A Generalized Expression for the Reliability of Measures

Horst, Paul

doi:10.1007/bf02290137

Cited by 100 publications

(34 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The symptomatic changes when assessed by the patient himself (table 4) yielded similar means. In addition, clinical experts rated the improvement of all indi vidual symptomatic criteria of all patients attaining an interrater reliability of >0.70 [Horst, 1949], and the following means in our rating scale: PT = 2.27, CC = 2.26 and If we compare the three subgroups con cerning the patient's self-assessment, their ability to cope with life as a more general indicator for improvement and well-being, we obtain the same results: no statistically significant differences between the groups though in the self-assessment NT achieved a less advantageous mean (2.11 rating mean against 1.54 for CC and 1.69 for PT).…”

Section: Lin Ic a L R Esu Ltsmentioning

confidence: 99%

12-Year Foliow-Up Study of the Hamburg Short Psychotherapy Experiment: An Overview

Meyer¹,

Stuhr²,

Wirth³

et al. 1988

Psychother Psychosom

View full text Add to dashboard Cite

In 1971/72 eighty-seven psychoneurotic and psychosomatic patients of the Psychosomatic Outpatient Department at the University Clinic of Hamburg-Eppendorf attained an indication for short-term psychotherapy. Nineteen patients did not accept our offer despite a sufficient prognosis; they form the nontreated control group. The remaining 68 patients were assigned at random to a psychodynamic conflict-centered (n = 34) or to a client-centered therapy (n = 34), each with a maximum of 30 sessions. In spite of this long interval of 12 years we gained information about 82.5% of our patients. The follow-up program consisted of a multidimensional approach to describe the posttreatment development of our patients: psychological tests, self-rating by the patient and assessments by the follow-up interviewer and independent ratings of experts. The efficacy of short psychotherapy is seen in the test scale ‘extraversion’ for both treatment groups, for ‘depression’ only for the client-centered therapy; for the psychodynamic therapy there is a tendency for late changes. The clinical ratings show no significant differences between the treatment and the control groups. Sex combination in the therapeutic dyads (i.e. same sex vs. opposite sex) is important for patients’ posttherapeutic development.

show abstract

Section: Lin Ic a L R Esu Ltsmentioning

confidence: 99%

12-Year Foliow-Up Study of the Hamburg Short Psychotherapy Experiment: An Overview

Meyer¹,

Stuhr²,

Wirth³

et al. 1988

Psychother Psychosom

View full text Add to dashboard Cite

show abstract

“…Subsumed under the internal consistency classification are the widely used split-half (i~), the Kuder-Richardson formulas (11), and the less well known Hoyt (9) and Horst (8) techniques. Under certain assumptions these four approaches yield similar estimates of reliability, as has been shown in the literature (3, 5, 8, I 5) .…”

mentioning

confidence: 99%

The Reliability of Letter Grades

Bendig

1953

Educational and Psychological Measurement

View full text Add to dashboard Cite

A professional problem that has been rather consistently ignored by both psychologists and educators concerns methods of determining the reliability of letter grades in single college courses. The stability of quality point averages and of other composite indices of academic achievement over various time periods has been investigated. The most recent study was reported by Clark (2). However, studies of the reliability of the single course letter grades comprising such achievement measures have yet to be reported. Such investigations are vitally needed for two reasons. One is that we need to know, as educators, how reliable are the achievement ratings that we give our students, especially as these ratings become part of the official academic record of the student and determine, in part at least, his future educational, employment, and military opportunities. Grading procedures found by such researches to lack reliability can then be modified to achieve a higher degree of consistency in the academic assessment. A second reason for such investigations is the widespread use in educational research of academic grades as predictor or criterion variables. All too often the disappointed psychologist who fails to find a high correlation between his pet freshmen entrance examination and later college grades glibly attributes the lack of relationship to the known unreliability of letter grades. The previously mentioned lack of evidence on grade reliability may be saving the professional reputation of many widely used freshmen entrance tests. In describing a &dquo;Grade&dquo; factor that has appeared in several factor analytic studies including academic marks as variables, French suggests (6) that this factor may be an artifact attributable to differences in the reliability of grading between different academic courses.

show abstract

“…T h e reliability coefficient, rii, developed by Horst ( 2 ) , is a measure of the degree to which we might expect another set of three subgroups to yield the same average ranks. T h u s the following reliability coefficients are measures of the conDownloaded by [University of Toronto Libraries] at 11:57 03 December 2014 and Education/Occupation Relationship, .74.…”

Section: Resultsmentioning

confidence: 99%