Causality and endogeneity: Problems and solutions

Most leadership and management researchers ignore one key design and estimation problem rendering parameter estimates uninterpretable: Endogeneity. We discuss the problem of endogeneity in depth and explain conditions that engender it using examples grounded in the leadership literature. We show how consistent causal estimates can be derived from the randomized experiment, where endogeneity is eliminated by experimental design. We then review the reasons why estimates may become biased (i.e., inconsistent) in non-experimental designs and present a number of useful remedies for examining causal relations with non-experimental data. We write in intuitive terms using nontechnical language to make this chapter accessible to a large audience.

Being Drunk to Have Fun or to Forget Problems? Identifying Enhancement and Coping Drinkers Among Risky Drinking Adolescents

Prevention programs in adolescence are particularly effective if they target homogeneous risk groups of adolescents who share a combination of particular needs and problems. The present work aims to identify and classify risky single-occasion drinking (RSOD) adolescents according to their motivation to engage in drinking. An easy-to-use coding procedure was developed. It was validated by means of cluster analyses and structural equation modeling based on two randomly selected subsamples of a nationally representative sample of 2,449 12- to 18-year-old RSOD students in Switzerland. Results revealed that the coding procedure classified RSOD adolescents as either enhancement drinkers or coping drinkers. The high concordance (Sample A: kappa - .88, Sample B: kappa - .90) with the results of the cluster analyses demonstrated the convergent validity of the coding classification. The fact that enhancement drinkers in both subsamples were found to go out more frequently in the evenings and to have more satisfactory social relationships, as well as a higher proportion of drinking peers and a lower likelihood to drink at home than coping drinkers demonstrates the concurrent validity of the classification. To conclude, the coding procedure appears to be a valid, reliable, and easy-to-use tool that can help better adapt prevention activities to adolescent risky drinking motives.

Computing infinite plans for LTL goals using a classical planner

Classical planning has been notably successful in synthesizing finite plans to achieve states where propositional goals hold. In the last few years, classical planning has also been extended to incorporate temporally extended goals, expressed in temporal logics such as LTL, to impose restrictions on the state sequences generated by finite plans. In this work, we take the next step and consider the computation of infinite plans for achieving arbitrary LTL goals. We show that infinite plans can also be obtained efficiently by calling a classical planner once over a classical planning encoding that represents and extends the composition of the planningdomain and the B¨uchi automaton representingthe goal. This compilation scheme has been implemented and a number of experiments are reported.

Once-yearly zoledronic acid and days of disability, bed rest, and back pain: randomized, controlled HORIZON Pivotal Fracture Trial.

The objective of this study was to determine the effect of once-yearly zoledronic acid on the number of days of back pain and the number of days of disability (ie, limited activity and bed rest) owing to back pain or fracture in postmenopausal women with osteoporosis. This was a multicenter, randomized, double-blind, placebo-controlled trial in 240 clinical centers in 27 countries. Participants included 7736 postmenopausal women with osteoporosis. Patients were randomized to receive either a single 15-minute intravenous infusion of zoledronic acid (5 mg) or placebo at baseline, 12 months, and 24 months. The main outcome measures were self-reported number of days with back pain and the number of days of limited activity and bed rest owing to back pain or a fracture, and this was assessed every 3 months over a 3-year period. Our results show that although the incidence of back pain was high in both randomized groups, women randomized to zoledronic acid experienced, on average, 18 fewer days of back pain compared with placebo over the course of the trial (p = .0092). The back pain among women randomized to zoledronic acid versus placebo resulted in 11 fewer days of limited activity (p = .0017). In Cox proportional-hazards models, women randomized to zoledronic acid were about 6% less likely to experience 7 or more days of back pain [relative risk (RR) = 0.94, 95% confidence interval (CI) 0.90-0.99] or limited activity owing to back pain (RR = 0.94, 95% CI 0.87-1.00). Women randomized to zoledronic acid were significantly less likely to experience 7 or more bed-rest days owing to a fracture (RR = 0.58, 95% CI 0.47-0.72) and 7 or more limited-activity days owing to a fracture (RR = 0.67, 95% CI 0.58-0.78). Reductions in back pain with zoledronic acid were independent of incident fracture. Our conclusion is that in women with postmenopausal osteoporosis, a once-yearly infusion with zoledronic acid over a 3-year period significantly reduced the number of days that patients reported back pain, limited activity owing to back pain, and limited activity and bed rest owing to a fracture.

