Article Navigation
Article Contents
-
Summary
-
1. Introduction
-
2. Model and missing mark data
-
3. Estimation procedure with missing marks
-
4. Testing of mark-specific vaccine efficacy
-
5. Simulation study
-
6. Analysis of the RV144 Thai trial
-
Acknowledgements
-
References
-
Appendix A: symptotic results
-
A.1. Proof of theorem 1
- < Previous
- Next >
Journal Article
, Peter B. Gilbert University of Washington , Seattle, USA Fred Hutchinson Cancer Research Center , Seattle, USA Address for correspondence: Peter B. Gilbert, Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue North, PO Box 19024, Seattle, WA 98109, USA. E-mail: pgilbert@scharp.org Search for other works by this author on: Oxford Academic Yanqing Sun University of North Carolina at Charlotte , USA Search for other works by this author on: Oxford Academic
Journal of the Royal Statistical Society Series C: Applied Statistics, Volume 64, Issue 1, January 2015, Pages 49–73, https://doi.org/10.1111/rssc.12067
Published:
03 July 2014
Article history
Received:
01 February 2013
Accepted:
01 March 2014
Published:
03 July 2014
- Split View
- Views
- Article contents
- Figures & tables
- Video
- Audio
- Supplementary Data
-
Cite
Cite
Peter B. Gilbert, Yanqing Sun, Inferences on Relative Failure Rates in Stratified Mark-Specific Proportional Hazards Models with Missing Marks, with Application to Human Immunodeficiency Virus Vaccine Efficacy Trials, Journal of the Royal Statistical Society Series C: Applied Statistics, Volume 64, Issue 1, January 2015, Pages 49–73, https://doi.org/10.1111/rssc.12067
Close
Search
Close
Search
Advanced Search
Search Menu
Summary
The paper develops hypothesis testing procedures for the stratified mark-specific proportional hazards model in the presence of missing marks. The motivating application is preventive human immunodeficiency virus (HIV) vaccine efficacy trials, where the mark is the genetic distance of an infecting HIV sequence to an HIV sequence represented inside the vaccine. The test statistics are constructed on the basis of two-stage efficient estimators, which utilize auxiliary predictors of the missing marks. The asymptotic properties and finite sample performances of the testing procedures are investigated, demonstrating double robustness and effectiveness of the predictive auxiliaries to recover efficiency. The methods are applied to the RV144 vaccine trial.
Augmented inverse probability weighting, Auxiliary marks, Competing risks failure time data, Genetic data, Proportional hazards model, Semiparametric model
1. Introduction
The primary objective of a preventive human immunodeficiency virus (HIV) vaccine efficacy trial is to assess vaccine efficacy VE to prevent HIV infection, where typically VE is defined as 1 minus the hazard ratio (vaccine/placebo) of HIV infection diagnosis. However, the great genetic variability of HIV poses a central challenge to developing a highly efficacious vaccine (Fauci et al., 2008). The trial population is exposed to many HIV genotypes but the vaccine contains only a few, and the vaccine is less likely to protect against HIVs with greater genetic distance from the sequences inside the vaccine (Gilbert et al., 1999). The trial has objectives to assess whether and how the vaccine impacts the infection rate with any HIV genotype and whether and how the vaccine effect varies by HIV genotype; assessment of this objective has been named ‘sieve analysis’ (Gilbert et al., 1998). Gilbert et al. (2008), Sun et al. (2009) and Sun and Gilbert (2012) developed sieve analysis methods using the competing risks failure time framework (Prentice et al., 1978), which attach a continuous ‘mark’ variable to HIV-infected subjects that measures the genetic distance of an infecting HIV sequence to a sequence inside the vaccine. The goal of the sieve analysis methods is evaluation of mark-specific vaccine efficacy, which here is defined as 1 minus the mark-specific hazard ratio (vaccine/placebo) of infection. Beyond HIV, the methods apply generally to any preventative vaccine efficacy trial for which the pathogen targeted by the vaccine is genetically diverse, which includes influenza, malaria, tuberculosis, dengue, streptococcus pneumoniae, human papilloma virus and hepatitis C virus.
Gilbert et al. (2008) and Sun et al. (2009) assumed no missing mark data in infected subjects, whereas Sun and Gilbert (2012) allowed marks missing at random. In practice there are missing marks; for example in the Vax004 trial 32 of 368 infected subjects had no HIV sequence data (Gilbert et al., 2008), owing to drop-out or to inability of the HIV sequencing technology to measure the infecting HIV sequence, and in the ‘Step’ trial 22 of 88 infected subjects had no HIV sequence data (Rolland et al., 2011). Whereas it is of scientific interest to evaluate a mark defined on the basis of the earliest available HIV sequence, a mark of particular scientific interest is defined on the basis of an HIV sequence measured near the time of acquisition, which is missing in a much larger fraction of infected subjects owing to the periodic (typically 6-monthly) diagnostic tests for HIV infection. Specifically, HIV sequences are measured from the earliest available post-infection blood sample, and a ‘near acquisition’ or ‘early’ sample may be defined as a sample that has been documented to be sufficiently near acquisition. In the Step trial, only 23 of the 66 infected subjects with sequence data had an early mark measured, defined as sampling within 3 weeks. Sun and Gilbert (2012) have provided details on the HIV testing algorithm that is used to define an early mark.
Sun and Gilbert (2012) is the only reference on sieve analysis that accommodates missing continuous marks. It develops two valid estimation approaches based on the stratified mark-specific proportional hazards model. The first uses inverse probability weighting (IPW) of the complete-case estimator, which leverages auxiliary predictors of whether the mark is observed, whereas the second, adapting Robins et al. (1994), augments the IPW complete-case estimator with auxiliary predictors of the missing marks. Sun and Gilbert (2012) restricted attention to estimation methods, and this paper is a sequel that develops corresponding inferential or hypothesis testing methods based on the augmented IPW estimator. An important new component of this work compared with the previous work is to centre it on the sieve analysis of the RV144 Thai trial, which recently delivered the landmark result that a prime boost HIV vaccine appeared to provide partial protection against HIV infection (estimated VE =31%; 95% confidence interval 1–51%; p = 0.04; Rerks-Ngarm et al. (2009)). This result has stimulated intense interest in sieve analysis, for two reasons. First, there is controversy about whether the vaccine is really partially working versus a false positive result (Gilbert et al., 2011), and the sieve analysis of HIV sequences can help to resolve this question. In particular, if evidence is found that the vaccine efficacy declines with genetic distance, and the distance is defined on the basis of known parts of HIV that contain putatively protective antibody epitopes, then an interpretation of real vaccine efficacy is supported. Secondly, the HIV vaccine field is grappling with how to modify the tested vaccine to increase its potential vaccine efficacy for the next efficacy trial, and understanding the relationship between vaccine efficacy and the genetic distance provides direct guidance on which HIV sequences to put inside the next generation vaccines.
This paper is organized as follows. Notation, assumptions and the stratified mark-specific proportional hazards model are introduced in Section 2. Background on the estimation procedures that are needed for the testing procedures are described in Section 3. The testing procedures are developed, and asymptotic properties described, in Section 4. The finite sample performances of the tests are evaluated via simulations in Section 5. The application to the Thai trial is given in Section 6, and the asymptotic results and their proofs are placed in Appendix A.
The programs that were used to analyse the data can be obtained from
https://academic.oup.com/jrsssc/issue/
2. Model and missing mark data
2.1. Stratified mark-specific proportional hazards model
Let T be the failure time, V a continuous mark variable with bounded support [0,1] and Z(t) a possibly time-dependent p-dimensional covariate. The mark V is observable only when T is observed. Suppose that the conditional mark-specific hazard function at time t given the covariate history Z(s), for s≤t, depends on the current value Z(t) only. We consider the stratified mark-specific proportional hazards model
(1)
where is the conditional mark-specific hazard function given covariate z(t) for an individual in the kth stratum, is the unspecified baseline hazard function for the kth stratum, β(v) is the p-dimensional unknown regression coefficient function of v and K is the number of strata. Model (1) allows different baseline functions for different strata and flexibly allows for arbitrary mark-specific infection hazards over time in the placebo group. In practice, different key subgroups (e.g. men and women in the Thai trial) are assigned different baseline mark-specific hazards of HIV infection.
Arranging , so that β1(v) is the coefficient for vaccination status and β2(v) for other covariates, the covariate- and stratum-adjusted mark-specific vaccine efficacy VE(v)=1− exp {β1(v)}. Sun et al. (2009) developed some statistical procedures for model (1) under K = 1 based on observations of the random variables (X,Z(·),V) for δ = 1 and (X,Z(·)) for δ = 0, where X = min{T,C}, δ = I(T≤C) and C is a censoring random variable. Sun and Gilbert (2012) developed estimation procedures for model (1) allowing V to be missing for some subjects with δ = 1; these methods incorporate auxiliary covariates and/or auxiliary mark variables that inform about the probability V that is observed and about the distribution of V. This paper develops parallel hypothesis testing procedures for assessing VE(v). As summarized in Section 1, the two objectives are to assess whether the vaccine efficacy ever deviates from 0 (i.e. test VE(v)=0) and to assess whether the vaccine efficacy changes with the mark (i.e. test VE(v)=VE).
2.2. Missing data assumptions
Let R be the indicator of whether all possible data are observed for a subject; R = 1 if either δ = 0 (right censored) or if δ = 1 and V is observed; and R = 0 otherwise. Auxiliary variables A may be helpful for predicting missing marks. Since the mark can only be missing for failures, supplemental information is potentially useful only for failures, for predicting missingness and for informing about the distribution of missing marks. For example, if V is defined on the basis of the early virus, then V*, the auxiliary mark information, may include sequences of later sampled viruses and can be considered a subset of A. In general, A could include multiple viral sequences per infected subject at multiple time points, giving information on intrasubject HIV evolution. The relationship between A and V can be modelled to help to predict V (see Section 5 for a simulated example).
We assume that C is conditionally independent of (T,V) given Z(·) and the stratum. We also assume that V is missing at random (Rubin, 1976), i.e., given δ = 1 and W=(T,Z(T),A), the probability that V is missing depends only on the observed W, not on the value of V; this assumption is expressed as
(2)
Let where Q=(δ,W). Then πk(Q)=δrk(W)+1−δ. The missingness at random (MAR) assumption (2) also implies that V is independent of R given Q:
(3)
Define and . The stratum-specific definitions of rk(w) and ρk(v,w) allow the models of the probability of complete-case and of the mark distribution to differ across strata.
Let τ be the end of the follow-up period, and be the number of subjects in the kth stratum; the total sample size is . Let be independent and identically distributed replicates of {X,Z(·),δ, R,V,A} from the kth stratum. The observed data are , where Oki={Xki,Zki(·),Rki, RkiVki,Aki} for δki=1 and Oki={Xki,Zki(·),Rki=1} for δki=0. We assume that the Oki are independent for all subjects.
2.3. Hypotheses to test
We develop procedures for testing the following two sets of hypotheses. Let [a,b]⊂(0,1). The first set of hypotheses is
The second set of hypotheses is
The null hypothesis H10 implies that the vaccine affords no protection (nor increased risk) against any HIV genotype. The ordered alternative H1m indicates that the vaccine provides protection for at least some of the HIV genotypes, whereas H1a indicates that the vaccine provides protection and/or increased risk for some HIV genotypes. The null hypothesis H20 implies that there is no difference in vaccine protection against different HIV genotypes. The ordered alternative H2m indicates that vaccine efficacy decreases with v and H2a indicates that the vaccine efficacy changes with v. With β1(v) the first component of β(v), the first set of hypotheses is equivalent to H10:β1(v)=0 for v ∈ [a,b] versusH1a:β1(v)≠0 for some v or H1m:β1(v)≤0 with strict inequality for some v. The second set of hypotheses is equivalent to H20:β1(v) does not depend on v ∈ [a,b] versusH2a:β1(v) depends on v or H2m:β1(v) increases as v increases. We develop testing procedures for detecting departures from H10 in the direction of H1m and H1a and for detecting departures from H20 in the direction of H2m and H2a. The procedures are developed on the basis of the augmented IPW complete-case estimator that was developed by Sun and Gilbert (2012).
3. Estimation procedure with missing marks
The augmented IPW (AIPW) estimator for model (1) is obtained in two stages. First the IPW complete-case estimator is derived and second the AIPW estimator is obtained, which improves efficiency by accounting for information in the conditional distribution of V given the auxiliaries.
