archive,category,category-r-code,category-24,bridge-core-2.4.3,ajax_fade,page_not_loaded,,qode_grid_1300,side_area_uncovered_from_content,overlapping_content,qode-content-sidebar-responsive,qode-theme-ver-22.8,qode-theme-bridge,disabled_footer_top,wpb-js-composer js-comp-ver-6.3.0,vc_responsive,elementor-default,elementor-kit-3194


R-code, Tips / 15.09.2018

We analyze within-subjects designs with repeated-measures regressions, aka random-effects models. Learn how to set up such models in R. This concerns analyzing data with grouping, clustering, aka. hierarchical data, data with correlated errors, or data with violations of sphericity.
Inter-rater reliability, R-code / 24.12.2014

Inter rater reliability

If you want to obtain inter-rater reliability measures for dichotomous ratings, by more than two raters, but not all raters rated all items, Fleiss and Cuzick (1979) will be the referece you'll find. For example, we asked researchers which model they consider a process model (dichotomous rating), and we asked about 60 researchers (more than two rater), of whom not everyone was familiar with every model (not all raters rated all items). They proposed a measure between a minimum value and 1. Here is an function to calculate their kappa measure in R.
R-code, Tips / 18.09.2014

In R's default round() functions, odd numbers ending with .5 are rounded up (1.5 becomes 2), but even numbers are rounded down (2.5 becomes 2, rather than 3). This is compliant with a norm on rounding as outlined in the R documentation, but sometimes inconvenient.