Showing papers on "Intraclass correlation published in 1972"
TL;DR: Six factors emerged from a factor analysis of the Structured and Scaled Interview to Assess Maladjustment (SSIAM) ratings made on 164 adults judged acceptable for outpatient psychotherapy: social isolation, work inadequacy, friction with family, dependence on family, sexual dissatisfaction, and friction outside family.
Abstract: Six factors emerged from a factor analysis of the Structured and Scaled Interview to Assess Maladjustment (SSIAM) ratings made on 164 adults judged acceptable for outpatient psychotherapy. These factors were social isolation, work inadequacy, friction with family, dependence on family, sexual dissatisfaction, and friction outside family. Their intraclass correlation coefficients of reliability ranged from 0.78 to 0.97. Other aspects of interrater agreement were examined, and agreement was found to be adequate. There was significant agreement (P
51 citations
01 Jun 1972-Cartographica: The International Journal for Geographic Information and Geovisualization
TL;DR: In this paper, the authors used rank correlation (Kendall's t) to measure the map correlation and drawing data from normal distributions, and found that there is a mathematically expressible curvilinear relationship between the product-moment correlation and the expected map correlation.
Abstract: The correlation between a pair of distributions is often measured with the product-moment correlation coefficient, r. But the choropleth maps of the same distributions may be more alike or less alike than that coefficient would indicate. The degree to which map likeness corresponds to the correlation between the data before mapping is affected by the class interval method used. Using rank correlation (Kendall's t) to measure the map correlation and drawing data from normal distributions, it is found that: (1) there is a mathematically expressible curvilinear relationship between the product-moment correlation and the expected map correlation; (2) this relationship appears to remain identical regardless of class interval system; (3) the standard deviation about the expected values does vary from one class interval system to another. Smaller standard deviations are associated with increased numbers of observations within each variable and with increased numbers of intervals. For a given r value, number of o...
6 citations
TL;DR: Although partial correlation is a correlation of residuals, the correlation of the true-score components of these residuals is not equivalent to the partial correlation of true scores themselves as mentioned in this paper, and this discrepancy is explained and its implications are briefly discussed.
3 citations
TL;DR: In this paper, the extent to which judges agree on their assignments is investigated, and the results of these categorical assignments may be used as a dependent or an independent variable for further research.
Abstract: MANY research problems require judges to assign objects or stimuli (often people) to discrete categories. The results of these categorical assignments may be used as a dependent or an independent variable for further research. Concerned about the reliability of these judgments, the investigator may wish to know, before proceeding with further study, the extent to which judges agree on their assignments.