Let rk(Wki, ψk) be the parametric model for the probability of the complete case, with rk(Wki) defined in equation (2), where Wki=(Tki,Zki(Tki),Aki) and ψk is a q-dimensional parameter. For example, we can assume the logistic model with for those with δki=1, where Wki=(Tki,Zki(Tki),Aki). By equation (2), the maximum likelihood estimator of ψ=(ψ1,…,ψK) is obtained by maximizing the observed data likelihood
(4)
Let K(x) be a kernel function with support [−1,1] and let h = hn be a bandwidth. Let Nki(t,v)=I(Xki≤t,δki=1,Vki≤v) and Yki(t)=I(Xki≥t). Let Qki=(δki,Wki) and πk(Qki, ψk)=δkirk(Wki, ψk)+1−δki. The first-stage estimator is the IPW estimator , which solves the estimating equation for β, where
(5)
Here
and
for j = 0,1, where z⊗0=1 and z⊗1=z for any . The score function (5) can be viewed as an extension of the score function that is used for the cause-specific Cox model (Prentice et al., 1978) for a particular failure cause J = j, for which the counting process counts only events of type j. It borrows strength from observations having marks in the neighbourhood of v. The kernel function is designed to give greater weight to observations with marks near v than those further away.
The baseline function λ0k(t,v) can be estimated by , obtained by smoothing the increments of the following estimator of the doubly cumulative baseline function :
(6)
For example, one can use the following kernel smoothing:
(7)
where and , with K(1)(·) and K(2)(·) the kernel functions and and the bandwidths.
Following Robins et al. (1994), Sun and Gilbert (2012) proposed a more efficient procedure for estimating equation (1) by incorporating the knowledge of ρk(w,v) in the estimation procedure. Let w=(t,z,a) and . Then
(8)
If no auxiliary variables are available or if Aki is conditionally independent of Vki given (Tki,Zki, δki), then
In this case, ρk(w,v) can be estimated by
where . When the auxiliary marks Aki are correlated with Vki conditional on Tki,Zki and δki=1, the conditional distribution ρk(w,v) involves the function , for which a parametric or semiparametric model may be developed to describe the dependence between Aki and Vki. Let be an estimator of with a convergence rate of at least (nh)−1/2. Then ρk(w,v) can be estimated by
(9)
Let and . The AIPW estimating equation for β is , where
(10)
and
and
for j = 0,1. The AIPW estimator of β(v) solves equation (10) and is denoted by . The estimator of the cumulative function is given by . Note that there is no in ; this is a difference between the IPW and AIPW estimators.
To implement the estimation procedures in practice, one can use arbitrary auxiliaries for estimating ; these auxiliaries may include covariates and marks at multiple time points pre infection and post infection respectively. In contrast, although in principle arbitrary auxiliaries may also be used for the terms in equation (9), owing to the curse of dimensionality the method is expected to perform best in practice with a univariate auxiliary, where semiparametric or fully parametric models for would be required to include multivariate auxiliaries.
Sun and Gilbert (2012) proved that the estimators and are consistent and that is more efficient than . In the next section, we develop some hypothesis testing procedures for assessing mark-specific vaccine efficacy based on .
4. Testing of mark-specific vaccine efficacy
The covariate-adjusted vaccine efficacy VE(v) is defined through the first component of β(v). Let be the first component of the cumulative coefficient function B(v). The hypothesis tests concerning VE(v) are constructed on the basis of the first component of the AIPW estimator . The cumulative estimator has more stable large sample behaviour and a faster convergence rate than .
Let for v ∈ [a,b]. In Appendix A we show that WB(v), v ∈ [a,b], converges weakly to a p-dimensional mean 0 Gaussian process with continuous sample paths on v ∈ [a,b]. Further, the distribution of WB(v), for v ∈ [a,b], can be approximated by using the Gaussian multipliers resampling method (of Lin et al. (1993)) based on
where are independent and identically distributed standard normal random variables and is defined in expression (22) in Appendix A. Let and be the first component of WB(v) and respectively. With the Gaussian multipliers method, the variance can be consistently estimated by , where is the first component on the diagonal of the covariance given in expression (23) in Appendix A.
4.1. Testing the null hypothesis H10
Consider the test process , v ∈ [a,b]. Then , v ∈ [a,b]. Under H10, for v ∈ [a,b], which motivates the following test statistics for testing H10:
The test statistics and capture general departures H1a, whereas the test statistics and are sensitive to the monotone departures H1m. It is easy to derive that all the test statistics , , and are consistent against their respective alternative hypotheses, and Appendix A derives their limiting distributions under H10.
Under H10, the distribution of Q(1)(v), v ∈ [a,b], can be approximated by the conditional distribution of , v ∈ [a,b], given the observed data sequence. Hence, the distributions of , , and under H10 can be approximated by the conditional distributions of , , and , given the observed data sequence, respectively. The critical values and of the test statistics and can be approximated by the (1−α)-quantile of and , which can be obtained by repeatedly generating a large number, say 500, of independent sets of normal samples while holding the observed data sequence fixed. Similarly, the critical values and of the test statistics and can be approximated by the α-quantile of and , which again can be obtained by repeatedly generating independent sets of normal samples . At the significance level α, the tests based on and reject H10 in favour of H1a if and respectively, and the tests based on and reject H10 in favour of H1m if and respectively.
4.2. Testing the null hypothesis H20
Let
Then
(11)
where
is a transformation of F1(·). We note that under H20 and under the alternatives, motivating Q(2)(v) as the test process and the following test statistics for testing H20:
where . We choose to avoid 0 in the denominator of Q(2)(v). In practice, one can choose close to a to make use of available data and to ensure that the tests are consistent.
By the asymptotic results shown in Appendix A and the continuous mapping theorem, under H20 the distribution of Q(2)(v), v ∈ [a,b], can be approximated by the conditional distribution of , v ∈ [a,b], given the observed data sequence. Hence, the distributions of , , and under H20 can be approximated by the conditional distributions of
and
given the observed data sequence, respectively. Similarly to Section 4.1, the respective critical values and of the test statistics and can be approximated by the (1−α)-quantiles of the conditional distributions of and obtained through repeatedly generating independent sets of normal samples while holding the observed data sequence fixed. The critical values and for and can be approximated similarly. At the significance level α, the tests based on and reject H20 in favour of H2a if and respectively, and the tests based on and reject H20 in favour of H2m if and respectively.
The tests and capture general departures H2a whereas the tests and are sensitive to the monotone departure H2m. Note that the derivative under H2m with strict inequality for at least some v ∈ [a,b]. This plus the fact that is non-decreasing with lead to the results that the tests based on and are consistent against H2m and the tests based on and are consistent against H2a. The proofs are given in the second paragraph following theorem 1 in Appendix A.
In Sections 4.1 and 4.2, we considered two types of test statistics, namely the integration-based test statistics and the supremum-based test statistics, for each pair of hypotheses. The former are generalizations of the Cramér–von Mises test statistic and involve integration of deviations over the whole range of the mark, whereas the latter are extensions of the classic Kolmogorov–Smirnov test statistic for testing the goodness of fit of a distribution function, and take the supremum of such deviations. As demonstrated in a comprehensive analysis of the relative powers of the classic Kolmogorov–Smirnov test and the Cramér–von Mises test by Stephens (1974), we expect that the two types of test statistic have different powers for different true alternative distributions. The integration-based test statistics are best suited for situations where the true alternative distribution deviates a little over the whole support of the mark and the supremum-based test statistics may have more power against situations where the true alternative has large deviations over a small section of the support. For example, for testing differential VE(v), H20, the supremum-based tests will tend to be relatively more powerful if is very high for a small range of marks near a and declines sharply to 0 and is constant at 0 for all other marks.
5. Simulation study
5.1. Numerical assessment of the tests under correctly specified models
We conduct a simulation study to evaluate the finite sample performance of the testing procedures proposed. The empirical sizes and powers of the test statistics are assessed for various models, sample sizes (500 and 800) and choices of bandwidths. The powers of the tests are evaluated in both situations where a correlated auxiliary variable is used and where it is absent.
We consider K = 1 stratum. Let Zki be the treatment indicator with P(Zki=1)=0.5. The (Tki, Vki) are generated from the following mark-specific proportional hazards model:
(12)
where α, β and γ are constants. Under model (12), λ0(t,v)= exp (γv) and VE(v)=1− exp (α+βv). For α = 0 and β = 0, VE(v)=0, indicating no vaccine efficacy, and, for β = 0, VE(v)=VE, indicating mark invariant vaccine efficacy, whereas β>0 indicates VE(v) decreasing in v. We examine the hypothesis testing procedures for the following specific models M1–M5 respectively:
implying that VE(v)=0;
implying that VE(v) does not depend on v;
implying that VE(v) decreases;
implying that VE(v) decreases;
implying that VE(v) decreases.
We generate the censoring times from an exponential distribution, independent of (T,V), with censoring rates ranging from 20% to 30%. We take τ = 2.0. The complete-case indicator Rki is generated with conditional probability , where
(13)
With ψk0=0.2 and ψk1=−0.2 about 50% of observed failures are missing marks.
Conditionally on (Tki, Zki, Vki), we assume that the auxiliary marks follow the model
(14)
for k = 1,…,K, where Vki are the possibly missing marks, Uki is uniformly distributed on [0,1] independent of Vki and θ>0 is an association parameter between Aki and Vki. The correlation coefficient ρ between Aki and Vki is 1 for θ = 0. Since Aki is observed for all observed failure times, the AIPW estimator in this case is the full data estimator. The Aki and Vki are independent for θ = ∞, yielding ρ = 0. In addition, the θ-values of 0.8, 0.4 and 0.2 correspond to ρ = 0.78, 0.92, 0.98.
Under model (14), the conditional density of Aki given (Tki, Zki, Vki) is
(15)
The likelihood function for θ is
It is easy to show that the maximum likelihood estimator equals
The density estimator is plugged into equation (9) to obtain , which is used to construct the AIPW estimator of β in equation (10).
The performances of the test procedures proposed are evaluated through simulations for the models described in expressions (12), (13) and (14) under the settings M1–M5, where M1 is a setting under the null hypothesis H10 and M2 is a setting under the null hypothesis H20. We consider the situations where no auxiliary information is provided and where the correlation between the auxiliary mark and the mark of interest is ρ = 0.92 (under model (14) with θ = 0.4). Table 1 presents the empirical sizes and powers of the tests , , and for testing H10 at the nominal level 0.05. Table 2 presents the empirical sizes and powers of the tests , , and for testing H20 at the nominal level 0.05. The results are presented for n = 500 with and , 0.2, and for n = 800 with and , 0.15. We take a = 0, b = 1 and for the tests. The Epanechnikov kernel is used throughout the numerical analysis.
