Richard F. Gunst

Spreading the word: Investigating antecedents of consumers’ positive word-of-mouth intentions and behaviors in a retailing context

Empirical studies investigating the antecedents of positive word of mouth (WOM) typically focus on the direct effects of consumers’ satisfaction and dissatisfaction with previous purchasing experiences. The authors develop and test a more comprehensive model of the antecedents of positive. WOM (both intentions and behaviors), including consumer identification and commitment. Specifically, they hypothesize and test commitment as a mediator and moderator of satisfaction on positive WOM and commitment as a mediator of identification on WOM. Using data obtained from customers of a retailer offering both products and services, they find support for all hypothesized relationships with WOM intentions and/or WOM behaviors as the dependent variable. The authors conclude with a discussion of their findings and implications for both marketing theory and practice.

Latent Root Regression Analysis

Least squares estimates of parameters of a multiple linear regression model are known to be highly variable when the matrix of independent variables is near singular. Using the latent roots and latent vectors of the “correlation matrix” of the dependent and independent variables a modified least squares estimation procedure is introduced. This technique enables one to determine whether the near singularity has predictive value and examine alternate prediction equations in which the effect of the near singrtlarity has been removed from the estimates of the regression coefficients. In addition a method for performing backward elimination of variables using standard least squares or the modified procedure is presented.

Biased Estimation in Regression: An Evaluation Using Mean Squared Error

Abstract A mean squared error criterion is used to compare five estimators of the coefficients in a linear regression model: least squares, principal components, ridge regression, latent root, and a shrunken estimator. Each of the biased estimators is shown to offer improvement in mean squared error over least squares for a wide range of choices of the parameters of the model. The results of a simulation involving all five estimators indicate that the principal components and latent root estimators perform best overall, but the ridge regression estimator has the potential of a smaller mean squared error than either of these.

Regression analysis and problems of multicollinearity

Multicollinearity or linear dependence among the vectors of regressor variables in a multiple linear regression analysis can have sever effects on the estimation of parameters and on variables selection techniques. This expository paper examines the sources of multicollinearity and discusses some of its harmful affects. Several methods proposed in the literature for detecting multicollinearity and dealing with the associated problems are also presented and discussed.

A Comparison of Least Squares and Latent Root Regression Estimators

Miilticollinesrity among the columns of regressor variables is known to cause severe distortion of the least squares estimates of the parameters in a multiple linear regression model. An alternate method of estimating the parameters which was proposed by the authors in a previous paper is Latent Root Regression Analysis. In this article several comparisons between the two methods of estimation are presented. The improvement of Latent Root Regression over ordinary least squares is shown to depend on the orientation of the parameter vector with respect to a vector defining the multicollinearity. Despite this dependence on orientation, the authors conclude that witch multicollinear data Latent Root, Regression Analysis is preferable to ordinary least squares for parameter estimation and variable selectJion.

Regresion analysis with multicollinear predictor variables: definition, derection, and effects

For over fifty years researchers have encountered difficulties with least squares estimators when predictor variables in a regression analysis are multicollinear. Extensive research efforts over the last ten to fifteen years have resulted in a clear understanding of many aspects of this problem and have, generated a great deal of controversy over possible solutiors. In this survey the nature and effects of predictor-variable multicollinearities are examined. Emphasis is placed on discussions of the multicollinearity problem itself rather than on classical or Bayesian solutions to the problem.

Peripherally Inserted Central Catheters May Lower the Incidence of Catheter-Related Blood Stream Infections in Patients in Surgical Intensive Care Units

BACKGROUND Long-term central venous catheterization is associated with a higher rate of catheter-related blood stream infections (CR-BSI). It is unclear whether there is a difference in the CR-BSI rate associated with central venous catheters (CVCs) and peripherally inserted central catheters (PICCs) in long-stay patients in surgical intensive care units (SICUs). We hypothesized that PICC use reduces the rate of CR-BSI compared with use of antiseptic CVCs in these patients. METHODS All 121 patients admitted to our SICU for ≥14 days between July 2005 and July 2006 were included. Central venous access was maintained with an antiseptic CVC (Arrow Guard silver/chlorhexidine; n = 263) or replacement with a PICC (n = 37). Experienced residents, using maximum barrier precautions and chlorhexidine skin preparation, placed central lines; a credentialed registered nurse placed PICCs similarly. A CR-BSI was defined by semi-quantitative catheter tip cultures with ≥15 colony-forming units and at least one positive blood culture with the same organism. Multivariable regression was performed to identify predictors of CR-BSI. RESULTS There were 13 CVC infections and one PICC infection, resulting in an infection rate of 6.0/1,000 catheter-days for CVCs and 2.2/1,000 for PICCs. Infected and non-infected CVCs were in place a mean of 25 ± 11 and 16 ± 9 days, respectively. The infected PICC was in place for 19 days, whereas the remainder of the PICCs were in place a mean of 14 ± 17 days. Logistic regression demonstrated that line days (duration of catheterization) was the only independent predictor of CVC infection (p = 0.015). CONCLUSION In this non-randomized study, PICC was associated with fewer CR-BSIs in long-stay SICU patients, although CVCs were in place longer than PICC lines. The only predictor of CVC infection was the duration the line was in place. These results suggest that minimizing the duration of central venous access and substituting PICC for CVC may reduce the incidence of CR-BSI in long-stay SICU patients.

Some Considerations in the Evaluation of Alternate Prediction Equations

Prediction equations constructed from multiple linear regression analyses are often intended for use in predicting response values throughout a region of the space of the predictor variables. Criteria for evaluating prediction equations, however, have generally concentrated attention on mean squared error properties of the estimated regression coefficients or on mean squared error properties of the predictor at the design points. If adequate prediction throughout a region of the space of predictor variables is the goal, neither of these criteria may be satisfactory in assessing the predictor. In this paper integrated mean squared error is used as a criterion to determine when the least squares, principal component, and ridge regression estimators of regression coefficients can produce satisfactory prediction equations in the presence of a multicollinear design matrix.

Outlier-Induced Collinearities

When an observation in a regression analysis has very large values on two or more predictor variables, artificial collinearities can be induced. The effects of such collinearities on a regression analysis are not well documented, although they can be shown to be similar in many respects to those resulting from approximate linear dependencies among the columns of predictor-variable values. The purpose of this article is to explore the effects of outlier-induced collinearities on the estimation of regression coefficients.

Selecting principal components in regression

Criteria for the deletion of principal components in regression are usually based on one of two indicators of components effects: (i) the magnitude of the eigenvalues of the predictor-variable correlation matrix or (ii) statistical tests of the significance of the components. Advocates of the first criterion cite guaranteed variance reduction properties as a rational for their proposals whereas proponents of inferential criteria point out that deletion solely on the basis of the magnitude of the eigenvalues ignores the potentials for bias. In this note we discuss the liminations of the second approach.