Premium
Assessing psychometric properties of scales: a case study
Author(s) -
Marshall Andrea P.,
Fisher Murray J.,
Brammer Jillian,
Eustace Paula,
Grech Carol,
Jones Bronwyn,
Kelly Michelle
Publication year - 2007
Publication title -
journal of advanced nursing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.948
H-Index - 155
eISSN - 1365-2648
pISSN - 0309-2402
DOI - 10.1111/j.1365-2648.2007.04316.x
Subject(s) - construct validity , scale (ratio) , construct (python library) , psychology , confirmatory factor analysis , test (biology) , factoring , sample (material) , population , nursing research , psychometrics , applied psychology , nursing , clinical psychology , structural equation modeling , medicine , computer science , statistics , mathematics , chemistry , biology , paleontology , chromatography , quantum mechanics , programming language , physics , environmental health , finance , economics
Title. Assessing psychometric properties of scales: a case studyAim. This paper is a report of a study to examine the construct validity of The Nursing Students’ Attitudes and Awareness of Research and Development within Nursing Scale. Background. The validity of instruments is critical in ensuring that data collected are sound and that the data measures what it purports to measure. When a new instrument is used in a different population or when it has been modified, it is useful to re‐examine the construct validity of the instrument. Method. A survey design was used in September 2004 with a sample of 615 undergraduate nursing students to test the factor structure of The Nursing Students’ Attitudes and Awareness of Research and Development within Nursing Scale and to estimate its similarity to the factor structure reported for the original scale developed and tested in a group of Registered Nurses. Results. Using Maximum Likelihood Factor Analysis and then Principal Axis Factoring, we were unable to obtain a similar factor structure to that originally identified for the scale. Our data resulted in a two‐factor structure. One factor consisted of 16 items that reflected a positive attitude to nursing research and the other consisted of 14 items that reflected a negative attitude to nursing research. Conclusion. The substantially different factor structure identified suggests that this scale requires further refinement and testing. This case study highlights the importance of a systematic and comprehensive approach to determining construct validity of scales, thus enabling researchers to determine their suitability as data collection instruments.