Table 1
Open in new tab
Empirical sizes and powers of the tests , , and for testing H10 at the nominal level 0.05 for ρ = 0 and 0.92 when 50% of the marks are missing†
Model | (α,β,γ) | n | h | Size/power | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|
ρ = 0 | ρ = 0.92 | ||||||||||
M1 | (0,0,0.3) | 500 | 0.15 | 5.4 | 4.0 | 4.0 | 5.0 | 4.6 | 4.2 | 3.8 | 4.2 |
0.20 | 5.0 | 4.4 | 4.6 | 5.2 | 4.8 | 4.0 | 4.2 | 3.6 | |||
800 | 0.10 | 3.8 | 3.6 | 4.2 | 4.2 | 3.8 | 3.8 | 5.4 | 4.8 | ||
0.15 | 4.0 | 3.8 | 4.6 | 4.6 | 5.0 | 4.4 | 5.4 | 5.6 | |||
M3 | (−0.6,0.6,0.3) | 500 | 0.15 | 68.2 | 67.0 | 79.4 | 76.0 | 73.2 | 74.6 | 83.2 | 85.4 |
0.20 | 63.2 | 65.0 | 75.8 | 74.2 | 69.2 | 71.4 | 79.8 | 82.6 | |||
800 | 0.10 | 88.2 | 86.2 | 94.6 | 90.4 | 92.0 | 93.0 | 95.0 | 97.2 | ||
0.15 | 87.4 | 86.6 | 92.8 | 90.8 | 89.2 | 90.6 | 93.4 | 95.2 | |||
M4 | (−1.2,1.2,0.3) | 500 | 0.15 | 99.6 | 99.4 | 99.8 | 99.8 | 99.8 | 100 | 99.8 | 100 |
0.20 | 99.4 | 99.0 | 99.6 | 99.8 | 99.6 | 99.8 | 99.8 | 100 | |||
800 | 0.10 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | ||
0.15 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | |||
M2 | (−0.69,0,0.3) | 500 | 0.15 | 100 | 100 | 100 | 99.8 | 100 | 100 | 100 | 100 |
0.20 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | |||
800 | 0.10 | 100 | 99.8 | 100 | 100 | 99.8 | 99.8 | 99.8 | 99.8 | ||
0.15 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 |
Model | (α,β,γ) | n | h | Size/power | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|
ρ = 0 | ρ = 0.92 | ||||||||||
M1 | (0,0,0.3) | 500 | 0.15 | 5.4 | 4.0 | 4.0 | 5.0 | 4.6 | 4.2 | 3.8 | 4.2 |
0.20 | 5.0 | 4.4 | 4.6 | 5.2 | 4.8 | 4.0 | 4.2 | 3.6 | |||
800 | 0.10 | 3.8 | 3.6 | 4.2 | 4.2 | 3.8 | 3.8 | 5.4 | 4.8 | ||
0.15 | 4.0 | 3.8 | 4.6 | 4.6 | 5.0 | 4.4 | 5.4 | 5.6 | |||
M3 | (−0.6,0.6,0.3) | 500 | 0.15 | 68.2 | 67.0 | 79.4 | 76.0 | 73.2 | 74.6 | 83.2 | 85.4 |
0.20 | 63.2 | 65.0 | 75.8 | 74.2 | 69.2 | 71.4 | 79.8 | 82.6 | |||
800 | 0.10 | 88.2 | 86.2 | 94.6 | 90.4 | 92.0 | 93.0 | 95.0 | 97.2 | ||
0.15 | 87.4 | 86.6 | 92.8 | 90.8 | 89.2 | 90.6 | 93.4 | 95.2 | |||
M4 | (−1.2,1.2,0.3) | 500 | 0.15 | 99.6 | 99.4 | 99.8 | 99.8 | 99.8 | 100 | 99.8 | 100 |
0.20 | 99.4 | 99.0 | 99.6 | 99.8 | 99.6 | 99.8 | 99.8 | 100 | |||
800 | 0.10 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | ||
0.15 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | |||
M2 | (−0.69,0,0.3) | 500 | 0.15 | 100 | 100 | 100 | 99.8 | 100 | 100 | 100 | 100 |
0.20 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | |||
800 | 0.10 | 100 | 99.8 | 100 | 100 | 99.8 | 99.8 | 99.8 | 99.8 | ||
0.15 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 |
† The bandwidths are and . Each entry is based on 500 Gaussian multipliers samples and 500 repetitions.
Table 1
Open in new tab
Empirical sizes and powers of the tests , , and for testing H10 at the nominal level 0.05 for ρ = 0 and 0.92 when 50% of the marks are missing†
Model | (α,β,γ) | n | h | Size/power | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|
ρ = 0 | ρ = 0.92 | ||||||||||
M1 | (0,0,0.3) | 500 | 0.15 | 5.4 | 4.0 | 4.0 | 5.0 | 4.6 | 4.2 | 3.8 | 4.2 |
0.20 | 5.0 | 4.4 | 4.6 | 5.2 | 4.8 | 4.0 | 4.2 | 3.6 | |||
800 | 0.10 | 3.8 | 3.6 | 4.2 | 4.2 | 3.8 | 3.8 | 5.4 | 4.8 | ||
0.15 | 4.0 | 3.8 | 4.6 | 4.6 | 5.0 | 4.4 | 5.4 | 5.6 | |||
M3 | (−0.6,0.6,0.3) | 500 | 0.15 | 68.2 | 67.0 | 79.4 | 76.0 | 73.2 | 74.6 | 83.2 | 85.4 |
0.20 | 63.2 | 65.0 | 75.8 | 74.2 | 69.2 | 71.4 | 79.8 | 82.6 | |||
800 | 0.10 | 88.2 | 86.2 | 94.6 | 90.4 | 92.0 | 93.0 | 95.0 | 97.2 | ||
0.15 | 87.4 | 86.6 | 92.8 | 90.8 | 89.2 | 90.6 | 93.4 | 95.2 | |||
M4 | (−1.2,1.2,0.3) | 500 | 0.15 | 99.6 | 99.4 | 99.8 | 99.8 | 99.8 | 100 | 99.8 | 100 |
0.20 | 99.4 | 99.0 | 99.6 | 99.8 | 99.6 | 99.8 | 99.8 | 100 | |||
800 | 0.10 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | ||
0.15 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | |||
M2 | (−0.69,0,0.3) | 500 | 0.15 | 100 | 100 | 100 | 99.8 | 100 | 100 | 100 | 100 |
0.20 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | |||
800 | 0.10 | 100 | 99.8 | 100 | 100 | 99.8 | 99.8 | 99.8 | 99.8 | ||
0.15 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 |
Model | (α,β,γ) | n | h | Size/power | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|
ρ = 0 | ρ = 0.92 | ||||||||||
M1 | (0,0,0.3) | 500 | 0.15 | 5.4 | 4.0 | 4.0 | 5.0 | 4.6 | 4.2 | 3.8 | 4.2 |
0.20 | 5.0 | 4.4 | 4.6 | 5.2 | 4.8 | 4.0 | 4.2 | 3.6 | |||
800 | 0.10 | 3.8 | 3.6 | 4.2 | 4.2 | 3.8 | 3.8 | 5.4 | 4.8 | ||
0.15 | 4.0 | 3.8 | 4.6 | 4.6 | 5.0 | 4.4 | 5.4 | 5.6 | |||
M3 | (−0.6,0.6,0.3) | 500 | 0.15 | 68.2 | 67.0 | 79.4 | 76.0 | 73.2 | 74.6 | 83.2 | 85.4 |
0.20 | 63.2 | 65.0 | 75.8 | 74.2 | 69.2 | 71.4 | 79.8 | 82.6 | |||
800 | 0.10 | 88.2 | 86.2 | 94.6 | 90.4 | 92.0 | 93.0 | 95.0 | 97.2 | ||
0.15 | 87.4 | 86.6 | 92.8 | 90.8 | 89.2 | 90.6 | 93.4 | 95.2 | |||
M4 | (−1.2,1.2,0.3) | 500 | 0.15 | 99.6 | 99.4 | 99.8 | 99.8 | 99.8 | 100 | 99.8 | 100 |
0.20 | 99.4 | 99.0 | 99.6 | 99.8 | 99.6 | 99.8 | 99.8 | 100 | |||
800 | 0.10 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | ||
0.15 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | |||
M2 | (−0.69,0,0.3) | 500 | 0.15 | 100 | 100 | 100 | 99.8 | 100 | 100 | 100 | 100 |
0.20 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | |||
800 | 0.10 | 100 | 99.8 | 100 | 100 | 99.8 | 99.8 | 99.8 | 99.8 | ||
0.15 | 100 | 100 | 100 | 100 | 100 | 100 | 100 | 100 |
† The bandwidths are and . Each entry is based on 500 Gaussian multipliers samples and 500 repetitions.
Table 2
Open in new tab
Empirical sizes and powers of the tests , , and for testing H20 at the nominal level 0.05 for ρ = 0 and 0.92 when 50% of the marks are missing†
Model | (α,β,γ) | n | h | Size/power | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|
ρ = 0 | ρ = 0.92 | ||||||||||
M2 | (−0.69,0,0.3) | 500 | 0.15 | 5.6 | 4.8 | 5.8 | 5.8 | 7.6 | 7.2 | 7.4 | 7.0 |
0.20 | 5.8 | 4.8 | 5.4 | 5.2 | 6.6 | 6.6 | 6.6 | 7.4 | |||
800 | 0.10 | 6.4 | 5.0 | 5.6 | 5.8 | 6.2 | 5.8 | 7.2 | 7.0 | ||
0.15 | 6.6 | 5.2 | 5.8 | 5.6 | 6.0 | 5.6 | 6.0 | 6.6 | |||
M3 | (−0.6,0.6,0.3) | 500 | 0.15 | 16.8 | 17.0 | 22.4 | 25.2 | 20.6 | 25.8 | 32.6 | 37.4 |
0.20 | 14.2 | 15.8 | 22.2 | 24.8 | 19.4 | 24.2 | 31.8 | 34.6 | |||
800 | 0.10 | 26.0 | 25.8 | 35.2 | 36.4 | 36.0 | 38.0 | 46.0 | 49.2 | ||
0.15 | 25.4 | 25.8 | 34.8 | 35.6 | 34.0 | 36.0 | 45.4 | 47.4 | |||
M4 | (−1.2,1.2,0.3) | 500 | 0.15 | 44.4 | 46.2 | 59.0 | 63.2 | 63.6 | 68.4 | 76.4 | 80.2 |
0.20 | 42.2 | 44.0 | 57.2 | 59.6 | 61.4 | 65.8 | 73.2 | 75.8 | |||
800 | 0.10 | 66.2 | 67.6 | 75.2 | 78.0 | 82.8 | 86.6 | 90.6 | 91.8 | ||
0.15 | 64.6 | 66.2 | 74.0 | 77.0 | 80.6 | 84.4 | 88.4 | 91.2 | |||
M5 | (−1.5,1.5,0.3) | 500 | 0.15 | 64.5 | 66.5 | 75.0 | 76.5 | 81.0 | 85.6 | 88.8 | 90.4 |
0.20 | 61.0 | 62.6 | 72.2 | 72.2 | 77.8 | 82.4 | 86.8 | 89.4 | |||
800 | 0.10 | 80.8 | 85.6 | 87.6 | 91.4 | 94.6 | 96.2 | 97.6 | 98.4 | ||
0.15 | 78.6 | 84.8 | 87.8 | 91.4 | 94.4 | 95.6 | 95.8 | 97.8 |
Model | (α,β,γ) | n | h | Size/power | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|
ρ = 0 | ρ = 0.92 | ||||||||||
M2 | (−0.69,0,0.3) | 500 | 0.15 | 5.6 | 4.8 | 5.8 | 5.8 | 7.6 | 7.2 | 7.4 | 7.0 |
0.20 | 5.8 | 4.8 | 5.4 | 5.2 | 6.6 | 6.6 | 6.6 | 7.4 | |||
800 | 0.10 | 6.4 | 5.0 | 5.6 | 5.8 | 6.2 | 5.8 | 7.2 | 7.0 | ||
0.15 | 6.6 | 5.2 | 5.8 | 5.6 | 6.0 | 5.6 | 6.0 | 6.6 | |||
M3 | (−0.6,0.6,0.3) | 500 | 0.15 | 16.8 | 17.0 | 22.4 | 25.2 | 20.6 | 25.8 | 32.6 | 37.4 |
0.20 | 14.2 | 15.8 | 22.2 | 24.8 | 19.4 | 24.2 | 31.8 | 34.6 | |||
800 | 0.10 | 26.0 | 25.8 | 35.2 | 36.4 | 36.0 | 38.0 | 46.0 | 49.2 | ||
0.15 | 25.4 | 25.8 | 34.8 | 35.6 | 34.0 | 36.0 | 45.4 | 47.4 | |||
M4 | (−1.2,1.2,0.3) | 500 | 0.15 | 44.4 | 46.2 | 59.0 | 63.2 | 63.6 | 68.4 | 76.4 | 80.2 |
0.20 | 42.2 | 44.0 | 57.2 | 59.6 | 61.4 | 65.8 | 73.2 | 75.8 | |||
800 | 0.10 | 66.2 | 67.6 | 75.2 | 78.0 | 82.8 | 86.6 | 90.6 | 91.8 | ||
0.15 | 64.6 | 66.2 | 74.0 | 77.0 | 80.6 | 84.4 | 88.4 | 91.2 | |||
M5 | (−1.5,1.5,0.3) | 500 | 0.15 | 64.5 | 66.5 | 75.0 | 76.5 | 81.0 | 85.6 | 88.8 | 90.4 |
0.20 | 61.0 | 62.6 | 72.2 | 72.2 | 77.8 | 82.4 | 86.8 | 89.4 | |||
800 | 0.10 | 80.8 | 85.6 | 87.6 | 91.4 | 94.6 | 96.2 | 97.6 | 98.4 | ||
0.15 | 78.6 | 84.8 | 87.8 | 91.4 | 94.4 | 95.6 | 95.8 | 97.8 |
† The bandwidths are and . Each entry is based on 500 Gaussian multipliers samples and 500 repetitions.
