Towards a Requirement Patterns-based Assistant for Low-Code/No-Code Software Development - Raw data, textual descriptions, and experimental objects

This repository contains the protocol, raw data, literal quotes, experimental objects, and other materials related to the semi-structured interviews and quasi-experiment from the paper "Towards a Requirement Patterns-based Assistant for Low-Code/No-Code Software Development," currently under review at the Array Journal. How to use Data from semi-structured interviews Interviews-Demo-Questionnaire.xlsx It contains the raw demographic data from interview participants. You will find a code for each participant, the main area of expertise, the time in years they have experience in such main area, the cluster type based on their demographics (i.e., domain expert, business analyst, and software developer), additional areas of expertise, programming experience in years, and their experience with modelling tools. Interview-experimental-object.png It contains screenshots of the LEMON assistant version (Figma mockups) that interview participants interacted with. Interview-Data-Repository.xlsx It contains the data extracted from the semi-structured interviews for desired features, qualities, and shortcomings. For each, we reported the clustered item (e.g., Desired feature: Recommendations for inputs), the total of participants that mention the feature, quality, or shortcoming, a representative quote for each group extracted from the transcripts, the IDs of each participant, and the frequency-divided for each type of participant (i.e., domain expert, business analyst, and software developer). Notice that we do not provide the complete transcript, but instead, we provide representative quotes for the sake of privacy, anonymity, and safety of the data. In case it is necessary, we can provide more details on a request basis. Personas#.PNG It contains results from a PERSONAS workshop conducted with experts (3 groups of 4 experts each, a total of 12 experts) in digital health software development (not the same participants from the semi-structured interviews). We reported the clustered personas profiles that helped us to cluster the interview participants into domain experts, business analysts, and software developers. Data from quasi-experiment G#-YYYY-Demo-anonym.csv/xlsx It contains the anonymized demographics from both G1, G2, and G3. It includes data about the study program, hands-on experience in industry, software development experience, among other relevant data to characterize subjects from G1, G2, and G3. LEMON-data-anonym.csv It contains the data from G1, G2, and G3, including: the group, the method (LEMON:Tool; or Without LEMON: Mendix), the starting time, the ending time, the duration in minutes, the true positives (TP), the true negatives (TN), the false positives (FP), and false negatives (FN), the precision, recall, F1, and accuracy. For G1, we also reported the perception variables, PEU, PU and ITU. Experimental-object.docx it contains the case and instructions given to G1 and G2 when using LEMON (G2 used similar instructions when using Mendix). Base_solution.png As a reference, we include the base solution for Treatment 1. This figure includes an Excerpt of the baseline solution for a blood pressure visualisation software specified as a PDS Query model for treatement 1.