How Factor Analysis Results May Change Due to Country Context

Piotr Brzyski, Maria Kózka, Allison Squires, Tomasz Brzostek

Research output: Contribution to journalArticle

Abstract

Purpose: To present how factor analysis results of a Practice Environment Scale of the Nursing Work Index (PES-NWI) translation changed due to country context. Design: Validity and reliability analysis of a cross-culturally adapted, Polish translation of the PES-NWI came from a cross-sectional, national survey that included 2,605 registered nurses working in surgical (50.4%) and medical (49.6%) units of 30 Polish hospitals. Methods: Exploratory factor analysis (EFA) using the principal component analysis (PCA) method with varimax rotation and confirmatory factor analysis (CFA) was used to examine factor structure of the instrument in the Polish context. Zero-order and partial Pearson correlation coefficients were used to establish the range of variance shared by the dimensions of the Polish version. The Cronbach's alpha coefficient determined internal consistency reliability. The internal consistency of the scale was also tested based on Kline's criterion. Findings: The PCA conducted in the sample of Polish nurses extracted six factors, explaining together 56% of the total variance. The varimax rotation, however, restricted results to five factors, explaining 52.7% of the total variance and generating a factor structure closer to that based on previous studies. The CFA model, based on a PCA solution with five nonorthogonal factors, fitted data better than the theoretically driven model. Conclusions: Dimensions of the PES-NWI nurses' work environments remain conceptually consistent in Poland, but load differently. Health system or nursing profession factors related to the country context are potential explanations for these differences. Clinical Relevance: When using a translation of the Practice Environment Scale of the Nursing Work Index-Revised to examine the influences of work environment quality on nursing and patient outcomes, it is important to consider contextual differences when using results to inform policy.

Original languageEnglish (US)
JournalJournal of Nursing Scholarship
DOIs
StateAccepted/In press - 2016

Fingerprint

Statistical Factor Analysis
Nursing
Principal Component Analysis
Nurses
Poland
Reproducibility of Results
Cross-Sectional Studies
Health

Keywords

  • Health services research
  • Nurses
  • PES-NWI
  • Poland
  • Survey research
  • Work environments

ASJC Scopus subject areas

  • Nursing(all)

Cite this

How Factor Analysis Results May Change Due to Country Context. / Brzyski, Piotr; Kózka, Maria; Squires, Allison; Brzostek, Tomasz.

In: Journal of Nursing Scholarship, 2016.

Research output: Contribution to journalArticle

@article{01614b3cde0b4bd3910a18bcade4912d,
title = "How Factor Analysis Results May Change Due to Country Context",
abstract = "Purpose: To present how factor analysis results of a Practice Environment Scale of the Nursing Work Index (PES-NWI) translation changed due to country context. Design: Validity and reliability analysis of a cross-culturally adapted, Polish translation of the PES-NWI came from a cross-sectional, national survey that included 2,605 registered nurses working in surgical (50.4{\%}) and medical (49.6{\%}) units of 30 Polish hospitals. Methods: Exploratory factor analysis (EFA) using the principal component analysis (PCA) method with varimax rotation and confirmatory factor analysis (CFA) was used to examine factor structure of the instrument in the Polish context. Zero-order and partial Pearson correlation coefficients were used to establish the range of variance shared by the dimensions of the Polish version. The Cronbach's alpha coefficient determined internal consistency reliability. The internal consistency of the scale was also tested based on Kline's criterion. Findings: The PCA conducted in the sample of Polish nurses extracted six factors, explaining together 56{\%} of the total variance. The varimax rotation, however, restricted results to five factors, explaining 52.7{\%} of the total variance and generating a factor structure closer to that based on previous studies. The CFA model, based on a PCA solution with five nonorthogonal factors, fitted data better than the theoretically driven model. Conclusions: Dimensions of the PES-NWI nurses' work environments remain conceptually consistent in Poland, but load differently. Health system or nursing profession factors related to the country context are potential explanations for these differences. Clinical Relevance: When using a translation of the Practice Environment Scale of the Nursing Work Index-Revised to examine the influences of work environment quality on nursing and patient outcomes, it is important to consider contextual differences when using results to inform policy.",
keywords = "Health services research, Nurses, PES-NWI, Poland, Survey research, Work environments",
author = "Piotr Brzyski and Maria K{\'o}zka and Allison Squires and Tomasz Brzostek",
year = "2016",
doi = "10.1111/jnu.12249",
language = "English (US)",
journal = "Journal of Nursing Scholarship",
issn = "1527-6546",
publisher = "Wiley-Blackwell",

