Ari Voutilainen1, Terhi Saaranen2, Marjorita Sormunen2. 1. Department of Nursing Science, University of Eastern Finland, Yliopistonranta 1C, Kuopio, Finland. Electronic address: ari.voutilainen@uef.fi. 2. Department of Nursing Science, University of Eastern Finland, Yliopistonranta 1C, Kuopio, Finland.
Abstract
BACKGROUND: By and large, in health professions training, the direction of the effect of e-learning, positive or negative, strongly depends on the learning outcome in question as well as on learning methods which e-learning is compared to. In nursing education, meta-analytically generated knowledge regarding the comparisons between conventional and e-learning is scarce. OBJECTIVES: The aim of this review is to discover the size of the effect of e-learning on learning outcomes in nursing education and to assess the quality of studies in which e-learning has been compared to conventional learning. METHODS: A systematic search of six electronic databases, PubMed, Ovid MEDLINE®, CINAHL (EBSCOhost), Cochrane Library, PsycINFO, and ERIC, was conducted in order to identify relevant peer-reviewed English language articles published between 2011 and 2015. The quality of the studies included as well as the risk of bias in each study was assessed. A random-effects meta-analysis was performed to generate a pooled mean difference in the learning outcome. RESULTS: Altogether, 10 studies were eligible for the quality assessment and meta-analysis. Nine studies were evaluated as good quality studies, but not without a risk of bias. Performance bias caused a high risk in nearly all the studies. In the meta-analysis, an e-learning method resulted in test scores that were, on average, five points higher than a conventional method on a 0-100 scale. Heterogeneity between the studies was very large. CONCLUSIONS: The size and direction of the effect of a learning method on learning outcomes appeared to be strongly situational. We suggest that meta-regressions should be performed instead of basic meta-analyses in order to reveal factors that cause variation in the learning outcomes of nursing education. It might be necessary to perform separate meta-analyses between e-learning interventions aimed at improving nursing knowledge and those aimed at improving nursing skills.
BACKGROUND: By and large, in health professions training, the direction of the effect of e-learning, positive or negative, strongly depends on the learning outcome in question as well as on learning methods which e-learning is compared to. In nursing education, meta-analytically generated knowledge regarding the comparisons between conventional and e-learning is scarce. OBJECTIVES: The aim of this review is to discover the size of the effect of e-learning on learning outcomes in nursing education and to assess the quality of studies in which e-learning has been compared to conventional learning. METHODS: A systematic search of six electronic databases, PubMed, Ovid MEDLINE®, CINAHL (EBSCOhost), Cochrane Library, PsycINFO, and ERIC, was conducted in order to identify relevant peer-reviewed English language articles published between 2011 and 2015. The quality of the studies included as well as the risk of bias in each study was assessed. A random-effects meta-analysis was performed to generate a pooled mean difference in the learning outcome. RESULTS: Altogether, 10 studies were eligible for the quality assessment and meta-analysis. Nine studies were evaluated as good quality studies, but not without a risk of bias. Performance bias caused a high risk in nearly all the studies. In the meta-analysis, an e-learning method resulted in test scores that were, on average, five points higher than a conventional method on a 0-100 scale. Heterogeneity between the studies was very large. CONCLUSIONS: The size and direction of the effect of a learning method on learning outcomes appeared to be strongly situational. We suggest that meta-regressions should be performed instead of basic meta-analyses in order to reveal factors that cause variation in the learning outcomes of nursing education. It might be necessary to perform separate meta-analyses between e-learning interventions aimed at improving nursing knowledge and those aimed at improving nursing skills.
Authors: Sunhea Choi; Ho Ming Yuen; Reginald Annan; Trevor Pickup; Andy Pulman; Michele Monroy-Valle; Nana Esi Linda Aduku; Samuel Kyei-Boateng; Carmen Isabel Velásquez Monzón; Carmen Elisa Portillo Sermeño; Andrew Penn; Ann Ashworth; Alan A Jackson Journal: J Med Internet Res Date: 2018-10-03 Impact factor: 5.428