Table 2
Open in new tab
Empirical sizes and powers of the tests , , and for testing H20 at the nominal level 0.05 for ρ = 0 and 0.92 when 50% of the marks are missing†
Model | (α,β,γ) | n | h | Size/power | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|
ρ = 0 | ρ = 0.92 | ||||||||||
M2 | (−0.69,0,0.3) | 500 | 0.15 | 5.6 | 4.8 | 5.8 | 5.8 | 7.6 | 7.2 | 7.4 | 7.0 |
0.20 | 5.8 | 4.8 | 5.4 | 5.2 | 6.6 | 6.6 | 6.6 | 7.4 | |||
800 | 0.10 | 6.4 | 5.0 | 5.6 | 5.8 | 6.2 | 5.8 | 7.2 | 7.0 | ||
0.15 | 6.6 | 5.2 | 5.8 | 5.6 | 6.0 | 5.6 | 6.0 | 6.6 | |||
M3 | (−0.6,0.6,0.3) | 500 | 0.15 | 16.8 | 17.0 | 22.4 | 25.2 | 20.6 | 25.8 | 32.6 | 37.4 |
0.20 | 14.2 | 15.8 | 22.2 | 24.8 | 19.4 | 24.2 | 31.8 | 34.6 | |||
800 | 0.10 | 26.0 | 25.8 | 35.2 | 36.4 | 36.0 | 38.0 | 46.0 | 49.2 | ||
0.15 | 25.4 | 25.8 | 34.8 | 35.6 | 34.0 | 36.0 | 45.4 | 47.4 | |||
M4 | (−1.2,1.2,0.3) | 500 | 0.15 | 44.4 | 46.2 | 59.0 | 63.2 | 63.6 | 68.4 | 76.4 | 80.2 |
0.20 | 42.2 | 44.0 | 57.2 | 59.6 | 61.4 | 65.8 | 73.2 | 75.8 | |||
800 | 0.10 | 66.2 | 67.6 | 75.2 | 78.0 | 82.8 | 86.6 | 90.6 | 91.8 | ||
0.15 | 64.6 | 66.2 | 74.0 | 77.0 | 80.6 | 84.4 | 88.4 | 91.2 | |||
M5 | (−1.5,1.5,0.3) | 500 | 0.15 | 64.5 | 66.5 | 75.0 | 76.5 | 81.0 | 85.6 | 88.8 | 90.4 |
0.20 | 61.0 | 62.6 | 72.2 | 72.2 | 77.8 | 82.4 | 86.8 | 89.4 | |||
800 | 0.10 | 80.8 | 85.6 | 87.6 | 91.4 | 94.6 | 96.2 | 97.6 | 98.4 | ||
0.15 | 78.6 | 84.8 | 87.8 | 91.4 | 94.4 | 95.6 | 95.8 | 97.8 |
Model | (α,β,γ) | n | h | Size/power | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|
ρ = 0 | ρ = 0.92 | ||||||||||
M2 | (−0.69,0,0.3) | 500 | 0.15 | 5.6 | 4.8 | 5.8 | 5.8 | 7.6 | 7.2 | 7.4 | 7.0 |
0.20 | 5.8 | 4.8 | 5.4 | 5.2 | 6.6 | 6.6 | 6.6 | 7.4 | |||
800 | 0.10 | 6.4 | 5.0 | 5.6 | 5.8 | 6.2 | 5.8 | 7.2 | 7.0 | ||
0.15 | 6.6 | 5.2 | 5.8 | 5.6 | 6.0 | 5.6 | 6.0 | 6.6 | |||
M3 | (−0.6,0.6,0.3) | 500 | 0.15 | 16.8 | 17.0 | 22.4 | 25.2 | 20.6 | 25.8 | 32.6 | 37.4 |
0.20 | 14.2 | 15.8 | 22.2 | 24.8 | 19.4 | 24.2 | 31.8 | 34.6 | |||
800 | 0.10 | 26.0 | 25.8 | 35.2 | 36.4 | 36.0 | 38.0 | 46.0 | 49.2 | ||
0.15 | 25.4 | 25.8 | 34.8 | 35.6 | 34.0 | 36.0 | 45.4 | 47.4 | |||
M4 | (−1.2,1.2,0.3) | 500 | 0.15 | 44.4 | 46.2 | 59.0 | 63.2 | 63.6 | 68.4 | 76.4 | 80.2 |
0.20 | 42.2 | 44.0 | 57.2 | 59.6 | 61.4 | 65.8 | 73.2 | 75.8 | |||
800 | 0.10 | 66.2 | 67.6 | 75.2 | 78.0 | 82.8 | 86.6 | 90.6 | 91.8 | ||
0.15 | 64.6 | 66.2 | 74.0 | 77.0 | 80.6 | 84.4 | 88.4 | 91.2 | |||
M5 | (−1.5,1.5,0.3) | 500 | 0.15 | 64.5 | 66.5 | 75.0 | 76.5 | 81.0 | 85.6 | 88.8 | 90.4 |
0.20 | 61.0 | 62.6 | 72.2 | 72.2 | 77.8 | 82.4 | 86.8 | 89.4 | |||
800 | 0.10 | 80.8 | 85.6 | 87.6 | 91.4 | 94.6 | 96.2 | 97.6 | 98.4 | ||
0.15 | 78.6 | 84.8 | 87.8 | 91.4 | 94.4 | 95.6 | 95.8 | 97.8 |
† The bandwidths are and . Each entry is based on 500 Gaussian multipliers samples and 500 repetitions.
Tables 1 and 2 show that all the tests have satisfactory empirical sizes close to the nominal level 0.05. The powers of the tests increase with sample size and they are not overly sensitive to the bandwidths selected. The powers of the tests for testing H10 increase as the model moves in the direction M1→M3→M4→M2, representing increased departure from the null hypothesis H10. The powers of the tests for testing H20 increase as the model moves in the direction M2→M3→M4→M5, representing increased departure from the null hypothesis H20. The tests utilizing the auxiliary marks have higher power than those without using the auxiliary marks.
As with any non-parametric smoothing procedure, one needs to select bandwidths carefully. In practice, the appropriate bandwidth selection can be based on a -fold cross-validation method (e.g. Efron and Tibshirani (1993), Hoover et al. (1998), Cai et al. (2000) and Tian et al. (2005)).
The testing procedures proposed properly handle missing marks under MAR with asymptotically correct significance levels. However, if only the observations with complete information are used, i.e. the complete-case analysis, then the testing procedures are expected often not to provide correct type I error control. We conduct a simulation study to evaluate the observed sizes of the proposed tests using the complete cases under two different models for missing the indicator Rki—model (13) and the model
(16)
For K = 1 both model (13) and model (16) yield about 50% missing marks among the observed failures. The sizes of , , and for testing H10 are evaluated under model M1 and the sizes of , , and for testing H20 are evaluated under model M2 (Table 3). Under model (13), the observed sizes for testing H10 are elevated (around 7–15%), whereas those for testing H20 remain around 5%. Under model (16), the observed sizes for testing H10 exceed 37% for all tests, whereas those for testing H20 reach 12% and 14% for tests and when n = 800.
Table 3
Open in new tab
Empirical sizes of the tests for H10 and H20 at the nominal level 0.05 using the complete cases under missingness completely at random when 50% of the marks are missing†
Model | Missingness model | n | h | Size | |||
---|---|---|---|---|---|---|---|
TestingH10 | |||||||
M1 | (13) | 500 | 0.20 | 0.14 | 0.10 | 0.12 | 0.15 |
800 | 0.15 | 0.10 | 0.07 | 0.11 | 0.11 | ||
(16) | 500 | 0.20 | 0.39 | 0.37 | 0.50 | 0.42 | |
800 | 0.15 | 0.50 | 0.46 | 0.63 | 0.55 | ||
Testing H20 | |||||||
M2 | (13) | 500 | 0.20 | 0.08 | 0.04 | 0.08 | 0.05 |
800 | 0.15 | 0.06 | 0.09 | 0.06 | 0.10 | ||
(16) | 500 | 0.20 | 0.08 | 0.07 | 0.08 | 0.05 | |
800 | 0.15 | 0.07 | 0.06 | 0.12 | 0.14 |
Model | Missingness model | n | h | Size | |||
---|---|---|---|---|---|---|---|
TestingH10 | |||||||
M1 | (13) | 500 | 0.20 | 0.14 | 0.10 | 0.12 | 0.15 |
800 | 0.15 | 0.10 | 0.07 | 0.11 | 0.11 | ||
(16) | 500 | 0.20 | 0.39 | 0.37 | 0.50 | 0.42 | |
800 | 0.15 | 0.50 | 0.46 | 0.63 | 0.55 | ||
Testing H20 | |||||||
M2 | (13) | 500 | 0.20 | 0.08 | 0.04 | 0.08 | 0.05 |
800 | 0.15 | 0.06 | 0.09 | 0.06 | 0.10 | ||
(16) | 500 | 0.20 | 0.08 | 0.07 | 0.08 | 0.05 | |
800 | 0.15 | 0.07 | 0.06 | 0.12 | 0.14 |
† The bandwidths are and . Each entry is based on 100 Gaussian multipliers samples and 100 repetitions.
Table 3
Open in new tab
Empirical sizes of the tests for H10 and H20 at the nominal level 0.05 using the complete cases under missingness completely at random when 50% of the marks are missing†
Model | Missingness model | n | h | Size | |||
---|---|---|---|---|---|---|---|
TestingH10 | |||||||
M1 | (13) | 500 | 0.20 | 0.14 | 0.10 | 0.12 | 0.15 |
800 | 0.15 | 0.10 | 0.07 | 0.11 | 0.11 | ||
(16) | 500 | 0.20 | 0.39 | 0.37 | 0.50 | 0.42 | |
800 | 0.15 | 0.50 | 0.46 | 0.63 | 0.55 | ||
Testing H20 | |||||||
M2 | (13) | 500 | 0.20 | 0.08 | 0.04 | 0.08 | 0.05 |
800 | 0.15 | 0.06 | 0.09 | 0.06 | 0.10 | ||
(16) | 500 | 0.20 | 0.08 | 0.07 | 0.08 | 0.05 | |
800 | 0.15 | 0.07 | 0.06 | 0.12 | 0.14 |
Model | Missingness model | n | h | Size | |||
---|---|---|---|---|---|---|---|
TestingH10 | |||||||
M1 | (13) | 500 | 0.20 | 0.14 | 0.10 | 0.12 | 0.15 |
800 | 0.15 | 0.10 | 0.07 | 0.11 | 0.11 | ||
(16) | 500 | 0.20 | 0.39 | 0.37 | 0.50 | 0.42 | |
800 | 0.15 | 0.50 | 0.46 | 0.63 | 0.55 | ||
Testing H20 | |||||||
M2 | (13) | 500 | 0.20 | 0.08 | 0.04 | 0.08 | 0.05 |
800 | 0.15 | 0.06 | 0.09 | 0.06 | 0.10 | ||
(16) | 500 | 0.20 | 0.08 | 0.07 | 0.08 | 0.05 | |
800 | 0.15 | 0.07 | 0.06 | 0.12 | 0.14 |
† The bandwidths are and . Each entry is based on 100 Gaussian multipliers samples and 100 repetitions.
These simulation results verify that the testing procedures applied to complete cases generally do not have nominal size, although for some of the scenarios the sizes are nominal. To explain this, it can be shown that, under MAR, , where . If hk(t,z) does not depend on z and MAR holds, then the observations for individuals with the observed marks only can be viewed as a random sample from a mark-specific proportional hazards model with a different baseline hazard function but the same regression function β(v). In this case, the tests for both H10 and H20 based on the complete cases are valid. If hk(t,z) depends on z but not on t and MAR holds, then hk(t,z) can be expressed as (the scenario under model (13)), and the tests of H10 based on the complete cases will be biased. However, the tests of H20 remain unbiased since the biases in the estimation of β(v) do not depend on v, such that the test process Q(2)(v) is still asymptotically a mean 0 process. In general, if hk(t,z) depends on both z and t and MAR holds, which is the scenario under the missingness model (16), then the test process Q(2)(v) is not an asymptotically mean 0 process. The magnitude of departure of the asymptotic sizes of the test statistics of H20 from the nominal level depends on hk(t,z) in a complicated manner.
