KitchenhamMadeyskiBrereton.ReportedEffectSizes: KitchenhamMadeyskiBrereton.ReportedEffectSizes data

In reproducer: Reproduce Statistical Analyses and Meta-Analyses

KitchenhamMadeyskiBrereton.ReportedEffectSizes data

Description

This data is used in the paper: Barbara Kitchenham, Lech Madeyski and Pearl Brereton. Meta-analysis for Families of Experiments: A Systematic Review and Reproducibility Assessment, Empirical Software Engineering (2019) doi:10.1007/s10664-019-09747-0. This file holds the individual effect sizes for each experiment, as reported by 13 primary studies in the systematic review.

Usage

KitchenhamMadeyskiBrereton.ReportedEffectSizes

Format

A text file with variables:

Study

This field includes the study identifier of each of the 13 primary studies which were included in the systematic review.

Type

This identifies the type of effect size used by the study authors. d or g refer to dIG and gIG, p is the p-value used for aggregation, if the repeated measures estimate was obtained it is appropriately specified as gRM, r refers to the point bi-serial correlation.

Source

Always set to Rep. This identifies that the data was as reported by the primary study authors.

Design

The refers to the design method used by the study author. 4GroupCO is a 4-group crossover design. Mixed means different experiments in a particular family used different methods (only S3 used mixed methods and 4 experiments used the 4 group crossover and one used an independent groups design). ABBACO is the standard 2-group crossover design. IndGroups is the independent groups design also called between groups design or a randomised design. PrePost is pretest and posttest design with a post test control.

Exp1

This is the reported standardized effect size for the first experiment in the family.

Exp2

This is the reported standardized effect size for the second experiment in the family.

Exp3

This is the reported standardized effect size for the third experiment in the family.

Exp4

This is the reported standardized effect size for the fourth experiment in the family. NA means there was no fourth experiment in the family.

Exp5

This is the reported standardized effect size for the fifth experiment in the family. NA means there was no fifth experiment in the family.

Source

https://madeyski.e-informatyka.pl/reproducible-research/

Examples

KitchenhamMadeyskiBrereton.ReportedEffectSizes

reproducer documentation built on Oct. 18, 2023, 5:10 p.m.