2.50
Hdl Handle:
http://hdl.handle.net/10755/161499
Type:
Presentation
Title:
A Tutorial on Sample Power Analysis for MANOVA
Abstract:
A Tutorial on Sample Power Analysis for MANOVA
Conference Sponsor:Midwest Nursing Research Society
Conference Year:2003
Author:Ryan-Wenger, Nancy
Contact Address:CON, 1585 Neil Avenue, Columbus, OH, 43210, USA
Co-Authors:Anne M. Mentro
Single variables rarely explain phenomena of interest in nursing research, therefore many research proposals involve more than one dependent variable (DV) and one or more independent variable (IV). By calculating one over-all analysis versus a series of individual ANOVAs or t-tests, MANOVA, a special case of multiple regression and set correlation analysis, prevents accumulative alpha error from DVs that are likely to be correlated, and thus increases power by minimizing alpha error. In addition, MANOVA takes the interaction of DVs into account, i.e. the effect of a single DV may not be visible except in combination with one or more other DVs. Finally, repeated measures MANOVA allows for examination of changes over time. Power analysis to determine optimum sample size is required for any research that proposes to use inferential statistics to ensure that: 1) an effect will be observed if it exists; 2) time and research resources are maximized; and 3) statistical results are meaningful. Although power analysis is frequently done for univariate statistics, power analysis for MANOVA is difficult to perform and is not available on most statistical software packages. This presentation provides an explanation of the logic underlying sample size analysis for multivariate statistics according to Cohen. This method takes into account the varying number of levels within each IV and DV. Step-by-step instructions to meet each specific aim, research question or hypothesis include: 1) Determination of the type of association and analysis required; 2) determination of the error model that fits the type of association and analysis; 3) selection of effect size, alpha and power; and 4) application of formulas to calculate a total sample size. Using appropriate power analysis methods for multivariate statistics will strengthen nursing research studies. Recommendations for reporting post hoc power analyses in research reports will be given. AN: MN030069
Repository Posting Date:
26-Oct-2011
Date of Publication:
17-Oct-2011
Sponsors:
Midwest Nursing Research Society

Full metadata record

DC FieldValue Language
dc.typePresentationen_GB
dc.titleA Tutorial on Sample Power Analysis for MANOVAen_GB
dc.identifier.urihttp://hdl.handle.net/10755/161499-
dc.description.abstract<table><tr><td colspan="2" class="item-title">A Tutorial on Sample Power Analysis for MANOVA </td></tr><tr class="item-sponsor"><td class="label">Conference Sponsor:</td><td class="value">Midwest Nursing Research Society</td></tr><tr class="item-year"><td class="label">Conference Year:</td><td class="value">2003</td></tr><tr class="item-author"><td class="label">Author:</td><td class="value">Ryan-Wenger, Nancy</td></tr><tr class="item-address"><td class="label">Contact Address:</td><td class="value">CON, 1585 Neil Avenue, Columbus, OH, 43210, USA</td></tr><tr class="item-co-authors"><td class="label">Co-Authors:</td><td class="value">Anne M. Mentro</td></tr><tr><td colspan="2" class="item-abstract">Single variables rarely explain phenomena of interest in nursing research, therefore many research proposals involve more than one dependent variable (DV) and one or more independent variable (IV). By calculating one over-all analysis versus a series of individual ANOVAs or t-tests, MANOVA, a special case of multiple regression and set correlation analysis, prevents accumulative alpha error from DVs that are likely to be correlated, and thus increases power by minimizing alpha error. In addition, MANOVA takes the interaction of DVs into account, i.e. the effect of a single DV may not be visible except in combination with one or more other DVs. Finally, repeated measures MANOVA allows for examination of changes over time. Power analysis to determine optimum sample size is required for any research that proposes to use inferential statistics to ensure that: 1) an effect will be observed if it exists; 2) time and research resources are maximized; and 3) statistical results are meaningful. Although power analysis is frequently done for univariate statistics, power analysis for MANOVA is difficult to perform and is not available on most statistical software packages. This presentation provides an explanation of the logic underlying sample size analysis for multivariate statistics according to Cohen. This method takes into account the varying number of levels within each IV and DV. Step-by-step instructions to meet each specific aim, research question or hypothesis include: 1) Determination of the type of association and analysis required; 2) determination of the error model that fits the type of association and analysis; 3) selection of effect size, alpha and power; and 4) application of formulas to calculate a total sample size. Using appropriate power analysis methods for multivariate statistics will strengthen nursing research studies. Recommendations for reporting post hoc power analyses in research reports will be given. AN: MN030069 </td></tr></table>en_GB
dc.date.available2011-10-26T23:22:21Z-
dc.date.issued2011-10-17en_GB
dc.date.accessioned2011-10-26T23:22:21Z-
dc.description.sponsorshipMidwest Nursing Research Societyen_GB
All Items in this repository are protected by copyright, with all rights reserved, unless otherwise indicated.