5.2. Numerical assessment of the tests under misspecified models
This subsection evaluates robustness of the proposed test procedures to misspecifications of rk(w) and/or , and to violation of the MAR assumption. The Zki, (Tki, Vki) and Cki are generated by using the same models as above, again with approximately 30% censoring.
Robustness of the tests to misspecification of rk(w) is examined by assuming model (13) whereas the actual complete-case indicator Rki is generated with the conditional probability , where
(17)
This model yields approximately 50% missing marks among observed failures under models M1–M5.
Robustness of the tests is also examined when is misspecified. This is carried out by assuming model (14) for the auxiliary mark, or, equivalently, model (15) for , whereas the actual mark for δki=1 is generated from
(18)
for Here Uki is uniformly distributed on [0,1] and is independent of Vki.
Robustness of the tests to violation of the MAR assumption (2) is examined by assuming model (13), whereas the actual Rki depends on Vki through the model
The proportion of missing marks among the observed failures is kept around 50% in all scenarios.
Models (17), (18) and (19) are similar to those used in Sun and Gilbert (2012) for examining robustness of the AIPW estimator. However, instead of examining biases and standard errors of the estimators, here we check whether the empirical sizes of the tests are close to their nominal level 0.05 and how the powers of the tests are affected by these misspecifications. For sample size n = 500 and bandwidths and , Table 4 shows the empirical sizes and powers of the tests of H10 and Table 5 shows the empirical sizes and powers of the tests of H20. In both Table 4 and Table 5, the first block shows the results when rk(w) is misspecified following model (17) and is correctly specified by model (15) with θ = 0.4; the second block shows the results when is misspecified following model (18) and rk(w) is correctly specified by model (13) with ψk1=0.2 and ψk1=−0.2; the third block shows the results when rk(w) is misspecified following model (17) and is misspecified following model (18); and the fourth block shows the results when rk(w) depends on Vki following model (19) and is correctly specified by model (15) with θ = 0.4.
Table 4
Open in new tab
Robustness of the tests for H10†
Model | (α,β,γ) | Size/power | |||
---|---|---|---|---|---|
rk(w) is misspecified | |||||
M1 | (0,0,0.3) | 4.2 | 5.2 | 3.6 | 4.2 |
M3 | (−0.6,0.6,0.3) | 62.0 | 74.4 | 74.0 | 81.8 |
M4 | (−1.2,1.2,0.3) | 99.6 | 99.8 | 99.8 | 99.8 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
is misspecified | |||||
M1 | (0,0,0.3) | 3.4 | 4.2 | 5.8 | 4.6 |
M3 | (−0.6,0.6,0.3) | 59.6 | 64.4 | 72.8 | 74.4 |
M4 | (−1.2,1.2,0.3) | 99.2 | 99.4 | 99.6 | 99.6 |
M2 | (−0.69,0,0.3) | 100 | 99.8 | 100 | 99.8 |
rk(w) andare misspecified | |||||
M1 | (0,0,0.3) | 4.0 | 4.0 | 3.8 | 3.4 |
M3 | (−0.6,0.6,0.3) | 61.8 | 61.8 | 71.8 | 73.8 |
M4 | (−1.2,1.2,0.3) | 99.6 | 98.6 | 99.8 | 99.8 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
MAR assumption is violated | |||||
M1 | (0,0,0.3) | 3.4 | 3.8 | 3.6 | 5.0 |
M3 | (−0.6,0.6,0.3) | 60.6 | 67.0 | 73.0 | 77.8 |
M4 | (−1.2,1.2,0.3) | 99.2 | 99.6 | 99.8 | 99.6 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
Model | (α,β,γ) | Size/power | |||
---|---|---|---|---|---|
rk(w) is misspecified | |||||
M1 | (0,0,0.3) | 4.2 | 5.2 | 3.6 | 4.2 |
M3 | (−0.6,0.6,0.3) | 62.0 | 74.4 | 74.0 | 81.8 |
M4 | (−1.2,1.2,0.3) | 99.6 | 99.8 | 99.8 | 99.8 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
is misspecified | |||||
M1 | (0,0,0.3) | 3.4 | 4.2 | 5.8 | 4.6 |
M3 | (−0.6,0.6,0.3) | 59.6 | 64.4 | 72.8 | 74.4 |
M4 | (−1.2,1.2,0.3) | 99.2 | 99.4 | 99.6 | 99.6 |
M2 | (−0.69,0,0.3) | 100 | 99.8 | 100 | 99.8 |
rk(w) andare misspecified | |||||
M1 | (0,0,0.3) | 4.0 | 4.0 | 3.8 | 3.4 |
M3 | (−0.6,0.6,0.3) | 61.8 | 61.8 | 71.8 | 73.8 |
M4 | (−1.2,1.2,0.3) | 99.6 | 98.6 | 99.8 | 99.8 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
MAR assumption is violated | |||||
M1 | (0,0,0.3) | 3.4 | 3.8 | 3.6 | 5.0 |
M3 | (−0.6,0.6,0.3) | 60.6 | 67.0 | 73.0 | 77.8 |
M4 | (−1.2,1.2,0.3) | 99.2 | 99.6 | 99.8 | 99.6 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
† Empirical sizes and powers of the tests , , and for testing H10 at the nominal level 0.05 for n = 500 and h = 0.2 when 50% of the marks are missing. The bandwidths are and . Each entry is based on 500 Gaussian multipliers samples and 500 repetitions.
Table 4
Open in new tab
Robustness of the tests for H10†
Model | (α,β,γ) | Size/power | |||
---|---|---|---|---|---|
rk(w) is misspecified | |||||
M1 | (0,0,0.3) | 4.2 | 5.2 | 3.6 | 4.2 |
M3 | (−0.6,0.6,0.3) | 62.0 | 74.4 | 74.0 | 81.8 |
M4 | (−1.2,1.2,0.3) | 99.6 | 99.8 | 99.8 | 99.8 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
is misspecified | |||||
M1 | (0,0,0.3) | 3.4 | 4.2 | 5.8 | 4.6 |
M3 | (−0.6,0.6,0.3) | 59.6 | 64.4 | 72.8 | 74.4 |
M4 | (−1.2,1.2,0.3) | 99.2 | 99.4 | 99.6 | 99.6 |
M2 | (−0.69,0,0.3) | 100 | 99.8 | 100 | 99.8 |
rk(w) andare misspecified | |||||
M1 | (0,0,0.3) | 4.0 | 4.0 | 3.8 | 3.4 |
M3 | (−0.6,0.6,0.3) | 61.8 | 61.8 | 71.8 | 73.8 |
M4 | (−1.2,1.2,0.3) | 99.6 | 98.6 | 99.8 | 99.8 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
MAR assumption is violated | |||||
M1 | (0,0,0.3) | 3.4 | 3.8 | 3.6 | 5.0 |
M3 | (−0.6,0.6,0.3) | 60.6 | 67.0 | 73.0 | 77.8 |
M4 | (−1.2,1.2,0.3) | 99.2 | 99.6 | 99.8 | 99.6 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
Model | (α,β,γ) | Size/power | |||
---|---|---|---|---|---|
rk(w) is misspecified | |||||
M1 | (0,0,0.3) | 4.2 | 5.2 | 3.6 | 4.2 |
M3 | (−0.6,0.6,0.3) | 62.0 | 74.4 | 74.0 | 81.8 |
M4 | (−1.2,1.2,0.3) | 99.6 | 99.8 | 99.8 | 99.8 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
is misspecified | |||||
M1 | (0,0,0.3) | 3.4 | 4.2 | 5.8 | 4.6 |
M3 | (−0.6,0.6,0.3) | 59.6 | 64.4 | 72.8 | 74.4 |
M4 | (−1.2,1.2,0.3) | 99.2 | 99.4 | 99.6 | 99.6 |
M2 | (−0.69,0,0.3) | 100 | 99.8 | 100 | 99.8 |
rk(w) andare misspecified | |||||
M1 | (0,0,0.3) | 4.0 | 4.0 | 3.8 | 3.4 |
M3 | (−0.6,0.6,0.3) | 61.8 | 61.8 | 71.8 | 73.8 |
M4 | (−1.2,1.2,0.3) | 99.6 | 98.6 | 99.8 | 99.8 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
MAR assumption is violated | |||||
M1 | (0,0,0.3) | 3.4 | 3.8 | 3.6 | 5.0 |
M3 | (−0.6,0.6,0.3) | 60.6 | 67.0 | 73.0 | 77.8 |
M4 | (−1.2,1.2,0.3) | 99.2 | 99.6 | 99.8 | 99.6 |
M2 | (−0.69,0,0.3) | 100 | 100 | 100 | 100 |
† Empirical sizes and powers of the tests , , and for testing H10 at the nominal level 0.05 for n = 500 and h = 0.2 when 50% of the marks are missing. The bandwidths are and . Each entry is based on 500 Gaussian multipliers samples and 500 repetitions.
Table 5
Open in new tab
Robustness of the tests for H20†
Model | (α,β,γ) | Size/power | |||
---|---|---|---|---|---|
rk(w) is misspecified | |||||
M2 | (−0.69,0,0.3) | 5.0 | 3.8 | 5.4 | 6.2 |
M3 | (−0.6,0.6,0.3) | 24.0 | 25.2 | 34.2 | 36.0 |
M4 | (−1.2,1.2,0.3) | 60.8 | 66.6 | 72.8 | 78.4 |
M5 | (−1.5,1.5,0.3) | 76.6 | 82.0 | 85.8 | 88.8 |
is misspecified | |||||
M2 | (−0.69,0,0.3) | 4.8 | 6.6 | 6.0 | 5.8 |
M3 | (−0.6,0.6,0.3) | 17.2 | 18.0 | 28.4 | 28.2 |
M4 | (−1.2,1.2,0.3) | 44.8 | 47.2 | 56.4 | 61.0 |
M5 | (−1.5,1.5,0.3) | 58.0 | 60.4 | 68.6 | 73.2 |
rk(w) andare misspecified | |||||
M2 | (−0.69,0,0.3) | 4.0 | 4.8 | 4.4 | 4.4 |
M3 | (−0.6,0.6,0.3) | 16.6 | 19.6 | 26.8 | 26.6 |
M4 | (−1.2,1.2,0.3) | 43.2 | 46.6 | 55.6 | 60.6 |
M5 | (−1.5,1.5,0.3) | 53.8 | 58.8 | 67.4 | 71.4 |
MAR assumption is violated | |||||
M2 | (−0.69,0,0.3) | 6.8 | 6.0 | 7.6 | 7.8 |
M3 | (−0.6,0.6,0.3) | 28.6 | 33.6 | 39.6 | 42.0 |
M4 | (−1.2,1.2,0.3) | 61.8 | 67.0 | 74.0 | 78.4 |
M5 | (−1.5,1.5,0.3) | 77.4 | 81.6 | 85.4 | 89.2 |
Model | (α,β,γ) | Size/power | |||
---|---|---|---|---|---|
rk(w) is misspecified | |||||
M2 | (−0.69,0,0.3) | 5.0 | 3.8 | 5.4 | 6.2 |
M3 | (−0.6,0.6,0.3) | 24.0 | 25.2 | 34.2 | 36.0 |
M4 | (−1.2,1.2,0.3) | 60.8 | 66.6 | 72.8 | 78.4 |
M5 | (−1.5,1.5,0.3) | 76.6 | 82.0 | 85.8 | 88.8 |
is misspecified | |||||
M2 | (−0.69,0,0.3) | 4.8 | 6.6 | 6.0 | 5.8 |
M3 | (−0.6,0.6,0.3) | 17.2 | 18.0 | 28.4 | 28.2 |
M4 | (−1.2,1.2,0.3) | 44.8 | 47.2 | 56.4 | 61.0 |
M5 | (−1.5,1.5,0.3) | 58.0 | 60.4 | 68.6 | 73.2 |
rk(w) andare misspecified | |||||
M2 | (−0.69,0,0.3) | 4.0 | 4.8 | 4.4 | 4.4 |
M3 | (−0.6,0.6,0.3) | 16.6 | 19.6 | 26.8 | 26.6 |
M4 | (−1.2,1.2,0.3) | 43.2 | 46.6 | 55.6 | 60.6 |
M5 | (−1.5,1.5,0.3) | 53.8 | 58.8 | 67.4 | 71.4 |
MAR assumption is violated | |||||
M2 | (−0.69,0,0.3) | 6.8 | 6.0 | 7.6 | 7.8 |
M3 | (−0.6,0.6,0.3) | 28.6 | 33.6 | 39.6 | 42.0 |
M4 | (−1.2,1.2,0.3) | 61.8 | 67.0 | 74.0 | 78.4 |
M5 | (−1.5,1.5,0.3) | 77.4 | 81.6 | 85.4 | 89.2 |
† Empirical sizes and powers of the tests , , and for testing H20 at the nominal level 0.05 for n = 500 and h = 0.2 when 50% of the marks are missing. The bandwidths are and . Each entry is based on 500 Gaussian multipliers samples and 500 repetitions.
