Dataset title: PROVglish Sentences Human Evaluation 
Author: Darren Richardson, University of Southampton

Description:
These are the raw, unprocessed results of a questionnaire attempting to evaluate the effectiveness of extracting linguistic information from URIs on the quality of provenance explanation generation.

Archive contents:
    trials.csv:
        This file maps Participant IDs to questions, showing the order in which the participant was shown the sentence pairs. Each number in the cell refers to a sentence pair ID, which can be found in sentences.csv
        
    sentences.csv:
        This file shows the sentence-pairs that were used for this experiment. Each pair has both a version with the URIs exploited, and a version without the URIs exploited.
        
    responses.csv:
        This file contains the participant responses to the questions. 
        The columns "Participant ID" and "Sentence Pair ID" are foreign keys to trials.csv and sentences.csv respectively. 
        
        Columns:
            - "Participant ID" is a foreign key to trial.csv
            - "Sentence Pair ID" is a foreign key to sentences.csv
            - "Sentence A" shows which of the two sentences in the sentence-pair was shown to the participant as Sentence A.
            - "Preferred Sentence" shows whether the Participant preferred Sentence (A), Sentence (B), or (N)either sentence


Ethics approval number: ERGO/FPSE/16731