Journal of Choice Modelling, 1(1), 2008, 26-39

www.jocm.org.uk

Specification testing of discrete choice

models: a note on the use of a nonparametric
test
Mogens Fosgerau∗
Institute for Transport, Technical University of Denmark, Knuth-Winterfeldts Allé,
Bygning 116 Vest, 2800 Kgs. Lyngby, Denmark

Received 21 August 2007, revised version received 4 December 2007, accepted 7 January 2008

Abstract
Model misspecification is a serious issue since misspecification generally
renders statistical inference invalid. However, specification testing of dis-
crete choice models is rarely applied. This paper describes a nonparametric
test procedure which uses a combination of smoothed residual plots and a
test statistic able to detect general misspecification. Nonparametric meth-
ods require large datasets when the number of independent variables is more
than a few. A way to circumvent this problem is indicated, increasing the
usefulness of the approach also with limited datasets.

Keywords: discrete choice, specification test, nonparametric, functional form

1 Introduction
It is standard practice in regression models to perform model control using the
residuals of the estimated model. Residuals are plotted to verify whether they
are in fact white noise unrelated to the independent variables. Residuals are
less easily defined in discrete choice models and similar model control is rarely
performed for such models.
In fact, a variety of specification tests are available in the literature for some
discrete choice models, but are not widely used. Lechner (1991) presents some
specification tests for the binary logit model. Gourieroux et al. (1987a) and
Gourieroux et al. (1987b) present tests based on generalised residuals for a range
of models including the multinomial logit model. McFadden (1987) presents
regression-based specification tests for the multinomial logit model similar in
nature to the test presented here. The seminal paper by McFadden and Train
∗
T: +45 4525 6521, F: +45 4593 6533, [email protected]

Licensed under a Creative Commons Attribution-Non-Commercial 2.0 UK: England & Wales License
http://creativecommons.org/licenses/by-nc/2.0/uk/

ISSN 1755-5345
 Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

(2000) provides specification tests of MNL and mixed logit against alternatives
with more mixing. Finally, software exists that allows the comparison of observed
choices to predictions when data are grouped on a categorical variable. This
approach may be viewed as a kind of residual test.1
The point of this paper is to describe how the nonparametric test of func-
tional form in Zheng (1996) may be applied to discrete choice models of general
form. This means that the test applies even to complicated models such as the
mixed generalised extreme value model. The Zheng test is based on nonparamet-
ric kernel regression of the parametric model residuals against the independent
variables. With residuals defined as the difference between choice 0-1 indicators
and predicted probabilities, the Zheng test applied to a discrete choice model is
based on the comparison of predicted and observed choices. Pagan and Ullah
(1999) review a range of nonparametric tests of functional form that are more or
less similar to the Zheng test. The procedure presented in this paper of applying
a test to a function of the independent variables is not restricted to the Zheng
test but may be applied with other tests as well. On the use of nonparametrics in
a discrete choice context and in transport applications see, e.g. Fosgerau (2006),
Fosgerau (2007) and Fosgerau and Bierlaire (2007). Pagan and Ullah (1999),
Yatchew (2003) and Härdle (1990) give general introductions to nonparametric
and semiparametric methods.2
A general problem in using nonparametric techniques is that the demands
on data increase exponentially in the dimension of the space of independent
variables, this is the so-called curse of dimensionality. This issue is addressed
in this paper by showing how such tests may be applied to a subset of variables
or more generally to functions of variables such as the index representing the
indirect utility of a choice alternative. It is thus possible to apply nonparametric
tests to a model, while regressing only on a low number of variables or even just
one. This reduces the demand on data and is of practical importance when the
model to be tested has several independent variables.
Furthermore, the test may be applied to variables that are not included in
the model. Thus the test may serve as a test of omitted variables.
The paper is organised as follows. Section 2 presents an exposition of the
Zheng test. Section 3 shows how such tests may be applied to low-dimensional
functions of the independent variables. Section 4 presents an example of appli-
cation of the test to a discrete choice model while section 5 concludes.

2 The Zheng test

Consider a model E(y|x), predicting the expectation of a variable y ∈ R condi-
tional on an m-dimensional vector of observables x ∈ Rm , which are also taken
1
E.g. Alogit produces so-called apply tables.
2
Tests based on regression of residuals against independent variables are also discussed in
Ellison and Ellison (2000).

