Timeline for Is it meaningful to test equality of two coefficients, one significant, the other insignificant?
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 13, 2017 at 12:44 | history | edited | CommunityBot |
replaced http://stats.stackexchange.com/ with https://stats.stackexchange.com/
|
|
| Mar 31, 2012 at 17:02 | comment | added | Andy W | (+1) Also to note, there is a paper by Andrew Gelman and Hal Stern that discusses this issue as well, The difference between “significant” and “not significant” is not itself statistically significant. | |
| Mar 31, 2012 at 9:59 | comment | added | Yang | Macro. I don't have enough reputation to upvote our answer. So I only choose accept. Thank you! and hope you to give me more help.@Macro | |
| Mar 31, 2012 at 9:57 | vote | accept | Yang | ||
| Mar 31, 2012 at 9:50 | comment | added | Yang | Actually the reason to test $b_{1}=b_{2}=b_{3}$ is to combine groups. If the joint test is significantly rejected, I will separate these three categories i.e. regression including three interaction terms. But if the joint test can not be rejected, I will put them together i.e. there is not interaction term $x_{i}$ but only using independent variable $x$. You point out the difference between pairwise test and ANOVA. That is correct. But I don't know whether it is correct to put three categories together when $b_{1}=b_{2}=b_{3}$ can not be rejected but $b_{1}=b_{3}$ is rejected. | |
| Mar 31, 2012 at 9:22 | history | answered | Macro | CC BY-SA 3.0 |