Table 5
Open in new tab
Robustness of the tests for H20†
Model | (α,β,γ) | Size/power | |||
---|---|---|---|---|---|
rk(w) is misspecified | |||||
M2 | (−0.69,0,0.3) | 5.0 | 3.8 | 5.4 | 6.2 |
M3 | (−0.6,0.6,0.3) | 24.0 | 25.2 | 34.2 | 36.0 |
M4 | (−1.2,1.2,0.3) | 60.8 | 66.6 | 72.8 | 78.4 |
M5 | (−1.5,1.5,0.3) | 76.6 | 82.0 | 85.8 | 88.8 |
is misspecified | |||||
M2 | (−0.69,0,0.3) | 4.8 | 6.6 | 6.0 | 5.8 |
M3 | (−0.6,0.6,0.3) | 17.2 | 18.0 | 28.4 | 28.2 |
M4 | (−1.2,1.2,0.3) | 44.8 | 47.2 | 56.4 | 61.0 |
M5 | (−1.5,1.5,0.3) | 58.0 | 60.4 | 68.6 | 73.2 |
rk(w) andare misspecified | |||||
M2 | (−0.69,0,0.3) | 4.0 | 4.8 | 4.4 | 4.4 |
M3 | (−0.6,0.6,0.3) | 16.6 | 19.6 | 26.8 | 26.6 |
M4 | (−1.2,1.2,0.3) | 43.2 | 46.6 | 55.6 | 60.6 |
M5 | (−1.5,1.5,0.3) | 53.8 | 58.8 | 67.4 | 71.4 |
MAR assumption is violated | |||||
M2 | (−0.69,0,0.3) | 6.8 | 6.0 | 7.6 | 7.8 |
M3 | (−0.6,0.6,0.3) | 28.6 | 33.6 | 39.6 | 42.0 |
M4 | (−1.2,1.2,0.3) | 61.8 | 67.0 | 74.0 | 78.4 |
M5 | (−1.5,1.5,0.3) | 77.4 | 81.6 | 85.4 | 89.2 |
Model | (α,β,γ) | Size/power | |||
---|---|---|---|---|---|
rk(w) is misspecified | |||||
M2 | (−0.69,0,0.3) | 5.0 | 3.8 | 5.4 | 6.2 |
M3 | (−0.6,0.6,0.3) | 24.0 | 25.2 | 34.2 | 36.0 |
M4 | (−1.2,1.2,0.3) | 60.8 | 66.6 | 72.8 | 78.4 |
M5 | (−1.5,1.5,0.3) | 76.6 | 82.0 | 85.8 | 88.8 |
is misspecified | |||||
M2 | (−0.69,0,0.3) | 4.8 | 6.6 | 6.0 | 5.8 |
M3 | (−0.6,0.6,0.3) | 17.2 | 18.0 | 28.4 | 28.2 |
M4 | (−1.2,1.2,0.3) | 44.8 | 47.2 | 56.4 | 61.0 |
M5 | (−1.5,1.5,0.3) | 58.0 | 60.4 | 68.6 | 73.2 |
rk(w) andare misspecified | |||||
M2 | (−0.69,0,0.3) | 4.0 | 4.8 | 4.4 | 4.4 |
M3 | (−0.6,0.6,0.3) | 16.6 | 19.6 | 26.8 | 26.6 |
M4 | (−1.2,1.2,0.3) | 43.2 | 46.6 | 55.6 | 60.6 |
M5 | (−1.5,1.5,0.3) | 53.8 | 58.8 | 67.4 | 71.4 |
MAR assumption is violated | |||||
M2 | (−0.69,0,0.3) | 6.8 | 6.0 | 7.6 | 7.8 |
M3 | (−0.6,0.6,0.3) | 28.6 | 33.6 | 39.6 | 42.0 |
M4 | (−1.2,1.2,0.3) | 61.8 | 67.0 | 74.0 | 78.4 |
M5 | (−1.5,1.5,0.3) | 77.4 | 81.6 | 85.4 | 89.2 |
† Empirical sizes and powers of the tests , , and for testing H20 at the nominal level 0.05 for n = 500 and h = 0.2 when 50% of the marks are missing. The bandwidths are and . Each entry is based on 500 Gaussian multipliers samples and 500 repetitions.
Tables 4 and 5 show that the empirical sizes of the tests are very close to the nominal level 0.05 when one of rk(w) and is misspecified, reflecting the double-robustness property of the AIPW estimator. The empirical sizes are also close to 0.05 when both rk(w) and are misspecified and when the MAR assumption is violated, which is intriguing. When only rk(w) is misspecified and MAR holds, the empirical powers in Tables 4 and 5 closely track the corresponding powers in Tables 1 and 2 under correct model specifications. The empirical powers are lower than those observed in Table 1 and 2 when is misspecified or when both rk(w) and are misspecified, whereas the empirical powers in Tables 4 and 5 are very close to those in Tables 1 and 2 when MAR is violated. Apparently for our particular data simulation, the bias due to the MAR violation counterbalances the bias due to misspecification of both rk(w) and ; however, in general these violations could distort sizes and powers.
5.3. Simulation study for the Thai trial
We conduct a simulation of the Thai trial, to gain insight about the power that is available for this real trial. Specifically, we simulated data to yield about the numbers of infections observed (74 in the placebo group and 51 in the vaccine group), the overall vaccine efficacy from the proportional hazards model is about 31%, and the true VE(v) curve decreases with v to be around 65–70% for v close to 0 and around 0% for v close to 1. The actual infection rate was only 0.3% over 3.5 years; to speed the simulations we use a 20% placebo infection rate and retain 74 infections on average.
Again with K = 1 stratum, the (Tki, Vki) are generated from the model
(20)
where α, β and γ are constants. Under model (20), VE(v)=1− exp (α+βv), the marginal hazards are λ0(t)=γ for z = 0 and λ1(t)=γ exp (α){ exp (β)−1}/β for z = 1, and the Cox proportional hazards vaccine efficacy is VEC=1−λ1(t)/λ0(t)=1− exp (α){ exp (β)−1}/β. We choose (α,β,γ)=(−1.1,1.3,0.068), yielding VEC=0.32, VE(0)=0.67 and VE(0.85)=0. We study 400 subjects each in the vaccine and placebo groups. Matching the actual trial, the censoring rate before τ is kept very low, just under 5%. The missing mark indicator is generated from model (13), with (ψk0, ψk1) set to yield about 0%, 25% (−1.2, −0.2), 50% (0.2, −0.2) and 75% (−1.0, −0.2) missing marks among observed failures. We assume that the auxiliary variable Aki follows model (14) given in Section 5.1, where the θ-values of ∞, 0.8, 0.4 and 0.2 correspond to ρ = 0, 0.78, 0.92, 0.98 for the correlation coefficient between Aki and Vki.
Because of lost information on the mark, we choose larger bandwidths for higher percentages of missing marks. We use h = 0.4 for the case with 75% missing marks, h = 0.3 for the case with 50% missing marks, h = 0.2 for the case with 25% missing marks and h = 0.15 for the case with 0% missing marks. The bandwidths and in equation (7) in the estimation of are taken to be 0.50 and in each case. Powers of the proposed tests , , , , , , and for the simulations based on the Thai trial at the nominal level 0.05 are reported in Table 6. The tests show similar performance as was found in the simulation study of Section 5.1. As only 10% of infected subjects had missing marks in the RV144 trial and the auxiliary was very weakly predictive, we focus on the entries with 0% or 25% missing marks and ρ = 0. There is 67–95% power to reject H10, and 33–60% power to reject H20. These results show that a fairly strong sieve effect with VE(v) declining from 67% to 0% could readily be missed in the Thai trial because of limited power. The only slightly improved power with an excellent auxiliary ρ = 0.98 shows that greater numbers of events would be needed to achieve high power for testing H20.
Table 6
Open in new tab
Power of the tests , , , , , , and for the Thai trial at the nominal level 0.05†
ρ | % missing marks | h | Power | |||||||
---|---|---|---|---|---|---|---|---|---|---|
TestingH10 | TestingH20 | |||||||||
0 | 0.15 | 77 | 85 | 86 | 95 | 48 | 48 | 59 | 60 | |
0 | 25 | 0.2 | 67 | 76 | 79 | 85 | 36 | 33 | 50 | 47 |
50 | 0.3 | 63 | 71 | 71 | 82 | 29 | 27 | 37 | 42 | |
75 | 0.4 | 41 | 51 | 59 | 58 | 21 | 18 | 35 | 31 | |
0.78 | 25 | 0.2 | 67 | 79 | 82 | 89 | 36 | 39 | 46 | 50 |
50 | 0.3 | 60 | 71 | 74 | 84 | 28 | 28 | 41 | 39 | |
75 | 0.4 | 49 | 53 | 63 | 65 | 25 | 25 | 34 | 34 | |
0.92 | 25 | 0.2 | 70 | 80 | 84 | 91 | 37 | 41 | 50 | 56 |
50 | 0.3 | 61 | 71 | 73 | 87 | 35 | 39 | 50 | 51 | |
75 | 0.4 | 54 | 58 | 62 | 71 | 30 | 33 | 40 | 44 | |
0.98 | 25 | 0.2 | 71 | 81 | 82 | 91 | 39 | 47 | 53 | 55 |
50 | 0.3 | 66 | 76 | 75 | 86 | 44 | 42 | 50 | 52 | |
75 | 0.4 | 56 | 66 | 68 | 76 | 41 | 43 | 51 | 49 |
ρ | % missing marks | h | Power | |||||||
---|---|---|---|---|---|---|---|---|---|---|
TestingH10 | TestingH20 | |||||||||
0 | 0.15 | 77 | 85 | 86 | 95 | 48 | 48 | 59 | 60 | |
0 | 25 | 0.2 | 67 | 76 | 79 | 85 | 36 | 33 | 50 | 47 |
50 | 0.3 | 63 | 71 | 71 | 82 | 29 | 27 | 37 | 42 | |
75 | 0.4 | 41 | 51 | 59 | 58 | 21 | 18 | 35 | 31 | |
0.78 | 25 | 0.2 | 67 | 79 | 82 | 89 | 36 | 39 | 46 | 50 |
50 | 0.3 | 60 | 71 | 74 | 84 | 28 | 28 | 41 | 39 | |
75 | 0.4 | 49 | 53 | 63 | 65 | 25 | 25 | 34 | 34 | |
0.92 | 25 | 0.2 | 70 | 80 | 84 | 91 | 37 | 41 | 50 | 56 |
50 | 0.3 | 61 | 71 | 73 | 87 | 35 | 39 | 50 | 51 | |
75 | 0.4 | 54 | 58 | 62 | 71 | 30 | 33 | 40 | 44 | |
0.98 | 25 | 0.2 | 71 | 81 | 82 | 91 | 39 | 47 | 53 | 55 |
50 | 0.3 | 66 | 76 | 75 | 86 | 44 | 42 | 50 | 52 | |
75 | 0.4 | 56 | 66 | 68 | 76 | 41 | 43 | 51 | 49 |
† Each entry is based on 100 Gaussian multipliers samples and 100 repetitions.
