Excel Techniques - 09 - t-Test - Two Sample.avi

Excel Techniques for Statistics DGGB 6820 - Summer 2010

t-Test: Two Sample, regular two sample t-Test of two different samples.

(Please be aware that the audio is a little out of sync)

Download the Excel Workbook at:
https://docs.google.com/leaf?id=0B3MhRfj-HDlqNTA0NDhmYWMtN2I5My00ZTJkLWJiNjYt...

Download the Stats Notes at:
https://docs.google.com/viewer?a=v&pid=explorer&chrome=true&srcid...

Category:

Education

Tags:

• Excel Techniques
• DGGB 6820
• Fordham Stats
• Data Analysis
• t-test
• t test
• t-test two sample
• two sample t test

[dillingersteelflex]

oh my god, thank you so much

[337camo]

@GynxDon Open your file. Click on that circular colorful button on the top left corner. Go to Excel Options at the bottom. Click on Add-Ins and there is a thing that says something like TTest or something. Make it to where that add-in is active.

[GynxDon]

@337camo what did you do i cant figure it out

[337camo]

@337camo Nevermind that i figured it out.

[337camo]

My excel doesnt have Data Analysis. It has Get External Data, Connection, Sort & Filter, Data Tools, and Outline but no Data Analysis. I have excel 2010.

[seal123bee123]

I don't get the part of one tail would reject the null hypothesis, but two tail not.

If two tail tells you there is no different, how comes one tail would tell you one area is doing better?

[tdn1991]

@bbslugger8 Dont think you can. It's microsoft so probably solely for windows, though there might be something similar for mac.

[bbslugger8]

How can you get data analysis on mac?

[goldblade84]

FYI: regarding what RobertShay and I was talking about...

By the central limit theorem, means of samples from a population with finite variance approach a normal distribution regardless of the distribution of the population. Rules of thumb say that the sample means are basically normally distributed as long as the sample size is at least 20 or 30. For a t-test to be valid on a sample of smaller size, the population distribution would have to be approximately normal.

[goldblade84]

@RobertShay Oh, btw I gave what you said more thought and looked it up (it's been a while since I took stats), and I definitely think its applicable for non stats majors... given that if the sample size is large enough (central tendency theorem) a regular manager could definitely benefit from using this test to compare sets of data if there was no clear alternative at hand.

[goldblade84]

@RobertShay Not at all. Though this course isn't for stats majors, it's less about theory and more to show viewers how to use the techniques on Excel. I think it gives people who know little or nothing about stats theory a few tools to compare data to make a slightly better judgement call rather than basing the decision on chance or gut feeling etc.

[RobertShay]

I'm a little confused. Looking at the data it seems obvious that one data set doesn't follow a normal distribution - so why even perform a t-test. Second, it also seems obvious from the data that the two data sets don't have a the same variance so why use a t-test with similar variance. Am I missing something here?

[poloplayertj]

very helpful... thanks

[revok1169]

Thank you very much. This was perfectly applicable to my Conservation Ecology Statistics assignment. Keep posting the statistics videos, they are much appreciated.