Search Results

You are looking at 31 - 40 of 90 items for :

  • "imputation" x
Clear All
Restricted access

Sarahjane Belton, Gavin Breslin, Stephen Shannon, Wesley O’Brien, Ben Fitzpatrick, Tandy Haughey, Fiona Chambers, Danielle Powell, Darryl McCullagh and Deirdre Brennan

authors are confident that the valid sample remained representative, given there were no significant differences in age or BMI between the original and valid sample. Further research may consider conducting imputation methods for improving sample size, such as multiple imputation or expectation

Restricted access

Chris Knoester and Theo Randolph

’s reports of relationship closeness, and d) child’s reports of relationship closeness. In all models, multiple imputation with chained equations is used to account for missing data. This is a preferred method for addressing missing data ( Johnson & Young, 2011 ), yet the presented results are robust to the

Restricted access

Janet Robertson, Eric Emerson, Susannah Baines and Chris Hatton

for the outcome “frequent participation in sport/exercise,” the strength of association between sociodemographic factors, and participation separately for participants with and without intellectual disability. Missing data among sociodemographic variables were imputed using multiple imputation

Restricted access

James J. Annesi

provided monthly to reinforce self-regulatory skills usage. All assessments were completed by study staff in a private area. Data Analyses After determining that the 11% of missing cases were missing at random, 15 the expectation–maximization algorithm 16 was applied for imputation. Based on an effect

Restricted access

Jonathan M. Miller, Mark A. Pereira, Julian Wolfson, Melissa N. Laska, Toben F. Nelson and Dianne Neumark-Sztainer

Institute, Cary, NC) Proc MI and MIANALYZE. Twenty imputed data sets were created using a Markov Chain Monte Carlo algorithm and all correlates and demographics. Multiple imputation methods for EAT-2010 have been previously described in greater detail. 22 Statistical Analyses All statistical analyses were

Restricted access

Jakob Tarp, Anna Bugge, Niels Christian Møller, Heidi Klakk, Christina Trifonov Rexen, Anders Grøntved and Niels Wedderkopp

missing data (n = 697). A flowchart of participants is available in Supplementary Figure S1 and multiple imputation analysis including all 1209 consenting participants at baseline is available in Supplementary Table S4 (available online). Results from imputed data yielded similar conclusions as

Restricted access

Jocelyn Kernot, Lucy Lewis, Tim Olds and Carol Maher

the trial remained in the sample for analysis. 43 Multiple imputation (fully conditional specification—as testing showed that missing data were not random) was implemented to account for missing data. Data were combined using Rubin’s rule. 44 To determine the effectiveness of the MSIU program, random

Restricted access

Jay Johnson, Michelle D. Guerrero, Margery Holman, Jessica W. Chin and Mary Anne Signer-Kroeker

imputation procedures using the fully conditioned specification approach. A total of five imputed data sets were generated, with each data set comprising slightly different imputed values. The parameter estimates of the five imputations were pooled and yielded a single set of estimates. Given that little is

Full access

Philippa M. Dall, Dawn A. Skelton, Manon L. Dontje, Elaine H. Coulter, Sally Stewart, Simon R. Cox, Richard J. Shaw, Iva Čukić, Claire F. Fitzsimons, Carolyn A. Greig, Malcolm H. Granat, Geoff Der, Ian J. Deary, Sebastien F.M. Chastin and On behalf of the Seniors USP Team

days were analyzed and therefore no data imputation was conducted. Attachment of the monitor with single-use attachment materials and removal of the monitor by a researcher allowed a high level of certainty of continuous monitor wear. Although it was possible that a monitor that was still worn on

Restricted access

Yara Fidelix, Mara C. Lofrano-Prado, Leonardo S. Fortes, James O. Hill, Ann E. Caldwell, João P. Botero and Wagner L. do Prado

for gender and baseline BMI. The procedures were performed following an intention-to-treat principle without data imputation. In addition, the effect size (ES) for each intervention was calculated as proposed by Cohen ( d ). The significance level was set at P  < .05. Statistical procedures were