27
 Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

as random having density p(x). The dependent variable y may be continuous,

but it may also be binary or a binary dummy indicator of an alternative in a
multinomial model. In such cases E(y|x) = P (y = 1|x). In models with many
alternatives it may also be useful to let y be a dummy indicator for the choice of
a group of alternatives. This situation is also accommodated. In situations with
unlabelled alternatives it could be useful to reorder the alternatives according to
some characteristic such that y indicates, e.g. choice of the fastest alternative.
The expectation of y conditional on x is a function of x. We let g(x) denote the
true but unknown conditional expectation. The researcher specifies a parametric
model f (x : θ) where θ ∈ Θ. We wish to test whether there is a parameter θ0 ∈ Θ
such that f (x : θ0 ) = g(x). Formally, we formulate our null hypothesis as

H0 : P [f (x : θ0 ) = g(x)] = 1. (1)

The null hypothesis thus states that f (· : θ0 ) coincides with g for almost all x.
The corresponding alternative hypothesis is that, for all θ ∈ Θ, f differs from g
on a set of probability greater than zero, or formally,

H1 : ∀θ ∈ Θ P [f (x : θ) = g(x)] < 1. (2)

The alternative hypothesis is the negation of the null hypothesis. It thus encom-
passes all possible departures from the null.
Zheng (1996) defines a statistic Tn , where n is sample size, and shows that,
under the null hypothesis, Tn converges in distribution to a standard normal
as sample size increases, whereas it converges in probability to infinity under
the alternative. Therefore the test is consistent against all departures from the
parametric model f .
The idea of the test is the following. Define residuals by ε = y − f (x :
θ0 ). Then under the null hypothesis, E(ε|x) = 0 almost surely and hence also
E[εE(ε|x)p(x)] = 0. However, under the alternative hypothesis,
E[εE(ε|x)p(x)] = E[E(ε|x)2 p(x)]
= E[(g(x) − f (x : θ0 ))2 p(x)]
> 0. (3)
where the first equality follows from the law of iterated expectations.
Consider now an i.i.d. sample (yi , xi ). The test statistic is formed from a
sample analogue of E[εE(ε|x)p(x)] = 0, constructed using kernel regression and
kernel density estimation (see for example Pagan and Ullah, 1999). In order to
apply these methods
R we need a non-negative, bounded, continuous and symmetric
kernel K with K(u)du = 1 and a bandwidth h depending on the sample size
n. The application in section 4 uses a standard normal density for K. Then we
construct first an estimate of the density of x ∈ Rm as
 
1 X 1 xi − xj
p̂(xi ) = K (4)
n−1 hm h
j≤n,j6=i

28
 Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

and next an estimate of the expected residual as

 
1 P 1 xi −xj
n−1 j≤n,j6=i hm K h εj
Ê(εi |xi ) = . (5)
p̂(xi )

Letting θ̂ be consistently estimated, define ei = yi − f (xi : θ̂) and define next a

sample analogue of E[εE(ε|x)p(x)] by
 
1 X X 1 xi − xj
K ei ej (6)
n(n − 1) hm h
i≤n j≤n,j6=i

Zheng then defines a standardised version of the test statistic by

 
P P xi −xj
i≤n j≤n,j6=i K h ei ej
Tn = h   i1 (7)
2 xi −xj e2 e2 2
P P
i≤n j≤n,j6=i 2K h i j

and shows that under some regularity conditions, if h → 0 and nhm → ∞,

then under the the null hypothesis, Tn →d N (0, 1), while under the alternative
hypothesis,
Tn
→p c > 0. (8)
nhm/2
Since Tn is a one-sided test statistic, we reject at level α if Tn > Zα , where Zα is
the upper α-percentile of a standard normal variable. For example, we reject H0
at the 5 percent level if Tn > 1.645 (Li and Racine, 2007).

2.1 Some comments

Although the Zheng statistic looks complicated, there is a straight-forward intu-
ition behind. The denominator in the expression for Tn takes care of the stan-
dardisation so we concentrate on the numerator. This is just a weighted sum of
products of residuals.
The statistic is built up as a sum of contributions from each observation i.
The weighting ensures that only observations near the i’th receive non-negligible
weight. The weighted product ei ej is positive if residuals ej near ei have the same
sign as ei . When the true and the estimated models are smooth functions of x,
this will happen if

0 6= g(xi ) − f (xi : θ̂) = E(yi − f (xi : θ̂)|xi ) = E(e|xi ). (9)

The weighting ensures that local discrepancies between g and f are detected even
when positive and negative discrepancies would net out globally.
The estimator in eq. (5) is a nonparametric kernel regression of the residuals
ε against x. It is useful to use this regression to produce smoothed plots of

