Show simple item record

dc.contributor.authorHammond, Sean M
dc.contributor.authorO'Rourke, Margaret
dc.contributor.authorKelly, Martina
dc.contributor.authorBennett, Deirdre
dc.contributor.authorO'Flynn, Siun
dc.date.accessioned2012-02-14T14:11:49Z
dc.date.available2012-02-14T14:11:49Z
dc.date.issued2012-01-12
dc.identifierhttp://dx.doi.org/10.1186/1472-6920-12-2
dc.identifier.citationBMC Medical Education. 2012 Jan 12;12(1):2
dc.identifier.urihttp://hdl.handle.net/10147/210729
dc.description.abstractAbstract Background The quality of the Educational environment is a key determinant of a student centred curriculum. Evaluation of the educational environment is an important component of programme appraisal. In order to conduct such evaluation use of a comprehensive, valid and reliable instrument is essential. One of most widely used contemporary tools for evaluation of the learning environment is the Dundee Ready Education Environment Measure (DREEM). Apart from the initial psychometric evaluation of the DREEM, few published studies report its psychometric properties in detail. The aim of this study was to examine the psychometric quality of the DREEM measure in the context of medical education in Ireland and to explore the construct validity of the device. Methods 239 final year medical students were asked to complete the DREEM inventory. Anonymised responses were entered into a database. Data analysis was performed using PASW 18 and confirmatory factor analysis performed. Results Whilst the total DREEM score had an acceptable level of internal consistency (alpha 0.89), subscale analysis shows that two subscales had sub-optimal internal consistency. Multiple group confirmatory factor analysis (using Fleming's indices) shows an overall fit of 0.76, representing a weak but acceptable level of fit. 17 of the 50 items manifest fit indices less than 0.70. We sought the best fitting oblique solution to the 5-subscale structure, which showed large correlations, suggesting that the independence of the separate scales is open to question. Conclusions There has perhaps been an inadequate focus on establishing and maintaining the psychometric credentials of the DREEM. The present study highlights two concerns. Firstly, the internal consistency of the 5 scales is quite variable and, in our sample, appears rather low. Secondly, the construct validity is not well supported. We suggest that users of the DREEM will provide basic psychometric appraisal of the device in future published reports.
dc.titleA psychometric appraisal of the DREEM
dc.typeJournal Article
dc.language.rfc3066en
dc.rights.holderHammond et al.; licensee BioMed Central Ltd.
dc.description.statusPeer Reviewed
dc.date.updated2012-02-10T20:05:58Z
refterms.dateFOA2018-08-22T15:46:45Z
html.description.abstractAbstract Background The quality of the Educational environment is a key determinant of a student centred curriculum. Evaluation of the educational environment is an important component of programme appraisal. In order to conduct such evaluation use of a comprehensive, valid and reliable instrument is essential. One of most widely used contemporary tools for evaluation of the learning environment is the Dundee Ready Education Environment Measure (DREEM). Apart from the initial psychometric evaluation of the DREEM, few published studies report its psychometric properties in detail. The aim of this study was to examine the psychometric quality of the DREEM measure in the context of medical education in Ireland and to explore the construct validity of the device. Methods 239 final year medical students were asked to complete the DREEM inventory. Anonymised responses were entered into a database. Data analysis was performed using PASW 18 and confirmatory factor analysis performed. Results Whilst the total DREEM score had an acceptable level of internal consistency (alpha 0.89), subscale analysis shows that two subscales had sub-optimal internal consistency. Multiple group confirmatory factor analysis (using Fleming's indices) shows an overall fit of 0.76, representing a weak but acceptable level of fit. 17 of the 50 items manifest fit indices less than 0.70. We sought the best fitting oblique solution to the 5-subscale structure, which showed large correlations, suggesting that the independence of the separate scales is open to question. Conclusions There has perhaps been an inadequate focus on establishing and maintaining the psychometric credentials of the DREEM. The present study highlights two concerns. Firstly, the internal consistency of the 5 scales is quite variable and, in our sample, appears rather low. Secondly, the construct validity is not well supported. We suggest that users of the DREEM will provide basic psychometric appraisal of the device in future published reports.


Files in this item

Thumbnail
Name:
1472-6920-12-2.xml
Size:
89.25Kb
Format:
XML
Thumbnail
Name:
1472-6920-12-2.pdf
Size:
196.3Kb
Format:
PDF

This item appears in the following Collection(s)

Show simple item record