|
t-Test
You will need to complete a t-test (more
specifically a two-tailed unpaired t-test*) on your data for your first two
laboratory reports. the following Internet Sites will either provide you with background
information about the t-test or will perform the test for you.
*For some of you a one-tailed test might be
more appropriate depending on your hypothesis. However, the two-tailed
test is more conservative and given the limited statistics background that
most of you have had, everyone can use the two-tailed test in this class
even if the data could be explored using a one-tailed test.
Basic Informational Site about What a t-test is and Does
This would also allow you to calculate the test by hand if
you wished. See Appendix 1 in the Laboratory manual as well.
http://www.socialresearchmethods.net/kb/stat_t.htm
- this is a good one
Online Programs that will run a T-test for You
Read the instructions for each. If one of the sites is
temporarily not operating use one of the others.
http://graphpad.com/quickcalcs/ttest1.cfm
1. Leave "Enter up to 50 rows" selected. (Select it if it's
not already selected.)
2. Put your raw data into the available boxes. (e.g., data for bindweed
slope condition 1 in
"Group One" spaces
and for bindweed slope condition 2 in "Group Two" spaces)
3. Select "Unpaired t test". (Select it if it's not
already selected.)
4. Left click on "Calculate now."
5. The p value is given at the top of the page and the means for each group are given below.
http://faculty.vassar.edu/~lowry/VassarStats.html
1. First left click on "t-Tests and Procedures" on the left navigation bar.
2. Left Click on "t-Test for Independent Samples."
3. In the small window that appears imput the number 5 (= the number of samples for each
plant in each area) and left click on "OK."
4. Scroll down and imput your raw data into the available small boxes. (e.g., data for
bindweed mowed in Xa spaces and for bindweed unmowed in Xb spaces)
5. Left click on "Calculate."
6. Scroll down the page. The means are reported in the "Data Summary" table.
Your P value in in the "P" table and is the "two-tailed" number.
7. You may print out the results by left clicking on "Printable Report" and then
left selecting the print command in your browser.
Chi-Square Goodness of Fit Test
and Hardy-Weinberg
You will need to complete Chi-square (more specifically Chi-square goodness of fit) tests on your
data for your fourth laboratory report. the following Internet Sites will either provide
you with background information about the Chi-square test or will perform the test for
you. Remember to use the NUMBERS OF INDIVIDUALS not genotype frequencies
when running a Chi-Square test on your population genetics data.
Basic Informational Sites about What a Chi-Square test is and Does
This would also allow you to calculate the test by hand if
you wished. A Chi-square is easier to calculate by hand than the t-test. (See
Appendix 2 in the Laboratory Manual for another explanation.)
http://grows.ups.edu/analysis/data_analysistext.htm
- explains both Hardy-Weinberg expected values and Chi-Square goodness of fit test
Online Programs that will run a Chi-square Test for You
For both of these sites you will NEED TO USE THE
CHI-SQUARE TABLE AT THE END OF APPENDIX 2 in the Lab Manual. Remember that
both the observed and expected data should be in the form of numbers of
individuals NOT genotype frequencies (See Appendix 2 in the
Laboratory Manual for a more detailed explanation.).
http://graphpad.com/quickcalcs/chisquared1.cfm
1. Leave "Enter up to 20 Categories (rows)" selected.
(Select it if it's not already selected.) You will be entering three
categories (NN, Nn, & nn).
2. Select "Actual". Select it if it's not already
selected.)
3. Enter your data.
- Label your three categories, NN, Nn, and nn.
- Enter the observed NUMBER OF INDIVIDUALS for each category in the
"Observed #" column .
- Enter the expected NUMBER OF INDIVIDUALS GIVEN THE TOTAL NUMBER IN
THE OBSERVED POPULATION (based on your Hardy-Weinberg calculations
using the observed allele frequencies) for each category in the
"Expected" column.
4. Left click on "Calculate now."
5. The Chi-square value is given at the top of the page and the
data you entered are summarized at the bottom.
The P VALUE GIVEN TO YOU ON THIS WEBSITE IS NOT THE ONE YOU
SHOULD USE. The computer program calculated using two degrees of freedom (df).
However, when looking at population genetic data in a one gene, two allele
system there is only one degree of freedom. Write down the Chi
value.
6. Use the Chi value and the 1 df (one degree of freedom) row of the Chi-square table in Appendix 1 of your lab
manual to estimate the p value..
http://www.kursus.kvl.dk/shares/vetgen/_Popgen/genetik/applets/kitest.htm
1. Put the number of individuals with the genotype NN observed in the top green box
for AA.
2. Put the number of individuals with the genotype Nn observed in the second green
box for AB.
3. Put the number of individuals with the genotype nn observed in the third green
box for BB.
4. Put the number of individuals predicted by the Hardy-Weinberg equation to have
the genotype NN in the top box in the "Expected" column for AA.
5. Put the number of individuals predicted by the Hardy-Weinberg equation to have
the genotype Nn in the second box in the "Expected" column for AB.
6. Put the number of individuals predicted by the Hardy-Weinberg equation to have
the genotype nn in the third box in the "Expected" column for BB.
7. Left click on the "Calculate!" button. (Do not click on
"Nullify!". It will erase your data.)
8. You will be given a Chi value (Chi=) that you should write down.
9. Use the Chi value and the 1 df (one degree of freedom) row of the Chi-square table in Appendix 1 of your lab
manual to estimate the p value.
Please send questions and comments about this site to the site administrator
07/07/2009
|