29
 Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

the residuals against the independent variables. Confidence bands around the
regression can be computed using that
 Z 
(nh)1/2 [Ê(εi |xi ) − E(εi |xi )] ∼ N 0, σ 2 (xi )p−1 (xi ) K 2 (u)du , (10)

where σ 2 is the variance of ε conditional on x.3 For the purpose of testing the
specification in the parametric model f we may estimate this variance using the
null hypothesis that E(y|x) = f (x|θ) a.e. In the case of a uni-dimensional x, we
may estimate σ(x) by

fˆ(x)(1 − fˆ(x))
Z
2
σ̂ (x) = K 2 (u)du, (11)
p̂(x)nh

where fˆ is a kernel regression of the parametric model probabilities against x and

p̂ is the estimated density of x. For a normal density kernel, K 2 (u)du = 2√1 π .
R

The idea that the bandwidth h should depend on sample size may be unfamil-
iar. If the bandwidth is too large then averaging over too large neighbourhoods
will smooth away relevant differences and hence introduce bias. If the bandwidth
is too small then estimates will be noisy, they will be too much influenced by
random fluctuations in data and hence variance will be high. Choosing the band-
width appropriately as a function of sample size balances bias and variance and
ensures that the bandwidth tends to zero at an appropriately slow rate, such that
the asymptotical results obtain.
These considerations suggest that using a large bandwidth will tend to reduce
the Zheng statistic as positive differences between f and g in some regions of the
data will then be averaged with negative differences in other regions.
For a given sample size n, the bandwidth may be chosen by a rule such as, e.g.
1
h = n− 2m . This choice is easy and agrees with the conditions for the Zheng test.
It is also possible to select a bandwidth using cross-validation in the regression of
the residuals against x. This is however time consuming and the gain from doing
it is not clear.4 In applied research, it may be sufficient or even preferable to
select a bandwidth by just inspecting the resulting regression and the confidence
bands visually, this is so-called eye-balling, suggested by Pagan and Ullah when
the dimension of x is low enough to make this feasible.

3 Reducing dimensionality
A concern with the application of tests like the Zheng test, just as with any
nonparametric technique, is the curse of dimensionality. The size of the dataset
3
See Pagan and Ullah (1999).
4
See the discussion in Li and Racine (2007). Generally all that is required for the test to be
consistent is that bandwidths tend to zero as sample size tends to infinity but slowly enough
that n multiplied by the product of bandwidths tends to infinity. This is, however, not very
helpful in finite samples.

30
 Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

necessary to achieve a given degree of precision increases exponentially in the

dimension of x. Rather than use the test directly it is therefore useful instead to
consider the test defined over a function of the data.
Let t(·) : Rm → Rmt be a measurable function such that t(x) has a density.
Note that mt ≤ m, such that t is not in general injective. The inverse function
of t is then a set-valued function defined by t−1 (t0 ) = {x : t(x) = t0 }. The idea
is now to work with the mt -dimensional stochastic variable t rather than with
the m-dimensional stochastic variable x. So we apply the Zheng test to E(y|t)
rather than to E(y|x). When mt < m there is a reduction of dimensionality and
the test will be easier to apply. Note that
E(y|t0 ) = E(y|x ∈ t−1 (t0 )) = E(E(y|x)|x ∈ t−1 (t0 )) (12)
such that g(t) = E(y|t) and f (t : θ) = E(f (x : θ)|t) are the corresponding true
and modelled conditional means, now defined over t rather than over x.
It is straight-forward that our null hypothesis H0 implies the version refor-
mulated in terms of functions t. If the model is true, then the model is also true
as a function of any function t.
H0 ⇒ H0∗ : ∃θ0 ∀t P [f (t : θ0 ) = g(t)] = 1. (13)
It is also clear that the converse hypothesis H1 implies the converse of H0∗ . That
is, if the model is false, then there exists a function t such that the model is false
as a function of t.
H1 ⇒ H1∗ : ∀θ ∈ Θ ∃t P [f (t : θ) = g(t)] < 1. (14)
To verify the latter, simply choose a function t as t(x) = g(x) − f (x : θ). The
reverse implications also hold: H0∗ implies the converse of H1∗ , which implies the
converse of H1 , which in turn implies H0 . So H0 is equivalent to H0∗ and H1 is
equivalent to H1∗ .

