The Use of Chance Corrected Percentage of Agreement to Interpret the Results of a Discriminant Analysis

Wiedemann, Carl F.; Fenster, C. Abraham

doi:10.1177/001316447803800105

Cited by 16 publications

(5 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Examination of the standardized discriminant function coefficients indicated that education and age were the most influential in determining group differences. Almost threefourths (73%) of the analysis group were correctly classified for the discriminant function (Table 3) with a significant Cohen's kappa (K = 5.6, p % 0.001) (Wiedemann and Fenster, 1978). Cross validation resulted in almost three-fourths (71%) correctly classified.…”

Section: Resultsmentioning

confidence: 96%

Accessibility and Ownership of Financial Accounts by Married U.S. Farm Women

Scholl

1985

Home Economics Research Journal

View full text Add to dashboard Cite

Married farm women's accessibility and ownership of farm and household credit and cash accounts was measured with a score constructed from various items in the 1980 National Farm Women Survey. The majority of women were found to be partially or not involved with the accounts, with variation among the women in regard to the degree of accessibility and ownership. Discriminant analysis indicated that farm women's education, age, and level of off‐farm income; stage of family life cycle; and the women's participation in farm decisionmaking, their perception of their farm involvement, and farm task participation distinguished between women with high scores and those with low scores. Attributes of farm women and characteristics which are unique to farm women in their interaction with their farms differentiated those who are financially involved with the farm and household accounts from those with little involvement. Farm characteristics, however, did not contribute to group differentiation.

show abstract

Section: Resultsmentioning

confidence: 96%

Accessibility and Ownership of Financial Accounts by Married U.S. Farm Women

Scholl

1985

Home Economics Research Journal

View full text Add to dashboard Cite

show abstract

“…Among those cases in the treatment completion group, 59.5% were correctly classified, and among those cases in the noncompletion treatment group, 65.8% were correctly classified. The Kappa coefficient, an index that accounts for chance agreements commonly used to assess the results of discriminant analysis procedures (Green & Salkind, 2005; Wiedemann & Fenster, 1978), was .24 ( p < .02), indicating fair accuracy in outcome group classification. The Wilks's lambda statistic was also used to test the differences across outcome group means for each predictor so that their effectiveness in differentiating among treatment completion/noncompletion within the model could be assessed separately.…”

Section: Resultsmentioning

confidence: 99%

Using fine-grained indexes of therapists' experience and training to predict treatment outcomes in a university-based training clinic for children and families.

Powell

Hunter

Beasley

et al. 2010

Training and Education in Professional Psychology

View full text Add to dashboard Cite

Over several decades, researchers have debated whether a clinician's training and experience are associated with treatment outcome. Recent studies have suggested that fine-grained variables representing training or experience are better predictors of clinical effectiveness than broader measures, such as level of education. The present study addressed whether a therapist's number of direct client contact hours, days in doctoral training, and number of settings in which services were delivered were positively associated with meeting treatment goals in a university-based training clinic serving children and families. Results suggested each of these variables significantly predicted successful treatment completion.

show abstract

“…The degree to which prediction of group membership exceeded chance was tested by means of kappa, a statistic that, when divided by its standard error, is analogous to a z score (Weidemann & Fenster, 1978). For the classification of cases in the discriminant‐function analysis of Survey I, !κ = .195, SE κ = .098, z = 3.98, p < .0001.…”

Section: Resultsmentioning

confidence: 99%

Sex Differences in Peer Responsiveness to Suicide Ideation

Wellman

1986

Suicide & Life Threat Behav

View full text Add to dashboard Cite

Two surveys assessing attitudes toward and beliefs about suicide were conducted with undergraduates. There were 473 participants in Survey I and 692 participants in Survey II. Sex differences were found in participants' receptivity to suicidal individuals, beliefs about the behavior of suicidal peers, attitudes concerning the worth of suicidal people, incidence of adolescent suicide, and morality of suicide. There was no sex difference in participants' own frequency or seriousness of suicide ideation or suicide attempts. Results are discussed in terms of hypotheses concerning traditional socialization of males and females, resulting in sex-role differentiation that may influence attitudes toward and beliefs about suicide.

show abstract

The Use of Chance Corrected Percentage of Agreement to Interpret the Results of a Discriminant Analysis

Cited by 16 publications

References 1 publication

Accessibility and Ownership of Financial Accounts by Married U.S. Farm Women

Accessibility and Ownership of Financial Accounts by Married U.S. Farm Women

Using fine-grained indexes of therapists' experience and training to predict treatment outcomes in a university-based training clinic for children and families.

Sex Differences in Peer Responsiveness to Suicide Ideation

Contact Info

Product

Resources

About