KitchenhamEtAl.CorrelationsAmongParticipants.Torchiano17JVLC {reproducer}

KitchenhamEtAl.CorrelationsAmongParticipants.Torchiano17JVLC data

Description

Data illustrate correlations between results from individual participants in a family of three cross-over experiments conducted by Torchiano et al: [1] M. Torchiano, G. Scanniello, F. Ricca, G. Reggio, and M. Leotta, 'Do UML object diagrams affect design comprehensibility? Results from a family of four controlled experiments.' Journal of Visual Languages and Computing, vol. 41, pp. 10–21, 2017. Although the paper reports four experiment, we only have data from three of those experiments. The experiments assess whether the comprehensibility of UML specifications improve when the software documents include UML object diagrams as well as the standard UML class diagrams. If you use this data set please cite: [1] M. Torchiano, G. Scanniello, F. Ricca, G. Reggio, and M. Leotta, 'Do UML object diagrams affect design comprehensibility? Results from a family of four controlled experiments.' Journal of Visual Languages and Computing, vol. 41, pp. 10–21, 2017. [2] Barbara Kitchenham, Lech Madeyski, Giuseppe Scanniello and Carmine Gravino, 'The importance of the Correlation between Results from Individual Participants in Crossover Experiments' (to be submitted as of 2020).

Usage

KitchenhamEtAl.CorrelationsAmongParticipants.Torchiano17JVLC

Format

A data frame with 214 rows and 8 variables:

ExperimentID

<fct>|ExperimentID: A unique identifier for each of the three experiments in the data set.

ParticipantID

<fct>|Participant ID: An identifier for each participant, unique for a specific experiment.

SequenceGroup

<fct>|Experimental Sequence Group: A , B , C, D

System

<fct>|Software systems used in the experiment: File System manager (FS) for folders, files, links. Roads system (R) handles maps made up of cities connected by means of roads. Train (T) a system to manage timetables, trains, and paths. Catalogue system (C). It collects category of items (e.g., cars) and items (e.g., car models) based on a set of features (e.g., number of doors). In PoliTo2, only FS and T were administered to the participants, while in UniBas1 and UniGe1 all the four experimental objects were used.

Period

<fct>|Time period of the cross-over experiment: 1 or 2

Treatment

<fct>|Experimental Treatment: Object Diagram (OD) vs No Object Diagram (NoOD)

Comprehension

<dbl>|Dependent variable: The comprehension level the software engineers. For PoliTo2 Comprehension was based on answering a set of 4 questions, for UniBas and UniGe comprehension was measured using the F metric.

CrossOverID

<fct>|CrossOver category: For 4 group crossover designs, the crossover category specifies the matching pairs of sequence groups, CO1 and CO2. For 2 group crossover, the category is set to CO1 only

Examples

KitchenhamEtAl.CorrelationsAmongParticipants.Torchiano17JVLC