Randomization tests for multiple regression

Four generic means of conducting randomization tests in the context of multiple regression are analysed. Based on their performance in traditional repeated samples, three of these are shown to be inappropriate or applicable only in special circumstances; their shortcomings are illustrated via Monte Carlo studies     

When Precaution Creates Misunderstandings: The Unintended Effects of Precautionary Information on Perceived Risks,the EMF Case

In the past decade, growing public concern about novel technologies with uncertain potential long‐term impacts on the environment and human health has moved risk policies toward a more precautionary approach. Focusing on mobile telephony, the effects of precautionary information on risk perception were analyzed. A pooled multinational experimental study based on a 5 × 2 × 2 factorial design was conducted in nine countries. The first factor refers to whether or not information on different types of precautionary measures was present, the second factor to the framing of the precautionary information, and the third factor to the order in which cell phones and base stations were rated by the study participants. The data analysis on the country level indicates different effects. The main hypothesis that informing about precautionary measures results in increased risk perceptions found only partial support in the data. The effects are weaker, both in terms of the effect size and the frequency of significant effects, across the various precautionary information formats used in the experiment. Nevertheless, our findings do not support the assumption that informing people about implemented precautionary measures will decrease public concerns.     

Demand for health risk reductions: A cross-national comparison between the U.S. and Canada

Using a large stated preference survey conducted across the U.S. and Canada, we assess differences in individual willingness to pay (WTP) for health risk reductions between the two countries. Our utility-theoretic choice model allows for systematically varying marginal utilities for avoided future time in different adverse health states (illness-years, recovered/remission years, and lost life-years). We find significant differences between Canadian and U.S. preferences. WTP also differs systematically with age, gender, education, and marital status, as well as a number of attitudinal and subjective health-perception variables. Age profiles for WTP are markedly different across the two countries. Canadians tend to display flatter age profiles, with peak WTP realized at older ages.     

Estimating the prevalence of sensitive behaviour and cheating with a dual design for direct questioning and randomized response

Randomized response is a misclassification design to estimate the prevalence of sensitive behaviour. Respondents who do not follow the instructions of the design are considered to be cheating. A mixture model is proposed to estimate the prevalence of sensitive behaviour and cheating in the case of a dual sampling scheme with direct questioning and randomized response. The mixing weight is the probability of cheating, where cheating is modelled separately for direct questioning and randomized response. For Bayesian inference, Markov chain Monte Carlo sampling is applied to sample parameter values from the posterior. The model makes it possible to analyse dual sample scheme data in a unified way and to assess cheating for direct questions as well as for randomized response questions. The research is illustrated with randomized response data concerning violations of regulations for social benefit.     

How non-normality affects the quadratic discriminant function

The quadratic discriminant function is commonly used for the two group classification problem when the covariance matrices in the two populations are substantially unequal. This procedure is optimal when both populations are multivariate normal with known means and covariance matrices. This study examined the robustness of the QDF to non-normality. Sampling experiments were conducted to estimate expected actual error rates for the QDF when sampling from a variety of non-normal distributions. Results indicated that the QDF was robust to non-normality except when the distributions were highly skewed, in which case relatively large deviations from optimal were observed. In all cases studied the average probabilities of misclassification were relatively stable while the individual population error rates exhibited considerable variability.     

Unobserved-Components Models for Seasonal Adjustment Filters

Time series models are presented, for which the seasonal-component estimates delivered by linear least squares signal extraction closely approximate those of the standard option of the widely-used Census X-11 program. Earlier work is extended by consideration of a broader class of models and by examination of asymmetric filters, in addition to the symmetric filter implicit in the adjustment of historical data. Various criteria that guide the specification of unobserved- components models are discussed, and a new preferred model is presented. Some nonstandard options in X-11 are considered in the Appendix.