Scribd
Upload
120 views16 pages

Addendum To Ordered Logit and Probit Models by Afees Salisu

The document discusses the proportional odds assumption in ordered logit/probit models. It states that this assumption, that the relationship between each pair of outcome groups is the same, is often violated in practice. Several tests can check for violation, including Wolfe Gould, Brant, score, likelihood ratio, and Wald tests. If violated, options include using a different model that does not assume proportionality, such as generalized ordered logit or heterogeneous choice models.

Uploaded by

Neema
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
120 views16 pages

Addendum To Ordered Logit and Probit Models by Afees Salisu

The document discusses the proportional odds assumption in ordered logit/probit models. It states that this assumption, that the relationship between each pair of outcome groups is the same, is often violated in practice. Several tests can check for violation, including Wolfe Gould, Brant, score, likelihood ratio, and Wald tests. If violated, options include using a different model that does not assume proportionality, such as generalized ordered logit or heterogeneous choice models.

Uploaded by

Neema
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 16

The Proportional Odds Assumption

in Ordered Logit/Probit Models

Afees Salisu

9/7/2016 CBN-ITI TRAINING (2016) 1


The proportional odds assumption

• One of the assumptions underlying


ordered logistic (and ordered probit)
regression is that the relationship between
each pair of outcome groups is the same.

9/7/2016 CBN-ITI TRAINING (2016) 2


Odds assumption …
• In other words, ordered logistic
regression assumes that the coefficients
that describe the relationship between,
say, the lowest versus all higher categories
of the response variable are the same as
those that describe the relationship
between the next lowest category and all
higher categories, etc.

9/7/2016 CBN-ITI TRAINING (2016) 3


Odds assumption …
• This is called the proportional odds
assumption or the parallel regression
assumption.

• In practice, violating this assumption may


or may not alter your substantive
conclusions. You need to test whether this
is the case.

9/7/2016 CBN-ITI TRAINING (2016) 4


Odds assumption …
• There are several tests for verifying this
assumption.
• Among these tests are Wolfe Gould, Brant,
Score, Likelihood ratio and Wald tests.
• The underlying null hypothesis is that the
relationship is proportional; that is,
parallel.

9/7/2016 CBN-ITI TRAINING (2016) 5


Odds assumption …
• To test in Stata, use ‘oparallel’ command.
• It is a user written program.
• To download the command type “findit
oparallel” in Stata.
• Once downloaded, you can type
“oparallel” immediately after you
estimate an ordered logit model (“ologit”)
to perform the test.

9/7/2016 CBN-ITI TRAINING (2016) 6


Odds assumption …
• In the case of our example, the oparallel
test is shown below
. oparallel

Tests of the parallel regression assumption

Chi2 df P>Chi2

Wolfe Gould 21.33 3 0.000


Brant 17.45 3 0.001
score 26.66 3 0.000
likelihood ratio 22.05 3 0.000
Wald 28.88 3 0.000

9/7/2016 CBN-ITI TRAINING (2016) 7


Odds assumption …
• Interpretation:
• The relationship is not proportional across
all the test statistics.

9/7/2016 CBN-ITI TRAINING (2016) 8


Dealing with violation …
• Option 1: Do nothing. Use ordered logistic
regression because the practical implications
of violating this assumption are minimal.
• Option 2: Use a multinomial logit model.
This frees you of the proportionality
assumption, but it is less parsimonious and
often dubious on substantive grounds.
• Option 3: Dichotomize the outcome and use
binary logistic regression. This is common,
but you lose information and it could alter
your substantive conclusions. 9
Dealing with violation …
• For instance, in the case of our example, one
can merge categories 1 & 2 since their marginal
effects are similar both in terms of sign and
significance.
• Stata command on how to dichotomize:
recode healthstatus (1 2 = 0)(3 = 1),
gen(health)
• Stata command for estimation:
logit health age logincome numberdiseases
9/7/2016 CBN-ITI TRAINING (2016) 10
Dealing with violation …

• Option 4: Use a model that does not


assume proportionality. Increasingly, this
is common. Two user-submitted Stata
commands fit these kinds of models:
• “gologit2” – generalized ordered logit
models (see Williams 2007, Stata Jn.).
• “oglm” – heterogeneous choice models
(see Williams 2010, Stata Jn.)

9/7/2016 CBN-ITI TRAINING (2016) 11


Dealing with violation …
• Note that they are user-written programs. You
have to install them before estimation.
• Stata estimation commands after installation:
oglm healthstatus age logincome
numberdiseases

gologit2 healthstatus age logincome


numberdiseases
• Note that the results obtained here are log odds
9/7/2016 CBN-ITI TRAINING (2016) 12
Dealing with violation …
• For the odds ratio, use:
oglm healthstatus age logincome
numberdiseases, or

gologit2 healthstatus1 age logincome


numberdiseases, or

9/7/2016 CBN-ITI TRAINING (2016) 13


Dealing with violation …
• For the marginal effects, use the Stata
commands below immediately after
estimating with oglm / gologit2;
margins, dydx(*) predict(outcome(1))
atmeans
margins, dydx(*) predict(outcome(2))
atmeans
margins, dydx(*) predict(outcome(3))
atmeans

9/7/2016 CBN-ITI TRAINING (2016) 14


Dealing with violation …
• Like the ologit command, you can also
perform relevant scenario analyses for the
oglm and gologit2 commands.
• For instance, you may be interested in
evaluating how certain fixed values of age
will affect the probability of reporting
excellent health status

9/7/2016 CBN-ITI TRAINING (2016) 15


Dealing with violation …
• . Stata commands:
margins, dydx(*) at(age=(25 30 35))
predict(outcome(3)) atmeans

margins, dydx(*) at(age=25/30 )


predict(outcome(3)) atmeans
• Please update your do-file appropriately.

9/7/2016 CBN-ITI TRAINING (2016) 16

You might also like