Power and Sample Size Determination When the Alternative Hypotheses are Given in Quantiles

The power of normal-theory tests about means depends on a noncentrality parameter which is a function of the unknown parameter σ. In order to calculate power and to solve sample-size problems based on power, differences between hypothesized and alternative values of the means are frequently selected as a multiple of σ, a choice which eliminates σ from the noncentrality parameter and permits a solution. Perhaps a more natural (but equivalent) way to express alternatives is to give one or more means as the quantile of order p (say Q_p ) of a distribution with another mean. As we will demonstrate, this kind of alternative also eliminates σ from the problem.     

A study on tuning parameter selection for the high-dimensional lasso

High-dimensional predictive models, those with more measurements than observations, require regularization to be well defined, perform well empirically, and possess theoretical guarantees. The amount of regularization, often determined by tuning parameters, is integral to achieving good performance. One can choose the tuning parameter in a variety of ways, such as through resampling methods or generalized information criteria. However, the theory supporting many regularized procedures relies on an estimate for the variance parameter, which is complicated in high dimensions. We develop a suite of information criteria for choosing the tuning parameter in lasso regression by leveraging the literature on high-dimensional variance estimation. We derive intuition showing that existing information-theoretic approaches work poorly in this setting. We compare our risk estimators to existing methods with an extensive simulation and derive some theoretical justification. We find that our new estimators perform well across a wide range of simulation conditions and evaluation criteria.     

Uncovering bias in order assignment

To mitigate sequencing effects in decision-making, many situations require a set of items to be considered in a random order. When such orderings are repeated, one can test whether randomization indeed obtains, or whether some orderings have been manipulated in order to achieve a favorable result. This paper articulates the key features of this problem and presents three general tests for randomness. These methods are used to analyze the order in which lottery numbers are drawn in Powerball, contestants perform on American Idol, and candidates are placed on election ballots. This last application features frequent manipulation, with potentially serious consequences.     

When “Life Is but a Dream”: Obliterating Politics Through Business Process Reengineering?

In this article, we explore the genesis andoperation of Business Process Reengineering (BPR) withina medium-sized U.K. bank from the late 1980s to themid-1990s. We dismiss the claims of those evangelical gurus who assume that BPR can bedecontextualized and decoupled from organizationalpolitics and posit that BPR can be managedinstantaneously and unproblematically. Instead we arguethat BPR is likely to be constituted by and through politicalrelations, and that BPR in turn will reconstituteorganizational forms and norms, in a highly politicalfashion. We endeavor to build upon current approachestoward organizational politics. We illustrate thatpolitics is not simply about resistance to some putativeorganizational norm of stability or uniformity as BPR'sgurus imply. Nor does it deriveexclusively from diverse interest groups pursuing separateor conflicting ends that can be juggled and managed asprocessual or pluralistic accounts of organizationalchange tend to assume. Neither, in this instance, can one interpret politics as being entirelyaxiomatic with labor's resistance to management(capital) which is characteristic of a traditional laborprocess analysis, although expressions of this were apparent in our case study. We suggest thatpolitics also needs to be understood interms of power and identity relations or how individualsseek, through political maneuverings, to further or secure their individual careers andidentities in an uncertain world. In view of this, weargue that politics are essential to the very fabric oforganizational life, which renders the outcomes of BPR uncertain and contested.     

A note on the solution of sample size problems using the Odeh and Fox charts

The usefulness of the Odeh and Fox (1975) charts is much enhanced by having φ for v₂ = ∞, a value quickly found from a table of Haynam, Govindarajulu and Leone (1970).     

Optimal detection of a jump in the intensity of a Poisson process or in a density with likelihood ratio statistics

We consider the problem of detecting a ‘bump’ in the intensity of a Poisson process or in a density. We analyze two types of likelihood ratio‐based statistics, which allow for exact finite sample inference and asymptotically optimal detection: The maximum of the penalized square root of log likelihood ratios (‘penalized scan’) evaluated over a certain sparse set of intervals and a certain average of log likelihood ratios (‘condensed average likelihood ratio’). We show that penalizing the square root of the log likelihood ratio — rather than the log likelihood ratio itself — leads to a simple penalty term that yields optimal power. The thus derived penalty may prove useful for other problems that involve a Brownian bridge in the limit. The second key tool is an approximating set of intervals that is rich enough to allow for optimal detection, but which is also sparse enough to allow justifying the validity of the penalization scheme simply via the union bound. This results in a considerable simplification in the theoretical treatment compared with the usual approach for this type of penalization technique, which requires establishing an exponential inequality for the variation of the test statistic. Another advantage of using the sparse approximating set is that it allows fast computation in nearly linear time. We present a simulation study that illustrates the superior performance of the penalized scan and of the condensed average likelihood ratio compared with the standard scan statistic.     

BEYOND BELIEF: ATHEISM,AGNOSTICISM, AND THEISTIC CERTAINTY IN THE UNITED STATES

Atheism, agnosticism, and undoubting belief in a one true god are pivotal issues distinguishing religious adherents. Shifts in the rates of atheism and agnosticism can influence the composition of religious markets. Demand side rational actor models and the new demographic approach to religious commitments emphasize the influence of life course dynamics on religious desires. I examine multiple predictors of different types of belief in God using data from the General Social Surveys. I employ ordinal logistic regression to test the effects of social status, gender, race, region and city size, family structure, and denominational affiliation on the odds of being an atheist, agnostic, doubting believer, or a consistent believer. I also examine trends in these belief configurations, and whether cohort variations might evidence a coming shift in the relative rates of belief and unbelief.     

Queer Is the New Black? Not So Much: Racial Disparities in Anti-LGBTQ Discrimination

The present study examines the intersection of race and sexual orientation in the experience of discrimination among lesbian, gay, bisexual, transgender, and queer (LGBTQ) people. The results of the study suggest that while a majority of LGBTQ individuals report being victims of anti-LGBTQ discrimination, racial minorities experience even greater levels of anti-LGBTQ discrimination than do White LGBTQ people. The findings suggest that the intersection of race and sexual orientation creates elevated levels of discrimination risk beyond the already elevated rates of discrimination experienced by members of the LGBTQ community for LGBTQ racial minorities.