Effects of a single outlier on arma identification

Fox (1972), Box and Tiao (1975), and Abraham and Box (1979) have proposed methods for detecting outliers in time series whose ARMA form is known (or identified). We show that the existence of a single aberrant observation, innovation, or intervention causes an ARMA model to be misidentified using unadjusted autocorrelation (acf) and partial autocorrelation estimates. The magnitude, location, type of outlier, and in some cases the ARMA's parameters, affect the identification outcome. We use variance inflation, signal-to-noise ratios, and acf critical values to determine an ARMA model's susceptibility to misidentifi-cation. Numerical and simulation examples suggest how to iteratively use the outlier detection methods in practice.     

Research Notes

In the summer of 1995, a required MSW course was taught simultaneously to two sections of students at the University of Georgia, 11 on the main campus and 9 at a branch campus. Each section was alternately taught using live instruction or distance learning (two-way interactive television). At the conclusion of the course, students evaluated each method of teaching using Biner’s (1993) instrument for measuring attitudes toward televised courses. Live instruction was rated significantly higher than distance learning. Although distance learning technology has potential for augmenting social work educational resources, it has not yet demonstrated comparable outcomes in terms of student learning.     

Multiple hypothesis test for parameter constancy based on recursive residuals

This article presents a multiple hypothesis test procedure that combines two well known tests for structural change in the linear regression model, the CUSUM test and the recursive t test. The CUSUM test is run through the sequence of recursive residuals as usual; if the CUSUM plot does not violate the critical lines, one more step is taken to perform the t test for hypothesis of zero mean based on all recursive residuals. The asymptotic size of this multiple hypothesis test is derived; power simulation results suggest that it outperforms the traditional CUSUM test and complements other tests that are currently stressed in econometrics.     

Celebrity and politics: Effects of endorser credibility and sex on voter attitudes,perceptions, and behaviors

While much research examines the effects of celebrity endorsements in commercial advertising, scholars have only recently sought to investigate the effects of celebrity endorsements of politicians on voter perceptions and behavior. This study expands existing research on celebrity political endorsement effects via an experiment exploring effects of different versions of a news story describing a celebrity's endorsement of a political candidate on participants’ voting attitudes, perceptions of candidate credibility, and voting behavioral intent. Although participants perceive credibility differences between high- and low-credibility celebrities, neither endorser credibility nor endorser sex impact attitudes toward the endorsed candidate, perceptions of the candidate's credibility, or intended voting behavior. Conceptual relationships to other studies on celebrity endorsement effects are discussed, as are implications, limitations, and directions for future research.     

Humps and Bumps in Lifetime Consumption

In this article we argue that the life-cycle model that allows demographics to affect household preferences and relaxes the assumption of certainty equivalence can generate hump-shaped consumption profiles over age that are very similar to those observed in household-level data sources and, in particular, match the differences in shape across different education groups. Liquidity constraints or myopia are not required to explain the empirical features of observed life-cycle patterns.     

Tests concerning two non-isotropic principal components

Kshirsagar & Gupta (1965) proposed test criteria for testing the null hypothesis that a p x p covariance matrix, with known smallest latent root of multiplicity p-2. has its two non-isotropic principal components in specified directions. Simula-tion results indicate that the power functions of these test criteria lack some desirable properties. However, another statis-tic discussed by Kshirsagar and Gupta is shown to have a power-function which performs quite well. The situation in which the smallest latent root of multiplicity p-2 is unknown, is also considered.