Txoj cai ntawm tus ntiv tes xoo hais txog kev sib koom tes yog tias koj muaj ntau dhau thaum VIF siab dua 10 (qhov no yog vim peb muaj 10 tus ntiv tes, yog li coj cov cai ntawm tus ntiv tes xoo rau qhov lawv muaj nqis). Qhov cuam tshuam yuav yog tias koj muaj ntau ntau collinearity ntawm ob qhov sib txawv yog r≥. 95.
Yuav ua li cas yog kev sib koom siab siab?
Pairwise kev sib raug zoo ntawm cov kev hloov pauv ywj pheej yuav siab (hauv tus nqi tiag). Txoj cai ntawm tus ntiv tes xoo: Yog tias qhov sib txheeb > 0.8 ces hnyav ntau yam tuaj yeem tshwm sim. Ua tau rau tus kheej regression coefficients tsis tseem ceeb tab sis rau tag nrho haum ntawm qhov sib npaug kom siab.
Dab tsi yog kev sib raug zoo?
VIF qhov tseem ceeb yuav tsum yog tsawg dua 5 kom lav kev sib koom tes tsis yog qhov teeb meem hauv koj tus qauv. Txawm li cas los xij, qee cov kws tshawb fawb pom zoo kom nws yog < 3.3 thaum thov PLS-SEM. … Txais VIF tsawg dua 5 lossis 10 nyob ntawm seb muaj pes tsawg tus piav qhia muaj feem cuam tshuam.
Thaum twg kuv yuav tsum txhawj xeeb txog kev sib koom ua ke?
Multicollinearity yog ib qho teeb meem tshwm sim thaum kwv yees kab lossis kab qauv dav dav, suav nrog kev thauj mus los thiab Cox regression. Nws tshwm sim thaum muaj kev sib raug zoo ntawm cov kev kwv yees qhov sib txawv, ua rau tsis muaj kev ntseeg siab thiab tsis ruaj khov ntawm kev rov qab coefficients.
Dab tsi yog suav tias yog kev sib tw siab ntau?
High: Thaum kev sib raug zoo ntawm covexploratory variables high or there is perfect correlation among them, ces nws hais tias yuav siab multicollinearity.
