I am currently interperting the results from an ordered logit regression, the following table shows these results with odds ratios. SCR is an ordinal variable that ranges from 1 to 22
Ordered logit, odds ratios clustered standard errors Bank claims and Bank liabilities | ||||||||
Bank claims | Bank liabilities | |||||||
(1) | (2) | (3) | (4) | (1) | (2) | (3) | (4) | |
VARIABLES | All countries | HIG | LIG | NIG | All countries | HIG | LIG | NIG |
SCR | . | . | . | . | . | . | . | . |
(.) | (.) | (.) | (.) | (.) | (.) | (.) | (.) | |
LogBclaims | 1.9068*** | 2.2969** | 1.6172** | 1.4832*** | ||||
(0.1715) | (0.7678) | (0.3621) | (0.1926) | |||||
LobBlib | 1.8183*** | 1.5799* | 1.2364 | 1.5330*** | ||||
(0.1993) | (0.3727) | (0.2955) | (0.1944) | |||||
LogGDPcapita | 4.7958*** | 5.6374* | 6.5603*** | 2.3693*** | 5.3839*** | 7.8473* | 6.0378*** | 2.6592*** |
(0.8085) | (5.8456) | (3.0772) | (0.5951) | (0.9556) | (8.3403) | (3.0996) | (0.6583) | |
VIXrate | 1.0626*** | 1.0730* | 1.0433* | 1.0424*** | 1.0638*** | 1.0743* | 1.0360* | 1.0450*** |
(0.0097) | (0.0424) | (0.0242) | (0.0147) | (0.0100) | (0.0427) | (0.0220) | (0.0149) | |
TEDspread | 0.9979 | 0.9994 | 0.9912** | 0.9974 | 0.9974* | 0.9988 | 0.9948 | 0.9963 |
(0.0013) | (0.0041) | (0.0044) | (0.0022) | (0.0014) | (0.0040) | (0.0043) | (0.0024) | |
FDIpGDP | 0.9972 | 0.9783 | 1.0011 | 0.9894*** | 0.9986 | 0.9834 | 1.0017 | 0.9872*** |
(0.0071) | (0.0250) | (0.0012) | (0.0038) | (0.0093) | (0.0240) | (0.0014) | (0.0034) | |
GDPchangeP | 1.0773** | 1.1613 | 1.0464 | 1.0529 | 1.0597* | 1.1447 | 1.0353 | 1.0375 |
(0.0333) | (0.1315) | (0.0591) | (0.0387) | (0.0357) | (0.1328) | (0.0609) | (0.0385) | |
DebtGDP | 0.9778*** | 0.9843 | 0.9902 | 0.9723*** | 0.9774*** | 0.9897 | 0.9947 | 0.9696*** |
(0.0055) | (0.0131) | (0.0063) | (0.0043) | (0.0063) | (0.0143) | (0.0059) | (0.0048) | |
infPchange | 0.8951*** | 0.8744 | 1.0217 | 0.9114*** | 0.8977*** | 0.9216 | 0.9845 | 0.9122*** |
(0.0148) | (0.0984) | (0.0603) | (0.0119) | (0.0150) | (0.1093) | (0.0684) | (0.0121) | |
Observations | 1,688 | 410 | 448 | 830 | 1,694 | 410 | 448 | 836 |
Number of groups | 86 | 28 | 38 | 60 | 86 | 28 | 38 | 60 |
Am i correct to say that if logBclaims goes up by 1, with everything else held constant, the odds of going up 1 notch in SCR is 90% for all the (1) regression all countries (since the value is 1.9068)
In addition for an odds ratio below 1 like infPchange, could i say: With everything else held constant if InfPchange goes up by 1 the odds of going up 1 notch decrease with 11%
And for logGDPcapita since it is 4,79 how do interpert this? is it if logGDPcapita goes up by 1 the odds of SCR increasing with 1 is 379%? so it increase by 3.79 notches? i am still a little confused on how to interpert it if it is more than 1.9999.
Thank you for taking the time to read this!
Kind regards,
Max Verheijen
0 Response to interpreting ordered logit regression results odds ratio
Post a Comment