Data Files and Variables

Data Files

We release the following data files:

• questionnaire.json: Anonymized participant demographics and language history questionnaire.

• session_summary.csv: Summary statistics and information on each participant’s experiment session, such as reading comprehension accuracy and experiment duration.

• [preview]_b[batch]_l[list]_[subj_id].edf: The raw eye-tracking data for the entire eye-tracking session of each participant, in EDF format.

• [preview]_b[batch]_l[list]_[subj_id].asc: Gaze location and additional features at 1ms intervals for the entire eye-tracking session of each participant, in ASCII format.

• Fixation Reports: Eye movement features, experiment and trial information, and linguistic word properties aggregated at the level of individual fixations, in tab-separated CSV format. A separate report is available for each of the Interest Periods.

• Interest Area Reports: Eye movement features, experiment and trial information, and linguistic word properties aggregated at the word level, in tab-separated CSV format. A separate report is available for each of the Interest Periods.

Fixation and Interest Area Report Files

Interest Period	Page	Content	Fixation Report	Interest Area Report
Title	Title	Article title	fixations_Title.csv	ia_Title.csv
Question Preview	1	Question	fixations_Question_preview.csv	ia_Question_preview.csv
Paragraph	2	Paragraph	fixations_Paragraph.csv	ia_Paragraph.csv
Question	3	Question	fixations_Question.csv	ia_Question.csv
Answers	4	Question and answers	fixations_Answers.csv	ia_Answers.csv
QA	3+4	Question and answers	fixations_QA.csv	ia_QA.csv
Feedback	5	Correct/Incorrect	fixations_Feedback.csv	ia_Feedback.csv

---

Data Variables

Participant Questionnaire Variables

Participant questionnaire variables. *Reading habits questions are based on Section 1 of the reading habits self-report of Acheson et al (2008).

Variable	Description	Values
Participant ID	Participant’s ID	Numeric
Age	Participant’s age	Years
Gender	Participant’s gender	Male / Female / Other
Home Country	Participant’s home country	List of countries
Education Level	Highest/current level of education	Secondary, Undergraduate, Postgraduate
Native English Speaker	Native English speaker	Yes / No
English AoA	English Age of Acquisition	Since birth, or numeric age
Reading Habits	Weekly time spent reading in various categories	0 to 7+ hours
Dyslexia	Presence of dyslexia	No / Dyslexia
Language Impairments	Language impairments	No / Impairment (free text)
Eye Conditions	Eye conditions	No / Specific condition (Amblyopia, etc.)

Session Summary Variables

Session summary file variables.

Variable	Description	Values
participant_id	Participant’s ID	38-547
article_batch	A 10-article batch assigned to the participant	1 (articles 1-10) / 2 (articles 11-20) / 3 (articles 21-30)
list_number	Experimental list	1 - 60
question_preview	Was the question presented before the paragraph (i.e., the reading regime)	True / False
data_collection_site	Location of data collection	MIT / Technion
comprehension_score-regular_trials	Participant’s overall reading comprehension score during first reading (10 articles, 54 regular trials)	0-100%
comprehension_score-repeated_reading	Participant’s overall reading comprehension score during repeated reading (2 articles, 8-14 repeated trials)	0-100%
recalibration_count	Number of times the session was interrupted to recalibrate the eye tracker	0 or more
total_recalibrations	Number of times the eye tracker was recalibrated during the session (in addition to the 3 obligatory calibrations)	0 or more
mean_validation_error	Mean validation error across all calibrations immediately preceding text reading	visual degrees
total_session_duration	Total duration of the experimental session (including breaks and calibrations)	minutes
session_duration	Duration of the experimental session excluding breaks and calibrations	minutes
dominant_eye	Participant’s dominant eye	L / R
tracked_eye	Eye that was tracked (typically the dominant eye)	L / R / LR **
lextale_score	Participant’s score on the LexTALE vocabulary test *	0 - 100

Notes:

• *LexTale scores are available for 100 participants.

• **L: left eye, R: right eye, LR: data was collected from both eyes (switched between eyes during the experiment).

Experiment, Trial, and Linguistic Annotation Variables

Experiment and trial variables, and linguistic annotations in the Fixation and Interest Area reports. UD annotations are extracted using spaCy. See the SR Data Viewer user manual for documentation of eye movement variables in these reports. Note, missing values are denoted by “.”.

Category	Feature	Description	Values
Experiment Variables	participant_id	Participant’s ID	38-547
	list_number	Experimental list	1 - 60
	question_preview	Was the question presented before the paragraph (i.e., the reading regime)	True / False
	article_batch	A 10-article batch assigned to the participant	1 (articles 1-10) / 2 (articles 11-20) / 3 (articles 21-30)
Trial Variables	trial_index	The trial index	1 - last trial
	practice_trial	Whether the trial was a practice trial	True / False
	article_id	The unique identifier for an article in a batch	1-10
	paragraph_id	The unique identifier for a paragraph in an article	1-7
	difficulty_level	Paragraph difficulty level	Adv / Ele
	repeated_reading_trial	Whether the trial was a repeated reading trial	True / False
	article_index	The index of the article in the session. Article 0 is the practice article	0-12
	article_title	The article title, presented before the first paragraph of each article	String
	paragraph	The paragraph presented in the trial	String
	question	The question presented in the trial	String
	same_critical_span	Whether there was another question with the same critical span	0 if no other question. 1 or 2 otherwise (arbitrarily per question set).
	selected_answer	The answer selected by the participant	A/B/C/D
	selected_answer_position	The position on the page of the answer selected by the participant	0/1/2/3 corresponding to answer positions: top, left, right, bottom
	correct_answer_position	The position on the page of the correct answer for the trial	0/1/2/3 corresponding to answer positions: top, left, right, bottom
	answers_order	Mapping between position on page and A/B/C/D	list of ABCD corresponding to answer positions: top, left, right, bottom
	answer_1	The answer presented in the trial in the top position	String
	answer_2	The answer presented in the trial in the left position	String
	answer_3	The answer presented in the trial in the right position	String
	answer_4	The answer presented in the trial in the bottom position	String
Linguistic Annotations - Big Three	word_length	Number of characters in the word	Integer
	word_length_no_punctuation	Number of characters in the word excluding punctuation	Integer
	subtlex_frequency	Log word frequency from the SUBTLEX-US database	Bits
	wordfreq_frequency	Log word frequency from the Wordfreq database	Bits
	gpt2_surprisal	Word surprisal extracted from the GPT-2 language model	Bits
Linguistic Annotations - Universal Dependencies (UD)	universal_pos	Universal part-of-speech tag	17 possible tags
	ptb_pos	Penn Treebank part-of-speech tag	See Label Scheme - TAGGER here
	head_word_index	Index of the syntactic head word in the dependency tree	Integer, 0 for root - number of words in the sentence
	dependency_relation	Dependency relation label to the head word in the dependency tree	See Label Scheme - PARSER here
	left_dependents_count	Number of syntactic dependents to the left	Integer
	right_dependents_count	Number of syntactic dependents to the right	Integer
	distance_to_head	Distance in words to the syntactic head	Integer, starting at 1 for adjacent words
	morphological_features	List of morphological features of the word	See list here
	entity_type	The entity type of the word (if applicable)	See Label Scheme NER here. None if not an entity.
STARC Auxiliary Spans	auxiliary_span_type	Whether a word is part of the critical span or the distractor span	critical / distractor / outside
	critical_span_indices	Start and end word indices of the critical span	list of tuples of integers
	distractor_span_indices	Start and end word indices of the distractor span	list of tuples of integers