Hypothesis testing
Six Sigma – iSixSigma › Forums › Old Forums › General › Hypothesis testing
 This topic has 7 replies, 5 voices, and was last updated 17 years, 2 months ago by Darth.

AuthorPosts

November 18, 2004 at 9:27 am #37584
When do we use Ftest and when should we use Ttest?
0November 18, 2004 at 12:06 pm #110915Chelle,
You should use the F test when you have variable data and want to compare the variance (or associated standard deviation) between sample sets….
You should use the t test when you have variable data and want to compare the means of two sample sets…
Practically, the F test is applicable when you are trying to reduce variation and the t test is applicable when you are working to shift (or stabilize) the mean.
Hope this helps…
Best Regards,
Bob J0November 19, 2004 at 2:06 am #110955But isn’t it right that we do ftest first before we do ttest?
0November 19, 2004 at 6:38 am #110962
Ron ManubayMember@RonManubay Include @RonManubay in your post and this person will
be notified via email.Since you have different ttest formula for equal and unequal variance, I would say that YES, it would be good to do ftest first.
0November 19, 2004 at 11:36 am #110982Chelle,
That is correct if you are working with means (t test) and you want to determine whether or not you want to assume equal variances when running the test.
Otherwise, you can use the F test as a stand alone to assess two populations to see if there is a significant difference in the respective variances.
Hope this helps….
Best Regards,
Bob J0November 19, 2004 at 11:55 am #110984
Mike CarnellParticipant@MikeCarnell Include @MikeCarnell in your post and this person will
be notified via email.Chelle,
You need to start with a normality test. If you have non normal data you need to test medians not means (you can also transform the data).
Good luck.0November 22, 2004 at 12:32 am #111083Hello Mike,How do you test medians?Thanks.
0November 22, 2004 at 1:08 am #111086This should keep you busy for a while.
Mood’s median test can be used to test the equality of medians from two or more populations and, like the KruskalWallis Test, provides a nonparametric alternative to the oneway analysis of variance. Mood’s median test is sometimes called a median test or sign scores test. Mood’s median test tests:
Ho: the population medians are all equal versus Ha: the medians are not all equal
An assumption of Mood’s median test is that the data from each population are independent random samples and the population distributions have the same shape. Mood’s median test is robust against outliers and errors in data and is particularly appropriate in the preliminary stages of analysis. Mood’s median test is more robust than is the KruskalWallis test against outliers, but is less powerful for data from many distributions, including the normal.
You can perform a KruskalWallis test of the equality of medians for two or more populations.This test is a generalization of the procedure used by the MannWhitney test and, like Mood’s Median test, offers a nonparametric alternative to the oneway analysis of variance. The KruskalWallis hypotheses are:
Ho: the population medians are all equal versus Ha: the medians are not all equal
An assumption for this test is that the samples from the different populations are independent random samples from continuous distributions, with the distributions having the same shape. The KruskalWallis test is more powerful than Mood’s median test for data from many distributions, including data from the normal distribution, but is less robust against outliers.
You can perform a twosample rank test (also called the MannWhitney test, or the twosample Wilcoxon rank sum test) of the equality of two population medians, and calculate the corresponding point estimate and confidence interval. The hypotheses are
Ho: h1 = h2 versus Ha: h1 ¹ h2 , where h is the population median.
An assumption for the MannWhitney test is that the data are independent random samples from two populations that have the same shape (hence the same variance) and a scale that is continuous or ordinal (possesses natural ordering) if discrete. The twosample rank test is slightly less powerful (the confidence interval is wider on the average) than the twosample test with pooled sample variance when the populations are normal, and considerably more powerful (confidence interval is narrower, on the average) for many other populations. If the populations have different shapes or different standard deviations, a 2Sample t without pooling variances may be more appropriate.
0 
AuthorPosts
The forum ‘General’ is closed to new topics and replies.