A variable-neighbourhood search algorithm for finding optimal run orders in the presence of serial correlation

The responses obtained from response surface designs that are run sequentially often exhibit serial correlation or time trends. The order in which the runs of the design are performed then has an impact on the precision of the parameter estimators. This article proposes the use of a variable-neighbourhood search algorithm to compute run orders that guarantee a precise estimation of the effects of the experimental factors. The importance of using good run orders is demonstrated by seeking D-optimal run orders for a central composite design in the presence of an AR(1) autocorrelation pattern.     

The performance of subset response surface designs for estimating third order terms

Response surface designs are widely used in industries like chemicals, foods, pharmaceuticals, bioprocessing, agrochemicals, biology, biomedicine, agriculture and medicine. One of the major objectives of these designs is to study the functional relationship between one or more responses and a number of quantitative input factors. However, biological materials have more run to run variation than in many other experiments, leading to the conclusion that smaller response surface designs are inappropriate. Thus designs to be used in these research areas should have greater replication. Gilmour (2006) introduced a wide class of designs called “subset designs” which are useful in situations in which run to run variation is high. These designs allow the experimenter to fit the second order response surface model. However, there are situations in which the second order model representation proves to be inadequate and unrealistic due to the presence of lack of fit caused by third or higher order terms in the true response surface model. In such situations it becomes necessary for the experimenter to estimate these higher order terms. In this study, the properties of subset designs, in the context of the third order response surface model, are explored.     

Evaluation of Response Surface Designs in Presence of Errors in Factor Levels

Experimenters are often confronted with the problem that errors in setting factor levels cannot be measured. In the robust design scenario, the goal is to determine the design that minimizes the variability transmitted to the response from the variables’ errors. The prediction variance performance of response surface designs with errors is investigated using design efficiency and the maximum and minimum scaled prediction variance. The evaluation and comparison of response surface designs with and without errors in variables are developed for second order designs on spherical regions. The prediction variance and design efficiency results and recommendations for their use are provided.     

The robustness of response surface designs with errors in factor levels

ABSRTACT

Since errors in factor levels affect the traditional statistical properties of response surface designs, an important question to consider is robustness of design to errors. However, when the actual design could be observed in the experimental settings, its optimality and prediction are of interest. Various numerical and graphical methods are useful tools for understanding the behavior of the designs. The D- and G-efficiencies and the fraction of design space plot are adapted to assess second-order response surface designs where the predictor variables are disturbed by a random error. Our study shows that the D-efficiencies of the competing designs are considerably low for big variance of the error, while the G-efficiencies are quite good. Fraction of design space plots display the distribution of the scaled prediction variance through the design space with and without errors in factor levels. The robustness of experimental designs against factor errors is explored through comparative study. The construction and use of the D- and G-efficiencies and the fraction of design space plots are demonstrated with several examples of different designs with errors.     

Estimation in Second-Order Models with Errors in the Factor Levels

The impact of errors in the factor levels is examined on the estimation of parameters in second-order response models. Errors can occur in setting the factor levels for response surface and robust parameter design models. These errors can lead to heterogeneity of variances in model errors that make ordinary least squares estimation inappropriate. Weighted least squares and maximum likelihood estimation approaches are developed as viable alternatives where it is assumed the variances and covariances of the errors are known. Performance of these estimation techniques are examined in simulation studies for two examples. Another example is given that applies these results.     

Optimal allocation of time points for the random effects model

In this article the problem of the optimal selection and allocation of time points in repeated measures experiments is considered. D‐ optimal designs for linear regression models with a random intercept and first order auto‐regressive serial correlations are computed numerically and compared with designs having equally spaced time points. When the order of the polynomial is known and the serial correlations are not too small, the comparison shows that for any fixed number of repeated measures, a design with equally spaced time points is almost as efficient as the D‐ optimal design. When, however, there is no prior knowledge about the order of the underlying polynomial, the best choice in terms of efficiency is a D‐ optimal design for the highest possible relevant order of the polynomial. A design with equally‐spaced time points is the second best choice     

Estimation of a response surface from-a general incomplete block design

If an incomplete block design is used to compare v treatments and later it is decided to fit a second order response surface using the fact that the v treatments are v combinations of levels of m different factors, how will one proceed to obtain this? This paper demonstrates how the usual analysis of an incomplete block design with or without recovery of interblock information can be augmented to yield estimates of response surface coefficients.     

