It is a false, but common, belief that statistical significance testing evaluates result replicability. In truth, statistical significance testing reveals nothing about results replicability. Since science is based on replication of results, methods that assess replicability are important. This is particularly true when multivariate methods, which…

The Likert rating scale procedure is often used in conjunction with a graded disagree-agree response scale to measure attitudes. Item characteristic curves associated with graded disagree-agree responses are generally single-peaked, nonmonotonic functions of true attitude. These characteristics are, thus, more generally consistent with an…

Considered the statistical comparison of groups under the condition of equal sample sizes. (Author)

A sampling theorem, based on the binomial probability distribution but otherwise distribution-free, is derived and some of its applications are illustrated numerically. (Author)

Work has been continuing on the correlation spectrometry techniques described at previous remote sensing symposiums. Advances in the techniques are described which enable accurate quantitative measurements of diffused atmospheric gases to be made using controlled light sources, accurate quantitative measurements of gas clouds relative to…

A study was conducted which has resulted in the development of tentative sampling and analysis of fluorides emitted from various stationary sources. The study was directed toward developing and understanding the kinds of species which are present in each source emission. The report presents the following information: review of the various unit…

The nature of the approximate chi-square test for hypotheses concerning multinomial probabilities is reviewed. Also, a BASIC computer program for calculating the sample size necessary to control for both Type I and Type II errors in chi-square tests for hypotheses concerning multinomial probabilities is described.

Four possible situations in trend analyses in the analysis of variance are discussed: unequal sample sizes with unequal intervals, unequal sample sizes with equal intervals, equal sample sizes with unequal intervals, and equal sample sizes with equal intervals. (Author/JKS)

In some linear regression problems samples may be unwittingly drawn from a heterogeneous population. The exploratory procedure described in this paper can be used to detect heterogeneity in a regression problem associated with the intercept. (Author)

A test for multisample sphericity based on the efficient scores criterion is obtained as an alternative to the likelihood ratio test developed by Mendoza. (Author)