3.1 Some comments

It is now possible to proceed by selecting various functions t of the data and to
conduct the Zheng test (or another similar test) on the model conditional on these
t’s. We cannot be sure to reject a false model for any given t, as the discrepancy
between the parametric model and the true model may not lie in the direction
of t. But if we reject the model for any t, then we are under the alternative
hypothesis H1∗ and hence also under H1 , which means that the model has been
rejected. Otherwise the model is accepted. The test based on functions t will be
less powerful since we may not be able to choose appropriate t’s. On the other
hand we gain the advantage that the dimension of t can be chosen to be low,
which greatly facilitates the use of nonparametric kernel methods.
Note that it is not necessary that every variable in x actually appears in the
model f . If a variable xo is omitted from the parametric model we may define
t(x) = x0 to form a nonparametric omitted variable test.

31
 Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

In a discrete choice model where the systematic utility of an alternative is

defined by βx with β fixed, we may test the model by defining t(x) = βx and
take y as the binary indicator for choice of the same or another alternative.
When the dimension of t is 1 or 2 it is feasible to select the bandwidth vi-
sually by plotting the nonparametric regression of the residuals against t(x) and
varying the bandwidth until an appropriate number of features arise. It may be
informative to inspect the data visually in this way and the Zheng test may then
be used to check if detected deviations of the expected residual from zero are in
fact significant.
It is hard to very specific about how this should be done. But usually one has
some sense of how complicated a potential relationship might be. A low band-
width tends to lead to an estimated relationship with many peaks. If there are
more peaks than one thinks is likely in the true relationship, then it is appropriate
to increase the bandwidth. On the other hand, choosing a too large bandwidth
will smooth away many differences. This may introduce bias so it might be better
to use a somewhat smaller bandwidth.

4 Examples
4.1 Mode choice
This section provides an illustrative application of the test to a mode choice
model. The model is not intended as a serious model. It merely serves the pur-
pose of showing how the Zheng test may be used to detect significant differences
between model and data. I use a mode choice data set from Sweden comprising
1799 observations of trips to Stockholm using either air, bus, car or train. For
each observation the data give travel time and cost for each alternative as well
as some background variables.
I estimated an MNL with alternative specific constants, mode-specific pa-
rameters for travel time and a common cost parameter. Write this model with
utilities Ui = Vi + εi , where εi are iid. extreme value and Vi = βxi . Then I com-
puted the predicted mode choice probabilities for the observations in the sample.
Residuals were computed for each mode by subtracting the predicted probability
from a dummy indicator for the observed choice. All the tests are carried out
using uni-dimensional kernel regressions and density estimates were carried out
using a normal density kernel and a bandwidth of n−1/2 , where n is sample size
and data (t(x)) are scaled to the unit interval. The choice of kernel is generally
not important. Judging from the graphs in the following, the bandwidth seems
appropriate. For computing confidence intervals around the kernel regression I
used equation (10).5 I trimmed away 1 percent of the sample, excluding the up-
5
The MNL was estimated in Biogeme (Bierlaire, 2003, 2005), which was also used to compute
the predicted choice probabilities. The Zheng test and the nonparametric regressions were
performed in Ox (Doornik, 2001). The Ox code for the test is given in the appendix.

32
Density_realincome 09:51:26 17-Jan-2008

Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 6.25 6.50

Figure 1: Density of log of real income

per and lower 0.5 percent after sorting on the independent variable, in order to
exclude regions with very thin data and hence uncertain estimates.

4.1.1 Omitted variable test

An income variable is available in the data but has not been included in the model.
Suppose we want to test whether income could improve the prediction of choices.
In this simple model one could of course just try estimating some specifications
including an income variable. For more complicated models it could however
be prohibitively time-consuming to estimate many different versions of the same
model. In such cases an omitted variable test could be useful.
Figure 1 shows first a kernel density estimate of the log of real income. The
Zheng statistics for the four alternatives are 3.3 (Air), 0.7 (Bus), 1.1 (Car) and
3.7 (Train), which means that the residuals depend systematically on income for
the Air and Train alternatives. Figures 2 and 3 show the corresponding plots for
the air and bus modes. It is clear that the residuals of choosing air increases with
income. Thus there is a dependency of choices on income that is not captured
by the model. A similarly clear relationship is not visible for the bus residuals.
The dependency on income is, of course, unsurprising. The point here is
that the test is well able to detect it. Income may be included in the model
in various ways and the test does not indicate how it should be done. The
smoothed residual plots suggest a specification that is linear in log income. It is

