- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: Graphing t-Test: Two-Sample Assuming Unequal Variances into Bell Curve
Posted on 11/20/15 at 3:54 pm to Epic Cajun
Posted on 11/20/15 at 3:54 pm to Epic Cajun
Or R, or Matlab.
0
Posted on 11/20/15 at 3:57 pm to TigerDik86
Google probably can find you someone who has done the exact same problem.
Posted on 11/20/15 at 4:05 pm to TigerDik86
quote:
non-gamer in-game deaths
20
18
8
13
28
Hardcore in-game deaths
10
12
4
5
15
Are you seriously collecting data about your video gamer friends for fun?
Im not even gonna make any faggy arse virgin loser mofo jokes.
Posted on 11/20/15 at 4:37 pm to TigerDik86
Can the OT help me solve this?
X+1=2
X=C
X+1=2
X=C
Posted on 11/20/15 at 4:39 pm to TigerDik86
Are you a 5* recruit?
Posted on 11/20/15 at 8:38 pm to TSLG
quote:
Are you seriously collecting data about your video gamer friends for fun?
I'm glad you brought that up. I wanted to ask the same.
Posted on 11/25/15 at 9:20 am to sullivanct19a
quote:
Are you seriously collecting data about your video gamer friends for fun?
Nope. I'm in college for video game design. The teacher is a frick nut who hadn't taught anyone how to create bell curves; we never were taught bell curves in Stats either. I'm part of a collaborative effort to playtest and provide an analysis of an actual (shitty) game that is in beta. The team I'm on is considered "hardcore gamers" and we facilitated these tests with non-gamer participants. I wanted to show the difference of Hardcore vs. Non-gamer types in deaths. I've attempted the google and even attempted searches. Over with now, but thanks guys. I knew the consequences.
Posted on 11/25/15 at 9:23 am to TigerDik86
quote:
we never were taught bell curves in Stats
What kinda shitty statistics class did you take?
Posted on 11/25/15 at 9:27 am to TigerDik86
Your bell curve should look similar to this.
You're welcome
You're welcome
This post was edited on 11/25/15 at 9:28 am
Posted on 11/25/15 at 9:28 am to Epic Cajun
Exactly. I started a BS student concern form over this and it has been acknowledged by the department head. Everyone is assuming points back on a test for something that was never covered.
Posted on 11/25/15 at 9:51 am to CptBengal
quote:Yeah. I'm confused because he said Bell curve, which I equate to the normal distribution. Yet, he's also talking about the T-Distribution, which we know is similar, but is distinct since it refers to a sample. Maybe it's semantics for the purpose of his question and it should look like a bell curve either way, but you never know.
What are you trying to graph?
The distributions?
If so, plot the probability distributions.
This isn't hard.
This post was edited on 11/25/15 at 9:56 am
Posted on 11/25/15 at 10:06 am to buckeye_vol
Homework Board
Posted on 11/25/15 at 10:09 am to TigerDik86
You don't plot the t-tests. You plot the sample distributions overlaid on the same graph and then report the statistical tests of independence.
Depending on the program you are using, you should be able to do this with the means and standard deviations for the two distributions.
Depending on the program you are using, you should be able to do this with the means and standard deviations for the two distributions.
This post was edited on 11/25/15 at 10:12 am
Posted on 11/25/15 at 11:57 am to TigerDik86
quote:Well Bell Curve is an informal name given to the normal distribution, which is the most commonly used distribution given the frequency it occurs a s the properties.
we never were taught bell curves in Stats either.
It's called the "Bell Curve" because the graphed probability density function looks like a Bell.
So what I'm saying is that, you may have learned about it, but in a stats class, they referred to the formal name instead.
Posted on 11/25/15 at 11:59 am to TigerDik86
You're not getting any meaningful statistical test with only 5 observations.
Posted on 11/25/15 at 2:17 pm to TigerDik86
Try estimated diff in means equals xbar1-xbar2 eq. 10.2
Est. std. Err of diff. Equals (root(56/5 plus 19.7/5))
Approx eq. root(11.2 plus 3.9)
Approx eq root(15.1)
Approx eq 4 or 3.something.
No reason to think normal underlying so probly not t
Estimated p-value of test of no diffs could be as large as 1/(sq(2.5))
Very Approx. Equal. .15.
Or worse cuz you only have smal sample estimates of variances and hence std. Errors.
So you can't reject possibility of no differences unless you have the ethics of global warming advocates.
Est. std. Err of diff. Equals (root(56/5 plus 19.7/5))
Approx eq. root(11.2 plus 3.9)
Approx eq root(15.1)
Approx eq 4 or 3.something.
No reason to think normal underlying so probly not t
Estimated p-value of test of no diffs could be as large as 1/(sq(2.5))
Very Approx. Equal. .15.
Or worse cuz you only have smal sample estimates of variances and hence std. Errors.
So you can't reject possibility of no differences unless you have the ethics of global warming advocates.
Posted on 11/26/15 at 12:16 pm to LordSaintly
Hey tigerdik86 just had another obvious thought. You could add a hint of exp design to your project by pairing observations. If you can find n willing subjects all inexperienced measure the subject variable which I presune is deaths per session or per unit time. Measure again after some number of hours or sessions then you will have pairs of ob on each. Only one variable d equal late - early score for each participant Calculate dbar and std dev of d and do straight forward one or two tailed t test, n-1 degrees of freedom.
Stronger design cause it eliminates exogenous differences like IQ,coordination,age, etc between individuals in independent samples.
Tell your instructor you didn't want to do a normal plot cause it looked like a penis with a diseased nutsack.
Stronger design cause it eliminates exogenous differences like IQ,coordination,age, etc between individuals in independent samples.
Tell your instructor you didn't want to do a normal plot cause it looked like a penis with a diseased nutsack.
Page 2 of 2
Popular
Back to top
Follow TigerDroppings for LSU Football News