It's a little known fact that all of the 'Equal Variances Not Assumed' post-hoc tests in SPSS are simultaneously robust to unequal sample sizes. Games-Howell is perhaps the most popular post-hoc test in this case.
Yes, strictly speaking, if you specify alpha, and two ps are lower than alpha, then they are both statistically significant. However, people tend to use the expression "more statistically significant" in the sense of the p value is lower than another p value. A p value is a probability. It's not as if two p values lower than .05 lose any relative information. A p = .049 vs. a p = .00001 do not imply the same exact information, even though both are "statistically significant".
Hi, many thanks for your very practical useful videos. I love that in some of your videos you mentioned sources (citations) for the approaches that you are taking or the arguments you are making. Would you please do me a favor and give me a citation for what you are doing in this video dealing with unequal sample sizes? Many Thanks
Hello! I'm not so good at statistics and I have a question, which would help me if you could answer it! So in short, what is the difference of a post hoc test which has equal varraince and unequal varriances. I mean, what is the difference in the procedure of Bonferroni and Tamhane's T2. Hope you understand my question and can help, Thanks!
I enjoyed your video. However, in comparing p values you state something to the effect "the lower p value is more significant that the higher p value". From my understanding this is incorrect. Before conducting the experiment, the researcher decides on a level of significance (say 0.05). If p
one more question. when with the reduced sample we confirm the results of the full sample, Are all other parameter estimates of the full sample (like F value, R2 and Beta) reliable or reportable?
How unequal is unequal? E.g. If Group 1 n = 55, but group 2 n =50 and unequal variances, does this cause havoc with interpretation of t-test for corrected unequal variances? Would you still recommend doing the random selection procedure? many thanks!
Hi, so I have an unequal sample size, and I have to do a Mann-Whitney U test, is your method suitable for this test? I have 55 measurements of tree trunks in the primary forest and 32 tree trunk measurements in the secondary forest. Thanks a lot
hi i have a data observations. i want to separate those observations which are located 2 Standard deviation away from mean. How can i screen out those observations? please guide.
Any suggestions if your group with the lowest sample size only has 12?? Do any post hoc options correct for unequal variance and sample size?? Thanks a ton!!
can u plz help me what are the assumptions of unequal variances assuming tests such as tamhane's test, Dunnette C and Dunnette T3 tests, PLz help Thanku
It's a little known fact that all of the 'Equal Variances Not Assumed' post-hoc tests in SPSS are simultaneously robust to unequal sample sizes. Games-Howell is perhaps the most popular post-hoc test in this case.
Yes, strictly speaking, if you specify alpha, and two ps are lower than alpha, then they are both statistically significant. However, people tend to use the expression "more statistically significant" in the sense of the p value is lower than another p value. A p value is a probability. It's not as if two p values lower than .05 lose any relative information. A p = .049 vs. a p = .00001 do not imply the same exact information, even though both are "statistically significant".
You would do the same thing; randomly sample from each group to ensure equal sample sizes.
Hi, many thanks for your very practical useful videos. I love that in some of your videos you mentioned sources (citations) for the approaches that you are taking or the arguments you are making.
Would you please do me a favor and give me a citation for what you are doing in this video dealing with unequal sample sizes?
Many Thanks
Yalda, did you ever find a citation for this method? Thanks!
Thank you for your wonderful video. Do you know how SPSS calculates the corrected p value for unequal variances?
this is just for simple t test...how about a two way anova? how to deal with unequal variance in a two way anova?
Hello! I'm not so good at statistics and I have a question, which would help me if you could answer it!
So in short, what is the difference of a post hoc test which has equal varraince and unequal varriances. I mean, what is the difference in the procedure of Bonferroni and Tamhane's T2.
Hope you understand my question and can help, Thanks!
thank you soo much for this
I enjoyed your video. However, in comparing p values you state something to the effect "the lower p value is more significant that the higher p value". From my understanding this is incorrect. Before conducting the experiment, the researcher decides on a level of significance (say 0.05). If p
one more question. when with the reduced sample we confirm the results of the full sample, Are all other parameter estimates of the full sample (like F value, R2 and Beta) reliable or reportable?
How unequal is unequal? E.g. If Group 1 n = 55, but group 2 n =50 and unequal variances, does this cause havoc with interpretation of t-test for corrected unequal variances? Would you still recommend doing the random selection procedure? many thanks!
Hi, so I have an unequal sample size, and I have to do a Mann-Whitney U test, is your method suitable for this test? I have 55 measurements of tree trunks in the primary forest and 32 tree trunk measurements in the secondary forest. Thanks a lot
hi i have a data observations. i want to separate those observations which are located 2 Standard deviation away from mean. How can i screen out those observations? please guide.
I'd say the logic applies to regression, yes.
how to deal with unequal variances and unequal sample sizes in one way ANOVA. my sample from groups is: 198, 387, 168, 87, 27 and 4. plz guide
Can I deal unequal sample sizes with random sample before working with regression? Thank you very much!
Any suggestions if your group with the lowest sample size only has 12?? Do any post hoc options correct for unequal variance and sample size?? Thanks a ton!!
can u plz help me what are the assumptions of unequal variances assuming tests such as tamhane's test, Dunnette C and Dunnette T3 tests, PLz help Thanku
How might you report this?