33
P_Airrealincome 09:52:01 17-Jan-2008

Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

0.5

0.4

0.3

0.2

0.1

0.0

-0.1

-0.2
P_Busrealincome 09:52:22 17-Jan-2008

-0.3

-0.4

2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 6.25 6.50

Figure 2: Residuals of air alternative on real income

0.5

0.4

0.3

0.2

0.1

0.0

-0.1

-0.2

-0.3

-0.4

2.75 3.00 3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25 5.50 5.75 6.00 6.25 6.50

Figure 3: Residuals of bus alternative on real income

34
P_TraV_Bus 10:57:58 26-Nov-2007

Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

0.5

0.4

0.3

0.2

0.1

0.0

-0.1

-0.2

-0.3

-0.4

-3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

Figure 4: Residuals of train alternative on systematic utility for bus

not possible to conclude from the test which alternatives income should enter as
an income variable in one alternative would affect the predicted probabilities of
all alternatives. I tested a model formulation with separate income parameters
for air, car and train and found those for air and car to be confidently positive,
while that for train to be insignificant.

4.1.2 Misspecification test

It depends on the application whether exclusion of a significant omitted vari-
able can be considered a misspecification. But a significant value of the Zheng
statistic for variables that are included in the model must be considered evidence
of misspecification. Table 1 shows the Zheng statistic for residuals of the four
modes on the systematic utilities for the four modes. The test does not find sig-
nificant misspecification over the systematic utility of the air mode, while there
are systematic differences along the directions of the systematic utilities of the
remaining modes. Figure 4 shows residual plots for the Train residuals on the
systematic utility for Bus.
Judging from the estimated parameters and the loglikelihood it was impossible
to detect whether the model is misspecified. But the test clearly shows that the
model is indeed misspecified. The test is general against misspecification and so
does not indicate the cause of misspecification. In a serious application it would
therefore be pertinent to seek to elaborate or otherwise change the model.

35
 Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

Table 1: Zheng statistics for tests on systematic utilities V

Air Bus Car Train

V(Air) 0.2 0.8 -0.2 0.0
V(Bus) 3.2 3.7 1.9 6.0
V(Car) 5.5 9.7 2.7 2.4
V(Train) 5.0 3.4 1.12 2.4

4.2 Demand for alternative vehicles

The example in this section is from McFadden and Train (2000), who illustrate
specification testing of the mixed logit model using a dataset on the demand
for alternative-fuelled vehicles. They take an existing model and apply a test to
detect that further mixing of some parameters is required.
Their dataset is available on the website of the Journal of Applied Economet-
rics. It has 4654 respondents who were asked to choose among six alternatives
characterised by price, range, acceleration, top speed etc. I reproduced the esti-
mates of McFadden and Train’s final model6 and increased the number of draws
used in the simulation to 500 Haltons. The model has a linear specification with
22 independent variables and 22 parameters for each alternative, six of which are
random with a normal distribution.
The only variable in the model with sufficiently many different values to be
called continuous is the price variable. I therefore first conducted the Zheng test
against the predicted probability for each alternative. The Zheng test statistics
were all very large, ranging from 10.1 to 81.6 for the residuals on the first alter-
native. Moreover, the residual plots revealed that residuals for each alternative
tended to all have the same sign. I therefore included alternative specific con-
stants in the model. They turned out to be extremely significant. The likelihood
improved from -7360.57 to -6977.27.7
Testing the model again still revealed evidence of some misspecification. The
Zheng statistics for residuals on the probability for alternative 1, e.g. ranged up
to 3.1, while the Zheng statistics for residuals on the price of the first alternative
did not indicate any misspecification. Repeating the test with an increased band-
width did, however, indicate misspecification. Inspection of the residual plots in-
dicated many cases where residuals seemed to follow a monotonous relationship
with the independent variable. I therefore extended the model by allowing for
full correlation between the six random parameters. This yielded a significant
improvement of the likelihood to -6969.95. From this point I did not find further
evidence of misspecification.
6
Except some cases where parameters were off by factors of ten.
7
These constants indicated that the first and third alternatives were preferred by respondents.
This could be an effect of the presentation of the choice situations or an effect of the design. I
have not pursued this issue.

