We analyse the effect of the anonymisation method multiplicative stochastic noise on the within estimation of a linear panel model. In particular, we concentrate on the panel model with serially correlated regressors. In addition to anonymisation as such, the serial correlation in a data set with only few points in time increases the bias of the within estimator and therefore must be taken into...

ABSTRACT Accounting studies often examine whether the relation between X and Y varies with a moderating variable, M, by including an interactive term, × in regression. We provide plain-English guidance on why, how, when to use control variables, Z, interaction tests. A simulation simple descriptions demonstrate how interacted controls affect coefficient estimates interpretations. In particular,...

We propose serial correlation robust asymptotic confidence bands for the receiver operating characteristic (ROC) curves estimated by quasi-maximum likelihood in the binormal model. Our simulation experiments confirm that this new method performs fairly well in finite samples. The conventional procedure is found to be markedly undersized in terms of yielding empirical coverage probabilities lowe...

Consider a linear model setting in which the explanatory variables are specified by time series. To sequentially test for the stability of the regression parameters in time, we introduce a detector which is based on the first excess time of a CUSUM-type statistic over a suitably defined threshold function. The main aim of this paper is to derive the limit distribution of the detector. By provid...

We consider statistical inference on a single component of a parameter vector that satisfies a finite number of moment inequalities. The null hypothesis for this single component is given a dual characterization as a composite hypothesis regarding point identified parameters. We also are careful in the specification of the alternative hypothesis that also has a dual characterization as a compos...

Relying on the well-established theoretical result that uncertainty has a common and an idiosyncratic component, we propose a new measure of earnings forecast uncertainty as the sum of dispersion among analysts and the variance of mean forecast errors estimated by a GARCH model. The new measure is based on both common and private information available to analysts at the time they make their for...

This paper explores the issue of efficiency in English higher education using data envelopment analysis and stochastic frontier analysis to estimate an output distance function (which incorporates measures of both quantity and quality of teaching and research inputs and outputs) over a thirteen-year period. The study compares the efficiency estimates derived from various estimation methods, and...

Aggregated qualitative survey data provide timely, but often imperfect, macroeconomic indicators. Exploiting a unique panel dataset for the UK, which contains matched firm-level responses from both qualitative and quantitative surveys, we find that firms’ responses are influenced not only by their own current and past output and lagged qualitative responses but also by an indicator of aggregate...

We have observed the recently discovered rich star cluster GLIMPSE-C01 for 46 ks with the Chandra X-ray Observatory. Seventeen X-ray sources with luminosities & 0.6×1031 ergs s−1 were discovered, one of which is likely a quiescent low-mass X-ray binary. The spatial distribution of these X-ray sources with respect to the NIR and IR images of the cluster, combined with the luminosity and spectral...

We present an exact test for whether two random variables that have known bounds on their support are negatively correlated. The alternative hypothesis is that they are not negatively correlated. No assumptions are made on the underlying distributions. We show by example that the Spearman rank correlation test as the competing exact test of correlation in nonparametric settings rests on an addi...