T-test
Moderators: kgudger, bfinio, MadelineB, Moderators
-
- Former Expert
- Posts: 1415
- Joined: Sat Nov 08, 2003 11:27 pm
- Occupation: Planetary Scientist
- Project Question: N/A
- Project Due Date: N/A
- Project Status: Not applicable
Re: T-test
Hi mehtarb,
I'm not sure I understand your question. Can you elaborate on what you mean by "data points on each side"?
In addition, please explain how your question relates to a science fair project you are working on. We don't provide help with other types of HW assignments, but there are other online forums that do. If your question is not related to a K-12 science project, then a different set of online forums will better be able to help you with your question.
I'm not sure I understand your question. Can you elaborate on what you mean by "data points on each side"?
In addition, please explain how your question relates to a science fair project you are working on. We don't provide help with other types of HW assignments, but there are other online forums that do. If your question is not related to a K-12 science project, then a different set of online forums will better be able to help you with your question.
All the best,
Terik
Terik
Re: T-test
Hi, mehtarb.
You can absolutely run a t-test with unequal sample sizes. What is actually more important in a t-test is:
- Do your two populations have equal variances, as that changes the formulas a bit. You can conduct an F-test to determine this. See the Excel link below for more information.
- Are you samples independent?
- Do your samples follow a normal distribution (can also be tested for online or in Excel)
What are you trying to do with your two populations? Are you comparing means to see if they are different? Comparing two populations is also known as a two-sample t-test. More information here.
http://stattrek.com/hypothesis-test/dif ... utorial=AP
The formulas are easy enough to do manually, but I highly recommend letting software do it for you. There are free online calculators where you simply plug in your sample data, or it is also easy to do in Microsoft Excel. A tutuorial is below.
http://www.excel-easy.com/examples/t-test.html
Please feel free to write back with more questions. You can also find more information by Googling "two-sample t test".
Good luck!
You can absolutely run a t-test with unequal sample sizes. What is actually more important in a t-test is:
- Do your two populations have equal variances, as that changes the formulas a bit. You can conduct an F-test to determine this. See the Excel link below for more information.
- Are you samples independent?
- Do your samples follow a normal distribution (can also be tested for online or in Excel)
What are you trying to do with your two populations? Are you comparing means to see if they are different? Comparing two populations is also known as a two-sample t-test. More information here.
http://stattrek.com/hypothesis-test/dif ... utorial=AP
The formulas are easy enough to do manually, but I highly recommend letting software do it for you. There are free online calculators where you simply plug in your sample data, or it is also easy to do in Microsoft Excel. A tutuorial is below.
http://www.excel-easy.com/examples/t-test.html
Please feel free to write back with more questions. You can also find more information by Googling "two-sample t test".
Good luck!
Deana
-
- Moderator
- Posts: 908
- Joined: Fri Jun 20, 2014 4:42 pm
- Occupation: Biostatistician/Data Scientist
- Project Question: Interested in volunteering as an expert
- Project Due Date: n/a
- Project Status: Not applicable
Re: T-test
Hello mehtarb,
At the risk of overwhelming you with "expert" advice, I would second the first expert's suggestion that you tell us more about your science project. Before deciding to do a "t-test", I wonder why you have fewer observations in one group? And, what are the observations? What is your hypothesis with respect to the differences between the two groups?
Here's the link to the Science Buddies project guide which should help you answers to these questions.
Be sure to let us know if you have more questions!
At the risk of overwhelming you with "expert" advice, I would second the first expert's suggestion that you tell us more about your science project. Before deciding to do a "t-test", I wonder why you have fewer observations in one group? And, what are the observations? What is your hypothesis with respect to the differences between the two groups?
Here's the link to the Science Buddies project guide which should help you answers to these questions.
Be sure to let us know if you have more questions!
-
- Moderator
- Posts: 908
- Joined: Fri Jun 20, 2014 4:42 pm
- Occupation: Biostatistician/Data Scientist
- Project Question: Interested in volunteering as an expert
- Project Due Date: n/a
- Project Status: Not applicable
Re: T-test
Hello Mehtarb,
Here's the link that I forgot to add to my previous post (mea culpa)!
https://www.sciencebuddies.org/science- ... ndex.shtml
Here's the link that I forgot to add to my previous post (mea culpa)!
https://www.sciencebuddies.org/science- ... ndex.shtml