36
 Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

5 Concluding remarks
The issue of potential model misspecification should be taken very seriously, since
statistical inference from a misspecified model is generally not valid. This paper
has pointed out how an existing nonparametric test of functional form, the Zheng
test, may be applied to discrete choice models in order to detect general misspec-
ification.
The curse of dimensionality is a general problem with nonparametric methods
and limits their usefulness. The paper points out how dimensionality may be
reduced by applying the test not to the raw data but to functions of the data.
This opportunity is not unique to the Zheng test but may be used with other
similar test.
Reducing the dimensionality of the data by applying the test to a function of
the data makes it a lot easier to apply the test and data requirements are reduced.
The associated cost is that the result is conditional on the chosen function, since
any given function may not be able to reveal a given misspecification. So while
significant rejection of the model using a given function of the data is conclusive
evidence against the model, it is possible that some misspecification will remain
undetected. It is therefore advisable to perform the test using several functions
of the data.
Application of the test is not difficult but requires some programming. In
the appendix I have included the Ox function that I used to compute the Zheng
statistic. The test has also been included in Biogeme along with facilities to
produce graphs, making the test very easy to use within that environment.

Acknowledgments
Mogens Fosgerau has received financial support from the Danish Social Science
Research Council.

A Ox code for Zheng test

I have here included a function in Ox that takes a column of unidimensional x,
a column of unidimensional residuals and a bandwidth to compute the Zheng
statistic. The bandwidth relates to data in the unit interval, the function scales
the bandwidth to the actual range of the data.

DoZheng(const x, const res, const bw)

{
decl Tnumerator, Tdenominator, Ttemp, T, i, bandwidth, density;
bandwidth = bw * (maxc(x) - minc(x));

Tnumerator = 0;

37
 Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

Tdenominator = 0;

for (i = 0; i < rows(x); i++)

{
density = densn((x[i] - x’)./bandwidth);
Ttemp = ( density .* res’) * res[i] ;
Ttemp[][i] = 0;
Tnumerator = Tnumerator + sumr(Ttemp);
Ttemp = 2 * ((density .^ 2) .* (res’).^2) * (res[i].^2) ;
Ttemp[][i] = 0;
Tdenominator = Tdenominator + sumr(Ttemp);
}
Tdenominator = sqrt(Tdenominator);
T = Tnumerator/Tdenominator;
// "T test value ~ N(0,1) is : ", T;
return T;
}

References
Bierlaire, M., 2003. BIOGEME: a free package for the estimation of discrete choice
models. Proceedings of the 3rd Swiss Transport Research Conference, Monte Verità,
Ascona, Switzerland.
Bierlaire, M., 2005. An introduction to Biogeme. biogeme.epfl.ch.
Doornik, J. A., 2001. Ox: An Object-Oriented Matrix Language. Timberlake Consultants
Press, London.
Ellison, G., Ellison, S. F., 2000. A simple framework for nonparametric specification
testing. Journal of Econometrics 96 (1), 1–23.
Fosgerau, M., 2006. Investigating the distribution of the value of travel time savings.
Transportation Research Part B: Methodological 40 (8), 688–707.
Fosgerau, M., 2007. Using nonparametrics to specify a model to measure the value of
travel time. Transportation Research Part A: Policy and Practice 41 (9), 842–856.
Fosgerau, M., Bierlaire, M., 2007. A practical test for the choice of mixing distribution
in discrete choice models. Transportation Research Part B: Methodological 41 (7),
784–794.
Gourieroux, C., Monfort, A., Renault, E., Trognon, A., 1987a. Generalised residuals.
Journal of Econometrics 34 (1-2), 5–32.
Gourieroux, C., Monfort, A., Renault, E., Trognon, A., 1987b. Simulated residuals.
Journal of Econometrics 34 (1-2), 201–252.
Härdle, W., 1990. Applied Nonparametric Regression. Vol. 19 of Econometric Society
Monograph Series. Cambridge University Press.
Lechner, M., 1991. Testing logit models in practice. Empirical Economics 16, 177–198.
Li, Q., Racine, J. S., 2007. Nonparametric Econometrics: Theory and Practice. Princeton
University Press, Princeton and Oxford.
McFadden, D., 1987. Regression-based specification tests for the multinomial logit model.
Journal of Econometrics 34 (1-2), 63–82.

38
 Fosgerau, Journal of Choice Modelling, 1(1), 2008, 26-39

McFadden, D., Train, K., 2000. Mixed mnl models for discrete response. Journal of
Applied Econometrics 15, 447–470.
Pagan, A., Ullah, A., 1999. Nonparametric Econometrics. Cambridge University Press.
Yatchew, A., 2003. Semiparametric Regression for the Applied Econometrician. Themes
in modern econometrics. Cambridge University Press.
Zheng, J. X., 1996. A consistent test of functional form via nonparametric estimation
techniques. Journal of Econometrics 75 (2), 263–289.

39