A predicted link for a pair of players results if their CF score is larger than the 90% quantile of the CF scores computed on all observed link pairs, or the TF score is larger than 99.99% quantile of the TF scores computed on all observed link pairs.
The final model gave a good prediction on all observed data [ 144].
Exploratory analyses examined good responders according to the European League Against Rheumatism (EULAR) 22 with odds ratios and the development of remission based on disease activity scores for 28 joints 23 with log rank tests on all observed data.
Multiple imputation [ 10] and model-based approaches, such as mixed models [ 11] and weighted generalized estimating equations (GEE) [ 12] for repeatedly measured outcomes, based on all observed data can be valid and unbiased methods for MAR data, as long as the models are specified correctly.
The analytic approach to handle missing data will be anchored on the assumption that data is missing at random (MAR), where the probability of missing can depend on all observed information such as measured weights and covariates but does not depend on variables that are not recorded.
In the model one latent factor (F, represented by a circle) influences on all observed AP measures (represented by squares), explaining the shared variance across ages and raters.