Limits on the validity of infinite length assumptions for modelling shallow landslides

The infinite slope method is widely used as the geotechnical component of geomorphic and landscape evolution models. Its assumption that shallow landslides are infinitely long (in a downslope direction) is usually considered valid for natural landslides on the basis that they are generally long relative to their depth. However, this is rarely justified, because the critical length/depth (L/H) ratio below which edge effects become important is unknown. We establish this critical L/H ratio by benchmarking infinite slope stability predictions against finite element predictions for a set of synthetic two-dimensional slopes, assuming that the difference between the predictions is due to error in the infinite slope method. We test the infinite slope method for six different L/H ratios to find the critical ratio at which its predictions fall within 5% of those from the finite element method. We repeat these tests for 5000 synthetic slopes with a range of failure plane depths, pore water pressures, friction angles, soil cohesions, soil unit weights and slope angles characteristic of natural slopes. We find that: (1) infinite slope stability predictions are consistently too conservative for small L/H ratios; (2) the predictions always converge to within 5% of the finite element benchmarks by a L/H ratio of 25 (i.e. the infinite slope assumption is reasonable for landslides 25 times longer than they are deep); but (3) they can converge at much lower ratios depending on slope properties, particularly for low cohesion soils. The implication for catchment scale stability models is that the infinite length assumption is reasonable if their grid resolution is coarse (e.g. >25?m). However, it may also be valid even at much finer grid resolutions (e.g. 1?m), because spatial organization in the predicted pore water pressure field reduces the probability of short landslides and minimizes the risk that predicted landslides will have L/H ratios less than 25. Copyright (c) 2012 John Wiley & Sons, Ltd.

Hyperactivity and attention problems in a Swiss sample of school-aged children: effects of school achievement, child gender, and informants.

OBJECTIVE: The sensitivity and tolerance regarding ADHD symptoms obviously differ from one culture to another and according to the informants (parents, teachers, or children). This stimulates the comparison of data across informants and countries. METHOD: Parents and teachers of more than 1,000 school-aged Swiss children (5 to 17 years old) fill in Conners's questionnaires on ADHD. Children who are older than 10 years old also fill in a self-report questionnaire. Results are compared to data from a North American sample. RESULTS: Swiss parents and teachers tend to report more ADHD symptoms than American parents and teachers as far as the oldest groups of children are concerned. Interactions are evidenced between school achievement, child gender, and informants. A relatively low rate of agreement between informants is found. CONCLUSION: These results strengthen the importance to take into account all informants in the pediatric and the child psychiatry clinic, as well as in the epidemiological studies.

On Certain Greedoid Polyhedra, Partially Indexable Scheduling Problems, and Extended Restless Bandit Allocation Indices

We present a polyhedral framework for establishing general structural properties on optimal solutions of stochastic scheduling problems, where multiple job classes vie for service resources: the existence of an optimal priority policy in a given family, characterized by a greedoid (whose feasible class subsets may receive higher priority), where optimal priorities are determined by class-ranking indices, under restricted linear performance objectives (partial indexability). This framework extends that of Bertsimas and Niño-Mora (1996), which explained the optimality of priority-index policies under all linear objectives (general indexability). We show that, if performance measures satisfy partial conservation laws (with respect to the greedoid), which extend previous generalized conservation laws, then the problem admits a strong LP relaxation over a so-called extended greedoid polytope, which has strong structural and algorithmic properties. We present an adaptive-greedy algorithm (which extends Klimov's) taking as input the linear objective coefficients, which (1) determines whether the optimal LP solution is achievable by a policy in the given family; and (2) if so, computes a set of class-ranking indices that characterize optimal priority policies in the family. In the special case of project scheduling, we show that, under additional conditions, the optimal indices can be computed separately for each project (index decomposition). We further apply the framework to the important restless bandit model (two-action Markov decision chains), obtaining new index policies, that extend Whittle's (1988), and simple sufficient conditions for their validity. These results highlight the power of polyhedral methods (the so-called achievable region approach) in dynamic and stochastic optimization.

Knowledge and attitudes of nurses towards alcohol and related problems: the impact of an educational intervention

An exploratory study of quasi-experimental approach that aimed to verify the impact of an educational intervention on attitudes and knowledge of nurses towards alcohol use and associated problems. The sample included 185 nurses, divided into two groups: 84 submitted to a training course and formed the experimental group. Data were collected through a knowledge survey and an attitude scale. The attitudes of the participants of both groups were positive. There were no significant differences between groups in relation to knowledge. The strongest predictors of positive attitudes were possessed preparation to act with chemical dependents (OR = 2.18), "have received increased workload during graduation on the theme, 'alcohol and other drugs'" (OR = 1.70), and "completed graduate school" (OR = 2.59). The educational intervention had a positive impact on the attitudes of nurses towards alcoholics, work and interpersonal relationships with such clientele.