8:24
So.. Kasey.. on a key you might see me write out two or three different CI calculations...
Kasey
8:23
cool.
AvatarGreg Miller
8:23
yes, the confidence coefficient should always partner with alpha if alpha has been chosen beforehand.. always.
Kasey
sorry just clarifying
which would be 99%
8:22
so you are saying if there significance level is .01 then we should go with the partner confidence coefficient?
AvatarGreg Miller
sometimes problems ONLY want a CI.. it's traditional to do 95% in those cases in the absence of any other info...
8:20
alpha = .05.. 95% CI... etc...
alpha = .01... 99% CI
But, if you've done a hypothesis test before.. then use 1 - alpha
8:19
Good question.. the general answer is yes.. .if the problem ONLY asks for a CI
Kasey
8:19
Sorry that was me... I have to hit refresh and it takes my name out of the guest box.
Guest
8:18
When the question asks for a confidence interval do we just pick the confidence coefficient that we choose?
Kasey
lol
8:13
yay! I'm glad I'm doing these calculations at home as if I were in a library or in public people would wonder why I'm talking to myself and congratulating myself when I put stuff together!
party 1
AvatarGreg Miller
8:11
BINGO
Kasey
8:11
Oh this is where I apply reject Ho only if th devalue is less that alph
AvatarGreg Miller
8:10
Not Rare is "higher than my alpha level"
Rare is.. ."lower than my alpha level.."
Being on chart literally has nothing to do with "rare or not"
8:09
because there are values like .005 and .01 and .025 represented on the chart.. and if you pick an alpha level bigger than those... then they are ALL RARE to you.
8:08
No..
Kasey
8:08
If our p value is on the chart does that mean it is not rare?
AvatarGreg Miller
all three of those posts flashed at the same time.. you must have hit enter at the exact time I did.. weird.
8:07
No problem
Kasey
8:06
Sorry didn't see that explanation before I sent that last one.
AvatarGreg Miller
"your test statistic value"
the p-value REQUIRES the "equal to part".. read the first part of definition
8:06
No..
Kasey
8:06
Our original Ha had greater than or equal to?
AvatarGreg Miller
8:06
"or more extreme" means.... "direction of Ha"
Here's the reason why.. the definition of p-value is the chance of your test statistic or one more extreme if Ho true
8:05
know*
Connecting…