Beyond Alpha: Lower Bounds for the Reliability of Tests

Bendermacher, Nol

doi:10.22237/jmasm/1272687000

Cited by 23 publications

(14 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We estimated score reliability with Guttman's (1945) λ2 and λ3 as well as 8-week testretest reliability. Guttman's λ3 or coefficient α, often incorrectly attributed to Cronbach (1951;Sijtsma, 2009), has advantages in ease of estimation (Guttman, 1945) but does not perform as well compared with λ2 (Bendermacher, 2010;Guttman, 1945;Sijtsma, 2009). We reported λ3 (coefficient α), however, due to its widespread use.…”

Section: Discussionmentioning

confidence: 99%

Validation of the elementary social behaviour assessment: teacher ratings of students’ social skills adapted to Norwegian, grades 1–6

Arnesen

Smolkowski

Ogden

et al. 2017

Emotional and Behavioural Difficulties

View full text Add to dashboard Cite

This study investigated the psychometric properties of the Elementary Social Behavior Assessment (ESBA), a teacher-report measure of students' social skills and academic engagement adapted for use in Norwegian schools. At two times, 8 weeks apart, 151 teachers rated 793 students in Grades 1-6 on the ESBA and the Social Skills Rating System Teacher's form (SSRS-T). Exploratory factor analysis with the first assessment suggested one or possibly two factors. Confirmatory factor analysis with the second assessment confirmed the one-or twofactor models but did not offer strong evidence for one over the other. The ESBA demonstrated good to excellent reliability. Strong concurrent and predictive correlations with the SSRS-T established criterion validity. The school and teacher levels explained only a limited amount of the total variance in ESBA scores, suggesting that teachers rate students quite similarly. The Norwegian-adapted ESBA paralleled the reliability and validity evidence from US samples.

show abstract

Section: Discussionmentioning

confidence: 99%

Validation of the elementary social behaviour assessment: teacher ratings of students’ social skills adapted to Norwegian, grades 1–6

Arnesen

Smolkowski

Ogden

et al. 2017

Emotional and Behavioural Difficulties

View full text Add to dashboard Cite

show abstract

“…Betrouwbaarheid bekijkt de onderlinge samenhang in relatie tot het aantal items (0,80 is de ondergrens) (31). Voor betrouwbaarheid berekenen we Cronbachs alfa, en een verbeterde versie daarvan die de 'hoogste ondergrens' wordt genoemd (GLB) (32). Soortgenootvaliditeit is er wanneer een meting voldoende samenhangt met andere variabelen waarvan aannemelijk is dat deze hetzelfde concept meten (33).…”

Section: Methodsunclassified

Hogere eenzaamheid onder Marokkaanse en Turkse ouderen in Nederland: Op zoek naar een verklaring

Tilburg

Fokkema

2018

Tijdschr Gerontol Geriatr

View full text Add to dashboard Cite

Eenzaamheid komt vaker voor onder ouderen van Marokkaanse en Turkse herkomst dan van Nederlandse herkomst. Twee verklaringen voor dit verschil zijn onderzocht. (1) Het concept en de meting verschillen tussen de herkomstgroepen. (2) Migranten verkeren in een kwetsbare situatie. Er is gebruik gemaakt van gegevens van de Longitudinal Aging Study Amsterdam. Interviews zijn gehouden met 176 mensen geboren in Marokko en 235 geboren in Turkije, in de leeftijd van 55-66 jaar, en wonend in stedelijke wijken. Gemiddeld migreerden zij 35 jaar geleden. Zij zijn vergeleken met een steekproef van 292 ouderen die, evenals hun ouders, geboren zijn in Nederland. De analyses laten zien dat de psychometrische eigenschappen van de eenzaamheidsschaal bevredigend zijn. Voor enkele items is er systematische vertekening. Migrantenouderen hebben meer sociale contacten, maar participeren minder, zijn minder tevreden over hun inkomen, ervaren minder regie en hebben een slechtere gezondheid waaronder een groter aantal depressieve symptomen. Als rekening wordt gehouden met deze verschillen halveert het verschil in eenzaamheid met Nederlandse ouderen. De ervaringen vanuit migratie en als minderheid in Nederland versterken waarschijnlijk hun gevoelens van eenzaamheid. Interventies moeten zich niet richten op het verbeteren van sociaal contact, maar bijvoorbeeld op het versterken van het ervaren van een sociaal gewaardeerde rol en het vermijden van negatieve interpretaties. Stronger feelings of loneliness among Moroccan and Turkish older adults in the Netherlands: A search for an explanation The prevalence of loneliness among Turkish-and Moroccan-Dutch older adults is higher than among Dutch older adults of non-migrant origin. Two explanations may account for this difference. (1) The meaning of the concept may differ, or there is differential item functioning. This might result in scores that not only differ in intensity but also in meaning across groups. (2) The position of older migrants is much more vulnerable than of non-migrant older people. Data from the Longitudinal Aging Study Amsterdam were used to examine support for both explanations. Feelings of loneliness are explored among 176 people born in Morocco and 235 people born in Turkey, aged 55-66 years, and living in urban areas. They migrated on average 35 years ago to the Netherlands. They are compared with a matched sample of 292 older people of Dutch origin. The psychometric properties of the loneliness scale are satisfying, although there is some differential item functioning. Older migrants have more frequent social contacts, but are at a disadvantage in other domains. Taking into account differences in social participation, satisfaction with their income, mastery and depressive symptoms, the difference between older migrants' and non-migrants' loneliness is reduced to more than half. Being an older migrant and belonging to a minority might further contribute to feelings of loneliness. Interventions should not be directed at stimulating social contact, but rather, for example, at ...

show abstract

“…Attitudes questions were grouped into three factors to assess feelings about research (affect), the value placed on research (value) and the ability to conduct research (ability) (Table ). Cronbach's alpha and Guttman's lambda‐2 ( λ 2) assessed internal reliability and were ≥0.7 for all factors.…”

Section: Methodsmentioning

confidence: 99%

Attitudes towards and experiences with research: Differences between dietetics students and professionals in Australia and the United States

Tan

Hemmelgarn

Baumgardner

et al. 2016

Nutrition & Dietetics

View full text Add to dashboard Cite

show abstract

Beyond Alpha: Lower Bounds for the Reliability of Tests

Abstract: The most common lower bound to the reliability of a test is Cronbach's alpha. However, several lower bounds exist that are definitely better, that is, higher than alpha. An overview is given as well as an algorithm to find the best: the greatest lower bound.

Cited by 23 publications

References 12 publications

Validation of the elementary social behaviour assessment: teacher ratings of students’ social skills adapted to Norwegian, grades 1–6

Validation of the elementary social behaviour assessment: teacher ratings of students’ social skills adapted to Norwegian, grades 1–6

Hogere eenzaamheid onder Marokkaanse en Turkse ouderen in Nederland: Op zoek naar een verklaring

Attitudes towards and experiences with research: Differences between dietetics students and professionals in Australia and the United States

Contact Info

Product

Resources

About