Test of Independence Against a Class of Ordered Alternatives in an R × C Contingency Table

Test procedures are developed for testing the hypotheses of independence in a two-way contingency table against a class of ordered alternatives defined in terms of pooled crossproduct ratios. The alternatives hypotheses of “positive dependence” defines new notions of dependence. The distribution of the test statistics is obtained. Admissibility of the test is deduced and power comparison with some known test procedures are given. A numerical example is presented to illustrate the techniques developed.     

Artificial nest site preferences of Black-capped Chickadees

ABSTRACT.   To facilitate study of the breeding biology of parids, Grubb and Bronson (1995; Condor 97: 1067–1070) designed artificial "snags" made from polyvinyl chloride (PVC) tubes. Because the cost of artificial snags is greater than that of traditional wooden boxes, we examined alternatives to PVC snags for attracting chickadees to artificial nesting sites. From 2005 to 2007, we compared the use of PVC snags and wooden nest boxes by Black-capped Chickadees ( Poecile atricapillus ) in Sapsucker Woods Sanctuary in Ithaca, New York. We also quantified the use of cavities with and without wood shavings. The probability of chickadee excavation was greater (60–70% per yr) in filled snags (with wood shavings) than in filled boxes (40–50%; logistic mixed model, P = 0.01), and chickadees initiated more nests in filled snags (25–30%) than filled boxes (15%; P = 0.03). Chickadees also initiated significantly more nests ( P = 0.03) in filled than unfilled boxes. Although wooden boxes filled with wood shavings were used more often by Black-capped Chickadees than unfilled wooden boxes, artificial snags filled with wood shavings were used most, were no more likely than boxes to be usurped by House Wrens ( Troglodytes aedon ), and were less likely than boxes to be occupied by mice. Thus, artificial snags may be the better option for investigators studying the breeding biology of chickadees.     

The reproducibility of test contractions for calibration of electromyographic measurements

The main purpose of this study was to evaluate the reproducibility of electromyographic (EMG) measurements and specifically to test a calibration procedure with submaximal test contractions. Bipolar surface electrodes (20 mm fixed distance) were repositioned by a tracing sheet on both trapezius muscles, halfway between acromion and processus prominens. Submaximal test contractions were performed by keeping both arms straight abducted 90 degrees and forward flexed 10 degrees for 15-s periods. The arm position could be precisely reproduced in the frontal plane, but deviated forwards by 4 degrees in the horizontal plane, where the sensitivity of the EMG response to arm position was lowest. The electrodes were repositioned within a radius of 3 mm with a probability of 90%. Large deviations in the EMG response were found within this radius and a significant depression of the EMG response was recorded over the middle part of the muscle (the innervation zone?). This change in sensitivity of the EMG response with electrode position occurred in parallel for the test and maximal contractions. The total coefficient of variation was estimated to be 23% for recurrent EMG measurements using the calibration procedure described.