Table 6
Open in new tab
Power of the tests , , , , , , and for the Thai trial at the nominal level 0.05†
ρ | % missing marks | h | Power | |||||||
---|---|---|---|---|---|---|---|---|---|---|
TestingH10 | TestingH20 | |||||||||
0 | 0.15 | 77 | 85 | 86 | 95 | 48 | 48 | 59 | 60 | |
0 | 25 | 0.2 | 67 | 76 | 79 | 85 | 36 | 33 | 50 | 47 |
50 | 0.3 | 63 | 71 | 71 | 82 | 29 | 27 | 37 | 42 | |
75 | 0.4 | 41 | 51 | 59 | 58 | 21 | 18 | 35 | 31 | |
0.78 | 25 | 0.2 | 67 | 79 | 82 | 89 | 36 | 39 | 46 | 50 |
50 | 0.3 | 60 | 71 | 74 | 84 | 28 | 28 | 41 | 39 | |
75 | 0.4 | 49 | 53 | 63 | 65 | 25 | 25 | 34 | 34 | |
0.92 | 25 | 0.2 | 70 | 80 | 84 | 91 | 37 | 41 | 50 | 56 |
50 | 0.3 | 61 | 71 | 73 | 87 | 35 | 39 | 50 | 51 | |
75 | 0.4 | 54 | 58 | 62 | 71 | 30 | 33 | 40 | 44 | |
0.98 | 25 | 0.2 | 71 | 81 | 82 | 91 | 39 | 47 | 53 | 55 |
50 | 0.3 | 66 | 76 | 75 | 86 | 44 | 42 | 50 | 52 | |
75 | 0.4 | 56 | 66 | 68 | 76 | 41 | 43 | 51 | 49 |
ρ | % missing marks | h | Power | |||||||
---|---|---|---|---|---|---|---|---|---|---|
TestingH10 | TestingH20 | |||||||||
0 | 0.15 | 77 | 85 | 86 | 95 | 48 | 48 | 59 | 60 | |
0 | 25 | 0.2 | 67 | 76 | 79 | 85 | 36 | 33 | 50 | 47 |
50 | 0.3 | 63 | 71 | 71 | 82 | 29 | 27 | 37 | 42 | |
75 | 0.4 | 41 | 51 | 59 | 58 | 21 | 18 | 35 | 31 | |
0.78 | 25 | 0.2 | 67 | 79 | 82 | 89 | 36 | 39 | 46 | 50 |
50 | 0.3 | 60 | 71 | 74 | 84 | 28 | 28 | 41 | 39 | |
75 | 0.4 | 49 | 53 | 63 | 65 | 25 | 25 | 34 | 34 | |
0.92 | 25 | 0.2 | 70 | 80 | 84 | 91 | 37 | 41 | 50 | 56 |
50 | 0.3 | 61 | 71 | 73 | 87 | 35 | 39 | 50 | 51 | |
75 | 0.4 | 54 | 58 | 62 | 71 | 30 | 33 | 40 | 44 | |
0.98 | 25 | 0.2 | 71 | 81 | 82 | 91 | 39 | 47 | 53 | 55 |
50 | 0.3 | 66 | 76 | 75 | 86 | 44 | 42 | 50 | 52 | |
75 | 0.4 | 56 | 66 | 68 | 76 | 41 | 43 | 51 | 49 |
† Each entry is based on 100 Gaussian multipliers samples and 100 repetitions.
6. Analysis of the RV144 Thai trial
In the RV144 Thai trial, 125 subjects (51 of 8197 in the vaccine group and 74 of 8198 in the placebo group) were diagnosed with HIV infection over a 42-month follow-up period, from whom full length HIV genomes were measured from 121; three missed data because their HIV viral load was too low for the Sanger sequencing technology to work, and one dropped out (Rerks-Ngarm et al., 2009; Rolland et al., 2012). We focus on the gp120 region of the HIV Env protein, because this region stimulates anti-HIV antibody responses which are the putative cause of the observed partial vaccine efficacy. Three gp120 sequences were included in the vaccine: 92TH023 in the ALVAC canarypox vector prime component, and CM244, MN in the AIDSVAX gp120 protein boost component. 92TH023 and CM244 are subtype E HIVs whereas MN is subtype B, and 110 of the 121 subjects were infected with subtype E sequences. The subtype E vaccine insert sequences are much closer genetically to the infecting (and regional circulating) sequences than MN and thus are more likely to stimulate protective immune responses. Accordingly, the analysis focuses on the 92TH023 and CM244 reference sequences, and right-censors the 15 subjects who were HIV infected with subtype B or with unknown subtype. One subject who acquired HIV infection during the trial was documented to have acquired HIV from another trial participant who had previously become HIV infected; the analysis excludes this subject because his or her inclusion would violate the independent observations assumption. In the context of our model set-up, T is the time to diagnosis of HIV infection with subtype E HIV. The time to diagnosis of HIV infection with subtype B or with unknown HIV subtype is treated as censoring.
We define V based on HIV sequence data measured from a blood sample drawn at or before the date of HIV diagnosis. (The trial documented acute phase or preseroconversion infection in only a few subjects, prohibiting defining the mark based on acute phase sequences.) 11 of the 109 (11%) infected subjects have sequences measured from a post-diagnosis sample and hence are missing V. To maximize biological relevance and statistical power, we restrict the gp120 distances to the published set of gp120 sites in contact with known broadly neutralizing monoclonal antibodies (Moore et al., 2009; Wei et al., 2003). For each HIV sequence from a subject and each of the two reference vaccine sequences, V is computed as a weighted Hamming distance by using the point accepted mutation between scoring matrix (Nickle et al., 2007). Between two and 13 sequences (a total of 1030 sequences) were measured per infected subject, and V is defined as the subject's sequence closest to his or her consensus sequence (the consensus sequence is comprised of the majority amino acids at each site, one site at a time). Finally, the distances are rescaled to values between 0 and 1.
In total, 109 infected subjects (43 vaccine; 66 placebo) are included in the analysis, of which 98 (39 vaccine; 59 placebo) have an observed mark V; Fig. 1 displays the observed Vs.
Fig. 1.
Scatter plots of the marks V versus the HIV-infection time T for the 98 HIV-infected subjects in the Thai trial with an observed mark: the mark V is the HIV-specific point accepted mutation matrix (Nickle et al., 2007) weighted Hamming distances between a subject's HIV envelope gp120 amino acid sequence (nearest to his or her consensus sequence) and (a) the 92TH023 or (b) CM244 vaccine reference sequence; the distances restrict to the 172 amino acid sites in gp120 documented to contact broadly neutralizing monoclonal antibodies (, vaccine LOWESS smooth fit (Cleveland, 1979);
, placebo LOWESS smooth fit)
Open in new tabDownload slide
To predict the probability of observing V among the 109 infected subjects, we use all-subsets logistic regression model selection considering demographics, host genetics and biomarker data post infection. The best model by the Bayesian information criterion BIC includes only the years from entry until diagnosis of HIV infection (X1), with model fit logit for the CM244 reference sequence. The model was very similar for the 92TH023 reference sequence (which is not shown). In addition, we consider linear and logistic regression models for relating the mean of various potential auxiliary variables (A) to V, X1 and treatment indicator Z. Model selection did not reveal any significantly predictive auxiliary variables; we expect that HIV sequence information measured after V has been defined would be a good predictor, but these data were not collected. Nevertheless, to implement the AIPW method we select the best available auxiliary variable, gender (A = X2; 1≡male; 0≡female) and use the logistic regression model that results; for CM244 the fitted model is logit, and the model was very similar for 92TH023 (which is not shown).
The AIPW estimation and testing procedures are applied to the Thai trial data set with bandwidths and , a = 0.05, b = 1 and (a and are near the minimum observed marks). As in the simulation study, 500 simulated Gaussian multipliers are used. Because the results are nearly identical with and without the auxiliary variable, only the latter results are presented. Fig. 2 shows the estimated VE(v) along with 95% pointwise confidence bands, indicating that vaccine efficacy appears to be high against HIVs near the 92TH023 reference sequence (estimated VE(0.01) = 56%) and declines to 0 against HIVs farthest from the 92TH023 reference sequence (estimated VE(1.0) = 2.4%). The decline is similar for the CM244 reference sequence, with estimated VE(0.01) = 45% and estimated VE(0.95) = −9.1%.
Fig. 2.
AIPW estimation of VE(v) and 95% pointwise confidence bands without using auxiliary variables for the Thai trial with bandwidths and for the monoclonal antibody contact site distances to (a) the 92TH023 and (b) CM244 reference sequences
Open in new tabDownload slide
Figs 3(a) and 3(b) show the test processes Q(1)(v) versus 20 realizations from the Gaussian multiplier process given the observed data, and Figs 3(c) and 3(d) show the parallel results for the test process Q(2)(v), each suggesting departures from the null hypothesis H10 and from the null hypothesis H20 for each reference sequence. The p-values of the tests based on the test statistics and for testing H10 against the monotone alternative over v ∈ [0,1] are 0.032 and 0.008 for 92TH023, and 0.014 and 0.010 for CM244. The p-values of the test statistics and for testing H10 against the general alternative are 0.054 and 0.018 for 92TH023 and 0.030 and 0.010 for CM244. For testing H20 over v ∈ [0,1], the p-values of the supremum-type tests based on the test statistics and are 0.53 and 0.27 for 92TH023 and 0.37 and 0.18 for CM244. The p-values of the integrated square type of tests based on the test statistics and are 0.35 and 0.14 for 92TH023 and 0.44 and 0.19 for CM244.
Fig. 3.
Diagnostic plots of the test processes for the Thai trial data set with bandwidths , and a = 0.05, b = 1 and : (a) Q(1)(v) () versus 20 realizations (
) from the Gaussian multiplier process (92TH023, CM244 reference; H10); (b) Q(1)(v) (
) versus 20 realizations (
) from the Gaussian multiplier process (92TH023, CM244 reference; H20); (c) Q(2)(v) (
) versus 20 realizations (
) from the Gaussian multiplier process (92TH023, CM244 reference; H10): (d) Q(2)(v) (
) versus 20 realizations (
) from the Gaussian multiplier process (92TH023, CM244 reference, H20)
Open in new tabDownload slide
These analyses provide more evidence that the vaccine had some protective efficacy than the original primary analysis that did not account for the mark information (Rerks-Ngarm et al., 2009): the primary analysis test for any vaccine efficacy yielded p = 0.04 whereas the tests for any vaccine efficacy against any mark reported here yielded a median p-value of 0.016 across the four test statistics and two reference sequences. The analyses also showed a non-significant trend (p-values around 0.14–0.19) that the vaccine protected better against HIVs closely matched to the vaccine strain HIVs in the monoclonal antibody contact sites but had less or absent protection against HIVs with many mismatches in these sites. Although the levels of significance are not compelling, the simulation study presented in Section 5.5.3 of the power available for detecting a vaccine sieve effect in the Thai trial showed that the study is well powered only to detect large sieve effects (with greater decline of VE(v) in v than was observed in the estimated VE(v) curves); thus a moderate-to-large sieve effect is consistent with the observed results. These results may guide future vaccine research by suggesting modifications of future vaccine candidates to include HIV sequences more closely matched to circulating HIVs in the monoclonal antibody contact sites. They may also motivate the design of future experiments to understand functional effects of amino acid mutations at the monoclonal antibody contact sites.
Acknowledgements
The authors thank Hasan Ahmed and Paul Edlefsen for generating the HIV sequence distances, and they thank the participants, investigators and sponsors of the RV144 Thai trial, including the US Military HIV Research Program, US Army Medical Research and Materiel Command, National Institute of Allergy and Infectious Diseases US and Thai Components, Armed Forces Research Institute of Medical Science Ministry of Public Health, Thailand, Mahidol University, SanofiPasteur and Global Solutions for Infectious Diseases. The authors thank the Joint Editor, Associate Editor and two referees for their helpful suggestions. The research of Yanqing Sun was partially supported by National Science Foundation grants DMS-0905777 and DMS-1208978, and the research of Dr Sun and Dr Peter Gilbert was partially supported by National Institutes of Health National Institute of Allergy and Infectious Diseases grant R37AI054165. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
References
Aalen O. O. Johansen S.
1978
)
An empirical transition matrix for non-hom*ogeneous Markov chains based on censored observations
.
Scand. J. Statist.
,
5
,
141
–
150
.
OpenURL Placeholder Text
Cai Z. Fan J. Li R.
2000
)
Efficient estimation and inferences for varying-coefficient models
.
J. Am. Statist. Ass.
,
95
,
888
–
902
.
Cleveland W. S.
1979
)
Robust locally weighted regression and smoothing scatterplots
.
J. Am. Statist. Ass.
,
74
,
829
–
836
.
Efron B. Tibshirani R. J.
1993
)
An Introduction to the Bootstrap
.
New York
:
Chapman and Hall
.
Fauci A. S. Johnston M. I. Dieffenbach C. W. Burton D. R. Hammer S. M. Hoxie J. A. Martin M. Overbaugh J. Watkins D. I. Mahmoud A. Greene W. C.
