Main content

Home

Menu

Loading wiki pages...

View
Wiki Version:
In this component, analysis scripts and secondary (aggregated) dataset are provided. For a description of the original analysis plan for each study please refer to the Confirmatory Analysis Plan section of the [ManyLabsProposal_V1.1.docx][1]. Please see the [Web Supplement][2] for the final analyses and notes of any deviations from the pre-registration. The [Syntax.Manylabs.sps][3] file is a script that works in SPSS 20. This code should be run first on the de-identified datafile (**Full_Dataset_De-Identified.sav**) provided in the [data component][4]. In this file, local paths have to be added when necessary (see comments marked by asterisks in the syntax file). This code generates confirmatory tests for all the studies and two SPSS data files named "effectsizes.all.sav" and "effectsizes.overall.sav". These files contain overall and disaggregated effect sizes for each study/sample and 99% asymmetric central CIs around the standardized mean difference. This file also generates the SPSS data file named "ManylabsData.withFactor&DVs.sav" which is the source file for moderation analyses (syntax files: "ModerationAnalyses(lab&web&flagpriming).sps" and "Order.effects.moderations.sps"). The file ICCs.sps estimates Intra-class Correlation Coefficients for both samples and effects. Original effect sizes, 95% CIs around them, and 99% noncentral CIs around the overall effects can be generated using the R code named "NONcentralCIoriginal&replicationstudies.R" and the data file called ML.xls. Some additional scripts may be found on [GitHub][5], including R scripts for analyzing heterogeneity, and for computing effect size + confidence intervals for the original effects. [1]: https://osf.io/6wijr/ [2]: https://osf.io/y36m8/ [3]: https://osf.io/er9xg/ [4]: https://osf.io/pqf9r/ [5]: https://github.com/ManyLabsOpenScience/ManyLabs1