Robust Estimation of Multi-response Surfaces Considering Correlation Structure

Response surfaces express the behavior of responses and can be used for both single and multi-response problems. A common approach to estimate a response surface using experimental results is the ordinary least squares (OLS) method. Since OLS is very sensitive to outliers, some robust approaches have been discussed in the literature. Although there are many methods available in the literature for multiple response optimizations, there are a few studies in model building especially robust models. Assuming correlated responses, in this paper, a robust coefficient estimation method is proposed for multi response problem based on M-estimators. In order to illustrate the performance of the proposed procedure, a contaminated experimental design using a numerical example available in the literature with some modifications is used. Both the classical multivariate least squares method and the proposed robust multivariate approach are used to estimate regression coefficients of multi-response surfaces based on this example. Moreover, a comparison of the proposed robust multi response surface (RMRS) approach with separate robust estimation of single response show that the proposed approach is more efficient.     

Dealing with emrqes in variables in response surface exploration

This paper is concerned with the analysis of data obtained from a designed experiment where the experimental design cannot be implemented exactly as planned, because errors in the levels of the variables cannot be avoided or measured. When the primary interest of the investigator lies In obtaining a satisfactory response surface model for the investigated relationship, the precision of the model estimates is essential for successful model building and accurate prediction of the response. An iterative procedure is proposed which estimates the effect of the variable in errors and obtains efficient weighted least squares estimates of the parameters of Interest.     

Some new augmented fractional Box–Behnken designs

The augmented Box–Behnken designs are used in the situations in which Box–Behnken designs (BBDs) could not estimate the response surface model due to the presence of third-order terms in the response surface models. These designs are too large for experimental use. Usually experimenters prefer small response surface designs in order to save time, cost, and resources; therefore, using combinations of fractional BBD points, factorial design points, axial design points, and complementary design points, we augment these designs and develop new third-order response surface designs known as augmented fractional BBDs (AFBBDs). These AFBBDs have less design points and are more efficient than augmented BBDs.     

Randomized response technique in a National Survey

In order to produce estimates of the number of women having abortions during a 12-month period in the conterminous United States, the randomized response technique was used in the 1973 National Survey of Family Growth. The model applied used 2 unrelated questions in separate half-samples, with a coin as the randomizing device. The randomized response technique resulted in a higher estimate for the number of women with abortions that has previously been obtained through direct questions or reporting systems. The overall estimated proportion who had abortions among women who had been married or who had their own children in the household is 3.0% with a standardized error of 0.8 percentage points; however, there is a wide variation in the half-sample estimates of abortion. Differences between the 2 half-samples led to an examination of possible measurement error. 3 types of errors in measurement which may affect the estimate based on the randomized response technique are: 1) error in the answer to the sensitive question on abortion under the randomized response conditions; 2) error in the answer to the innocuous question under randomized response conditions; and 3) error in the answer to the innocuous question when asked directly. Comparisons between data from the different sources for currently married women suggest that all differences are not due to measurement error and that a large number of women had an unreported and/or illegal abortion in 1973. Although the randomized response models have been in use for at least 10 years, there continues to be a need for work on the field administration and subsequent analysis of these models.     

Asymptotic properties of a quast-maximum likelihood estimator in truncated regression model with serial correlation

Robinson (1982a) presented a general approach to serial correlation in limited dependent variable models and proved the strong consistency and asymptotic normality of the quasi-maximum likelihood estimator (QMLE) for the Tobit model with serial correlation, obtained under the assumption of independent errors. This paper proves the strong consistency and asymptotic normality of the QMLE based on independent errors for the truncated regression model with serial correlation and gives consistent estimators for the limiting covariance matrix of the QMLE.     

Asymptotic properties of a quast-maximum likelihood estimator in truncated regression model with serial correlation

Robinson (1982a) presented a general approach to serial correlation in limited dependent variable models and proved the strong consistency and asymptotic normality of the quasi-maximum likelihood estimator (QMLE) for the Tobit model with serial correlation, obtained under the assumption of independent errors. This paper proves the strong consistency and asymptotic normality of the QMLE based on independent errors for the truncated regression model with serial correlation and gives consistent estimators for the limiting covariance matrix of the QMLE.     