2008
)
HIV vaccine research: the way forward
.
Science
,
321
,
530
–
532
.
Gilbert P. B. Berger J. O. Stablein D. Becker S. Essex M. Hammer S. M. Kim J. H. Degruttola V. G.
2011
)
Statistical interpretation of the RV144 HIV vaccine efficacy trial in Thailand: a case study for statistical issues in efficacy trials
.
J. Infect. Dis.
,
203
,
969
–
975
.
Gilbert P. B. Lele S. Vardi Y.
1999
)
Maximum likelihood estimation in semiparametric selection bias models with application to AIDS vaccine trials
.
Biometrika
,
86
,
27
–
43
.
Gilbert P. B. McKeague I. W. Sun Y.
2008
)
The two-sample problem for failure rates depending on a continuous mark: an application to vaccine efficacy
.
Biostatistics
,
9
,
263
–
276
.
Gilbert P. B. Self S. G. Ashby M. A.
1998
)
Statistical methods for assessing differential vaccine protection against human immunodeficiency virus types
.
Biometrics
,
54
,
799
–
814
.
Hoover D. R. Rice J. A. Wu C. O. Yang P.-L.
1998
)
Nonparametric smoothing estimates of time-varying coefficient models with longitudinal data
.
Biometrika
,
85
,
809
–
822
.
Lin D. Y. Wei L. J. Ying Z.
1993
)
Checking the Cox model with cumulative sums of martingale-based residuals
.
Biometrika
,
80
,
557
–
572
.
Moore P. L. Ranchobe N. Lambson B. E. Gray E. S. Cave E. Abrahams M.-R. Bandawe G. Mlisana K. Abdool Karim S. S. Williamson C. Morris L.
the CAPRISA 002 Study and the NIAID Center for HIV/AIDS Vaccine Immunology (CHAVI)
(
2009
)
Limited neutralizing antibody specificities drive neutralization escape in early HIV-1 subtype C infection
.
PLOS Path.
,
5
, article e1000598.
OpenURL Placeholder Text
Nickle D. C. Heath L. Jensen M. A. Gilbert P. B. Mullins J. I. Kosakovsky Pond S. L.
2007
)
HIV-specific probabilistic models of protein evolution
.
PLOS ONE
,
2
, no. 6, article e503.
OpenURL Placeholder Text
Prentice R. L. Kalbfleisch J. D. Peterson, Jr A. V. Flournoy N. Farewell V. T. Breslow N. E.
1978
)
The analysis of failure times in the presence of competing risks
.
Biometrics
,
34
,
541
–
554
.
Rerks-Ngarm S. Pitisuttithum P. Nitayaphan S. Kaewkungwal J. Chiu J. Paris R. Premsri N. Namwat C. de Souza M. Adams E. Benenson M. Gurunathan S. Tartaglia J. McNeil J. G. Francis D. P. Stablein D. Birx D. L. Chunsuttiwat S. Khamboonruang C. Thongcharoen P. Robb M. L. Michael N. L. Kunasol P. Kim J. H.
2009
)
Vaccination with ALVAC and AIDSVAX to prevent HIV-1 infection in Thailand
.
New Engl. J. Med.
,
361
,
2209
–
2220
.
Robins J. M. Rotnitzky A. Zhao L. P.
1994
)
Estimation of regression coefficients when some regressors are not always observed
.
J. Am. Statist. Ass.
,
89
,
846
–
866
.
Rolland M. Edlefsen P. T. Larsen B. B. Tovanabutra S. Sanders-Buell E. Hertz T. deCamp A. C. Carrico C. Menis S. Magaret C. A. Ahmed H. Juraska M. Chen L. Konopa P. Nariya S. Stoddard J. N. Wong K. Zhao H. Deng W. Maust B. S. Bose M. Howell S. Bates A. Lazzaro M. O’Sullivan A. Lei E. Bradfield A. Ibitamuno G. Assawadarachai V. O’Connell R. J. deSouza M. S. Nitayaphan S. Rerks-Ngarm S. Robb M. L. McLellan J. S. Georgiev I. Kwong P. D. Carlson J. M. Michael N. L. Schief W. R. Gilbert P. B. Mullins J. I. Kim J. H.
2012
)
Increased HIV-1 vaccine efficacy against viruses with genetic signatures in Env V2
.
Nature
,
490
,
417
–
420
.
Rolland M. Tovanabutra S. Decamp A. C. Frahm N. Gilbert P. B. Sanders-Buell E. Heath L. Magaret C. A. Bose M. Bradfield A. O’Sullivan A. Crossler J. Jones T. Nau M. Wong K. Zhao H. Raugi D. N. Sorensen S. Stoddard J. N. Maust B. S. Deng W. Hural J. Dubey S. Michael N. L. Shiver J. Corey L. Li F. Self S. G. Kim J. Buchbinder S. Casimiro D. R. Robertson M. N. Duerr A. McElrath M. J. McCutchan F. E. Mullins J. I.
2011
)
Genetic impact of vaccination on breakthrough HIV-1 sequences from the STEP trial
.
Nat. Med.
,
17
,
366
–
371
.
Rubin D. B.
1976
)
Inference and missing data
.
Biometrika
,
63
,
581
–
592
.
Stephens M. A.
1974
)
Edf statistics for goodness of fit and some comparisons
.
J. Am. Statist. Ass.
,
69
,
730
–
737
.
Sun Y. Gilbert P. B.
2012
)
Estimation of stratified mark-specific proportional hazards models with missing marks
.
Scand. J. Statist.
,
39
,
34
–
52
.
Sun Y. Gilbert P. B. McKeague I. W.
2009
)
Proportional hazards models with continuous marks
.
Ann. Statist.
,
37
,
394
–
426
.
Sun Y. Wu H.
2005
)
Semiparametric time-varying coefficients regression model for longitudinal data
.
Scand. J. Statist.
,
32
,
21
–
47
.
Tian L. Zucker D. Wei L. J.
2005
)
On the Cox model with time-varying regression coefficients
.
J. Am. Statist. Ass.
,
100
,
172
–
183
.
Wei X. Decker J. M. Wang S. Hui H. Kappes J. C. Wu X. Salazar-Gonzalez J. F. Salazar M. G. Kilby J. M. Saag M. S. Komarova N. L. Nowak M. A. Hahn B. Kwong P. D. Shaw G. M.
2003
)
Antibody neutralization and escape by HIV-1
.
Nature
,
422
,
307
–
312
.
Appendix A: symptotic results
The following regularity conditions from Sun and Gilbert (2012) are assumed.
Condition 1.
- (a)
β(v) has componentwise continuous second derivatives on [0,1]. For each k = 1,…,K, the second partial derivative of λ0k(t,v) with respect to v exists and is continuous on [0,τ]×[0,1]. The covariate process Zk(t) has paths that are left continuous and of bounded variation, and satisfies the moment condition
where M is a constant such that (v,β(v)) ∈ [0,1]×(−M,M)p for all v and for a matrix A = (akl).
- (b)
Each component of is continuous on [0,τ]×[−M,M]p, is continuous on [0,τ]×[−M,M]p×[−L,L]q for some M,L>0 and j = 0,1,2. and .
- (c)
The limit exists and 0<pk<∞. on [0,τ]×[−M,M]p and the matrix
is positive definite, where
and .
- (d)
The kernel function K(·) is symmetric with support [−1,1] and of bounded variation. The bandwidth h satisfies nh2→∞ and nh4→0 as n→∞.
- (e)
There is a σ>0 such that rk(Wki)≥σ for all k and i with δki=1.
Let be the (right continuous) filtration generated by the full data processes . Assume that , i.e. the mark-specific instantaneous failure rate at time t given the observed information up to time t depends on only the failure status and the current covariate value. By the definition of the conditional mark-specific hazard function, . Hence, the mark-specific intensity of Nki(t,v) with respect to equals . Let
By Aalen and Johansen (1978), Mki(·,v1) and Mki(·,v2)−Mki(·,v1) are orthogonal square integrable martingales with respect to for any 0≤v1≤v2≤1.
The weak convergence of for v ∈ [a,b] is given in theorem 1 below.
Theorem 1. Under conditions 1–5, ,uniformly in v ∈ [a,b], where
(21)
The processes WB(v) converge weakly to a p-dimensional mean 0 Gaussian process with continuous sample paths on v ∈ [a,b], where and .
Theorem (2) provides the basis for obtaining asymptotically correct critical values for the testing procedures for H10 and for H20. In particular, let G(v) be the limiting Gaussian process of , v ∈ [a,b], as n→∞. Then, under , as n→∞. By theorem 1 and the continuous mapping theorem, , , and under H10 as n→∞. Under H20, , v ∈ [a,b], as n→∞. Applying the continuous mapping theorem, under H20, , and , as n→∞.
The proof of the consistency of the tests for testing H10 are straightforward. To show the consistency of the tests for testing H20, we note that the derivative under H2m with strict inequality for at least some v ∈ [a,b]. The function is non-decreasing with . We have, under H2m, with strict inequality for at least some v ∈ [a,b]. Let v0 ∈ [a,b] be such that . Then for v≤v0. Now, defining , we have for and for v*≤v<b. It follows from expression (11) and theorem 1 that and under H2m as n→∞ for . Thus the tests based on and are consistent against H2m. Similarly, let . Then, under H2a, for , and for . Hence and under H2a as n→∞ for , resulting in the consistent tests against H2a.
We use the Gaussian multiplier resampling method (Lin et al., 1993) to approximate the distribution of WB(v), v ∈ [a,b]. Let be independent and identically distributed standard normal random variables. Replacing each term of equation (26), which is asymptotically equivalent to equation !(21), by its empirical counterpart and multiplying by ξki, we obtain , where
(22)
where .
Following an application of lemma 1 of Sun and Wu (2005), the distribution of WB(v), v ∈ [a,b], can be approximated by the conditional distribution of , v ∈ [a,b], given the observed data sequence, which can be obtained through repeatedly generating independent sets of . Hence, the distribution of Q(1)(v), v ∈ [a,b], under H10, can be approximated by the conditional distribution of , v ∈ [a,b], given the observed data sequence. By the continuous mapping theorem, the distribution of Q(2)(v), v ∈ [a,b], under H20, can be approximated by the conditional distribution of , v ∈ [a,b], given the observed data sequence.
With the Gaussian multiplier method, the variance can be consistently estimated by , where is the first component on the diagonal of
(23)
A.1. Proof of theorem 1
Let
(24)
Following the proof of theorem 4 of Sun and Gilbert (2012) (their Web appendix (W.19)) and under nh4→0,
(25)
Hence
which, by exchanging the order of integrations, equals
(26)
Let
It follows that
(27)
Since the kernel function K(·) has compact support on [−1,1], equation (27) equals
(28)
It can be shown that converges weakly to a mean 0 Gaussian process with continuous paths. Under assumption 4, has bounded variation and converges uniformly to Σ(x)−1 for x ∈ (h,v−h). By lemma 2 of Gilbert et al. (2008), the first term in expression (28) is equal to . Similar arguments lead to the second and the third terms in expression (28) to be op(1). Hence,
which converges weakly to a p-dimensional mean 0 Gaussian process on v ∈ [a,b] with continuous sample paths by lemma 1 of Sun and Wu (2005). Theorem 1 follows since is a linear transformation of .
© 2014 Royal Statistical Society
This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
Issue Section:
Original Article
Download all slides
Advertisem*nt
Citations
Views
124
Altmetric
More metrics information
Metrics
Total Views 124
94 Pageviews
30 PDF Downloads
Since 3/1/2023
Month: | Total Views: |
---|---|
March 2023 | 5 |
April 2023 | 9 |
June 2023 | 5 |
July 2023 | 6 |
August 2023 | 5 |
September 2023 | 2 |
October 2023 | 4 |
November 2023 | 1 |
December 2023 | 10 |
January 2024 | 9 |
February 2024 | 27 |
March 2024 | 2 |
April 2024 | 10 |
May 2024 | 17 |
June 2024 | 12 |
Altmetrics
Email alerts
Article activity alert
Advance article alerts
New issue alert
Receive exclusive offers and updates from Oxford Academic
Citing articles via
Google Scholar
-
Latest
-
Most Read
-
Most Cited
More from Oxford Academic
Mathematics
Probability and Statistics
Science and Mathematics
Books
Journals
Advertisem*nt