Home › Forums › Old Forums › General › P-value in ANOVA table

This topic contains 12 replies, has 11 voices, and was last updated by Ernikunj 9 years, 2 months ago.

Viewing 13 posts - 1 through 13 (of 13 total)

- AuthorPosts
- September 15, 2003 at 4:44 am #46976
What actually the “P” value tells us in ANOVA table. We normally compare the test statistics with the critical value (F0, obtained from F-table at given d.o.fs) to determine whether the given parameters effects are significant or not.

ThanksSeptember 15, 2003 at 1:17 pm #46993What used to be done, before computers, was that you would look up a critical value and compare it to your observed F-statistic. That critical value represented the cutoff point for the alpha level of significance, usually 0.05. Well, 0.05 is also the p-value that is associated with a certain F-statistic. When you see a p-value in an ANOVA table what you see is really the area under the F distribution curve that is to the right of your observed F-statistic. It also represents the probability that your observed F-statistic occurs under the assumption that the null hypothesis of your test (in this case, that the means of the factor levels are equal).

September 15, 2003 at 5:38 pm #47010The p-value is the area to the right of the F statistic, F0, obtained from ANOVA table. It is the probability of observing a result (Fcritical) as big as the one which is obtained in the experiment (F0), assuming the null hypothesis is true. Low p-values are indications of strong evidence against the null hypothesis. I read in the book that it is common to declare a result as significant if the p-value is less than 0.05 or 0.01.

Hope this helpsSeptember 15, 2003 at 8:57 pm #47018This is more of a question than a statement, but I was always taught if you use a P-Value in the ANOVA then you are assuming normality in the data. Correct???

September 15, 2003 at 9:16 pm #47020In Analysis of variance test (ANOVA), we have two fundamental assumptions. First that the “means” are normally distributed; secondly the “variances” are equal. So, what I think is that in ANOVA test, we can use both methods, either “F” test or “P” test in oder to see whether the means are significantly different or not. Please correct me if i am wrong!

RegardsSeptember 22, 2003 at 1:53 pm #47345See this link. It offers one of the best explanations I’ve seen.

http://www.sportsci.org/resource/stats/pvalues.htmlSeptember 22, 2003 at 9:05 pm #47365P values have 2 specific uses.

1. Normality of Data – a significant P value (P < or = .05) means that the sample data you are testing (either with Normal Probability Plot or Descriptive Stats, using Minitab) is a normal subset of the population data.

2. Hypothesis Testing – a significant P value (P < or = .05) means you would reject your null hypothesis (that there is NO difference between the 2 or more sets of data you are testing).September 22, 2003 at 11:26 pm #47369In Bobby’s reply, paragraph1, is it that P < or = 0.05 implies normal data or

P > or = 0.05 implies normal data?

Thanks

September 23, 2003 at 12:31 am #47374I agree with 2nd statement, but a P Value in a normality test must be >=.05 to indicate normality.

September 23, 2003 at 2:04 am #47377The P value is the probability of Alpha (or type 1) error, which is the probability that an F value this large could occur by chance and not be due to some assignable cause.

September 24, 2003 at 10:56 am #47445Yes, I will say I placed my sign backwards. Just goes to show that in a few instances inspection may be a value-added step.

Always proofread your email before you hit send.

Thanks Dan and Raja for catching my miscue.September 24, 2003 at 6:36 pm #47465I remember it this way.If the p is low, the null must go.

October 17, 2008 at 1:30 am #151761In 2nd use…I would like to add that…p-Value is probability of making a type I error that is calculated from data.

so p-value is probability of rejecting a true hypothesis. - AuthorPosts

Viewing 13 posts - 1 through 13 (of 13 total)

The forum ‘General’ is closed to new topics and replies.

© Copyright iSixSigma 2000-2017. User Agreement. Any reproduction or other use of content without the express written consent of iSixSigma is prohibited. More »