Testing for autocorreiation in nested analysis of variance

This paper considers the problem in which repeated measures are taken on each subject in an experimental design and the model errors follow a first order autoregressive process. The appropriate form of the Durbin-Watson test statistic for autocorrelated errors is presented. Critical values for the test statistic are reported.     

Optimal designs for non-competitive enzyme inhibition kinetic models

In this paper, we present a new method for determining optimal designs for enzyme inhibition kinetic models, which are used to model the influence of the concentration of a substrate and an inhibition on the velocity of a reaction. The approach uses a nonlinear transformation of the vector of predictors such that the model in the new coordinates is given by an incomplete response surface model. Although there exist no explicit solutions of the optimal design problem for incomplete response surface models so far, the corresponding design problem in the new coordinates is substantially more transparent, such that explicit or numerical solutions can be determined more easily. The designs for the original problem can finally be found by an inverse transformation of the optimal designs determined for the response surface model. We illustrate the method determining explicit solutions for the D-optimal design and for the optimal design problem for estimating the individual coefficients in a non-competitive enzyme inhibition kinetic model.     

Minimax designs for approximately linear models with AR(1) errors

We obtain designs for linear regression models under two main departures from the classical assumptions: (1) the response is taken to be only approximately linear, and (2) the errors are not assumed to be independent, but to instead follow a first-order autoregressive process. These designs have the property that they minimize (a modification of) the maximum integrated mean squared error of the estimated response, with the maximum taken over a class of departures from strict linearity and over all autoregression parameters ρ,|ρ,| < 1, of fixed sign. Specific methods of implementation are discussed. We find that an asymptotically optimal procedure for AR(1) models consists of choosing points from that design measure which is optimal for uncorrelated errors, and then implementing them in an appropriate order.     

Statistical methodology for binary data from a set of partially structured two-way random cross-classifications

A complete two-way cross-classification design is not practical in many settings. For example, in a toxicological study where 30 male rats are mated with 30 female rats and each mating outcome (successful or unsuccessful)is observed, time and resource considerations can make the use of the complete design prohibitively costly. Partially structured variations of this design are, therefore, of interest (e.g., the balanced disjoint rectangle design, the fully diagonal design, and the "S"-design). Methodology for analyzing binary data from such incomplete designs is illustrated with an example. This methodology, which is based on infinite population sampling arguments, allows the estimation of the mean response, among-row correlation coefficient, among-column correlation coefficient, and the within-cell correlation coefficient as well as their standard errors.     

Robust designs for experiments with blocks

ABSTRACT

For experiments running in field plots or over time, the observations are often correlated due to spatial or serial correlation, which leads to correlated errors in a linear model analyzing the treatment means. Without knowing the exact correlation matrix of the errors, it is not possible to compute the generalized least-squares estimator for the treatment means and use it to construct optimal designs for the experiments. In this paper, we propose to use neighborhoods to model the covariance matrix of the errors, and apply a modified generalized least-squares estimator to construct robust designs for experiments with blocks. A minimax design criterion is investigated, and a simulated annealing algorithm is developed to find robust designs. We have derived several theoretical results, and representative examples are presented.     

Non-linear design problem in a chemical kinetic model with non-constant error variance

For a locally optimum non-linear design problem for a chemical kinetic model, we investigate the influence of the dispersion structure of the random observation errors on the design and its efficiency. We find that there are two kinds of design determined by the model parameters and the error variance function: “interior” designs, and “boundary” designs that depend also on the design range. We give an exact criterion for determining which kind of design will arise and we illustrate the qualitative difference between the two kinds of design in terms of the design locus and the equivalence theorem. We tabulate quantitative details of the designs for a range of parameter values.     

Testing for lack of fit in linear multiresponse models based on exact or near replicates

Three procedures for testing the adequacy of a proposed linear multiresponse regression model against unspecified general alternatives are considered. The model has an error structure with a matrix normal distribution which allows the vector of responses for a particular run to have an unknown covariance matrix while the responses for different runs are uncorrelated. Furthermore, each response variable may be modeled by a separate design matrix. Multivariate statistics corresponding to the classical univariate lack of fit and pure error sums of squares are defined and used to determine the multivariate lack of fit tests. A simulation study was performed to compare the power functions of the test procedures in the case of replication. Generalizations of the tests for the case in which there are no independent replicates on all responses are also presented.