}

TY - JOUR

T1 - How Factor Analysis Results May Change Due to Country Context

AU - Brzyski, Piotr

AU - Kózka, Maria

AU - Squires, Allison

AU - Brzostek, Tomasz

PY - 2016

Y1 - 2016

N2 - Purpose: To present how factor analysis results of a Practice Environment Scale of the Nursing Work Index (PES-NWI) translation changed due to country context. Design: Validity and reliability analysis of a cross-culturally adapted, Polish translation of the PES-NWI came from a cross-sectional, national survey that included 2,605 registered nurses working in surgical (50.4%) and medical (49.6%) units of 30 Polish hospitals. Methods: Exploratory factor analysis (EFA) using the principal component analysis (PCA) method with varimax rotation and confirmatory factor analysis (CFA) was used to examine factor structure of the instrument in the Polish context. Zero-order and partial Pearson correlation coefficients were used to establish the range of variance shared by the dimensions of the Polish version. The Cronbach's alpha coefficient determined internal consistency reliability. The internal consistency of the scale was also tested based on Kline's criterion. Findings: The PCA conducted in the sample of Polish nurses extracted six factors, explaining together 56% of the total variance. The varimax rotation, however, restricted results to five factors, explaining 52.7% of the total variance and generating a factor structure closer to that based on previous studies. The CFA model, based on a PCA solution with five nonorthogonal factors, fitted data better than the theoretically driven model. Conclusions: Dimensions of the PES-NWI nurses' work environments remain conceptually consistent in Poland, but load differently. Health system or nursing profession factors related to the country context are potential explanations for these differences. Clinical Relevance: When using a translation of the Practice Environment Scale of the Nursing Work Index-Revised to examine the influences of work environment quality on nursing and patient outcomes, it is important to consider contextual differences when using results to inform policy.

AB - Purpose: To present how factor analysis results of a Practice Environment Scale of the Nursing Work Index (PES-NWI) translation changed due to country context. Design: Validity and reliability analysis of a cross-culturally adapted, Polish translation of the PES-NWI came from a cross-sectional, national survey that included 2,605 registered nurses working in surgical (50.4%) and medical (49.6%) units of 30 Polish hospitals. Methods: Exploratory factor analysis (EFA) using the principal component analysis (PCA) method with varimax rotation and confirmatory factor analysis (CFA) was used to examine factor structure of the instrument in the Polish context. Zero-order and partial Pearson correlation coefficients were used to establish the range of variance shared by the dimensions of the Polish version. The Cronbach's alpha coefficient determined internal consistency reliability. The internal consistency of the scale was also tested based on Kline's criterion. Findings: The PCA conducted in the sample of Polish nurses extracted six factors, explaining together 56% of the total variance. The varimax rotation, however, restricted results to five factors, explaining 52.7% of the total variance and generating a factor structure closer to that based on previous studies. The CFA model, based on a PCA solution with five nonorthogonal factors, fitted data better than the theoretically driven model. Conclusions: Dimensions of the PES-NWI nurses' work environments remain conceptually consistent in Poland, but load differently. Health system or nursing profession factors related to the country context are potential explanations for these differences. Clinical Relevance: When using a translation of the Practice Environment Scale of the Nursing Work Index-Revised to examine the influences of work environment quality on nursing and patient outcomes, it is important to consider contextual differences when using results to inform policy.

KW - Health services research

KW - Nurses

KW - PES-NWI

KW - Poland

KW - Survey research

KW - Work environments

UR - http://www.scopus.com/inward/record.url?scp=84990821397&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84990821397&partnerID=8YFLogxK

U2 - 10.1111/jnu.12249

DO - 10.1111/jnu.12249

M3 - Article

JO - Journal of Nursing Scholarship

JF - Journal of Nursing Scholarship

SN - 1527-6546

ER -