Raw eye tracking data of healthy adults reading aloud words, pseudowords and numerals

This paper describes data from de Chambrier et al. (2023). The dataset [2] contains raw eye tracking data of 36 healthy adults, collected using an EyeLink 1000 (SR Research Ltd., ON, Canada) during an on-screen reading task. Participants read 96 items including words, pseudowords and numerals. Each item was presented at the center of the screen until the participant produced an oral response and pressed the keyboard's space bar. Part of the data were analyzed to extract key metrics such as fixation number, fixation duration, saccade number, and saccade amplitude identified by the EyeLink 1000 [1]. Reuse potential includes (but is not limited to) pupil diameter data analysis, identification of fixations and saccades using custom algorithms, and secondary analyses using participant demographics (age, gender) as independent variables.

Experimental and Cognitive Psychology Specific subject area Reading Type of data Eye tracking raw data sampled at 10 0 0 Hz using an EyeLink 10 0 0 (SR Research Ltd., ON, Canada) How the data were acquired The items were presented at the center of a 24-inch LCD monitor (visible screen width and height: 520 × 325 mm) with a resolution of 1'920 × 1'200 (refresh rate: 60 Hz). The dominant eye-gaze position was recorded at a sampling frequency of 10 0 0 Hz with a desktop-mounted EyeLink 10 0 0 eye tracker, placed at 530 mm distance in front of the participant. Data format Raw Description of data collection Each trial included the sequential presentation of a target item and a blank screen.
The target item was presented on the screen until the participant produced an oral response and pressed the space bar. Immediately following the response, a blank screen was presented for 100 ms, followed by the drift check indicating the beginning of the next trial. Data source location

Value of the Data
• Reproducibility: Other researchers could reproduce the results and verify our scientific claims. • These data can be useful for researchers interested in reading (e.g., developmental psychologists, neuropsychologists), as well as for people interested in event (fixations, saccades, microsaccades, blinks, and so forth) detection from raw eye tracking samples (e.g., computer scientists, neuroscientists, and so forth). • Possible data use/reuse cases include (but are not limited to) pupil diameter data analysis, identification of fixations and saccades using custom algorithms, as well as secondary analyses (e.g., using participant demographics as independent variables). Pseudowords were precisely matched to words, and eye movements that occurred on these two categories of items could also be compared.

Objective
Reading numbers is a skill required in daily life and is fundamentally different com pared to reading words. Word reading follows a phonographic system, in which graphic units (e.g. letters, syllables) do not have a meaning on their own but by being assembled with other symbols to form words. Number reading follows a logographic (or ideographic) system, in which each symbol (e.g. 1, 5, 0) has a meaning in itself but can be further combined to express quantities larger than 9. During the last four decades, eye movements have allowed to generate extensive knowledge on the cognitive processes occurring while reading text, however we could not find research that targeted possible eye movement differences when reading words compared to numbers. We generated this dataset to spot such differences.
This data article adds to the published article (a) a detailed description of the raw data publicly available in a data repository and (b) some suggestions on how these data could be used/reused. Table 1 contains the names of variables in the dataset. For further details, see http:// sr-research.jp/support/files/dvmanual.pdf .

Participants
Participants were 36 students from the University of Lausanne psychology course (27 women and 9 men; mean age = 21.3 years old; SD = 4.15). All had normal or lenses-corrected vision, no history of learning disorders, and were native French speakers. Table 2 contains age and gender of each participant.

Procedure
Participants read 96 items in total. They were individually placed in a quiet, low-lit room, seated 930 mm from the PC screen. We presented items at the center of a 24-inch LCD monitor with visible screen width and height of 520 × 325 mm, resolution of 1 920 × 1 200 px and refresh rate of 60 Hz. Table 3 contains all the stimuli displayed. A chin and forehead rest ensured a correct head position. We determined eye dominance using the hole-in-card test with participants' hands and centered gaze. Gaze position of the dominant eye was measured at a sampling rate of 10 0 0 Hz with a table-mounted EyeLink 10 0 0 positioned 530 mm in front of the participant.
Each trial was preceded by a drift check, during which the experimenter ensured that the participant centered his or her gaze within a black circle 0.48 °in diameter and then validated  1  18  female  2  unknown  male  3  22  female  4  21  female  5  23  female  6  19  female  7  22  female  8  21  female  9  23  male  10  21  female  11  21  female  12  20  female  13  19  male  14  43  male  15  19  female  16  23  female  17  21  female  18  19  male  19  23  female  20  21  female  21  22  female  22  22  female  23  18  male  24  18  female  25  19  female  26  25  female  27  21  male  28  22  female  29  19  female  30  22  female  31  19  female  32  20  female  33  22  male  34  19  male  35  19  female  36 20 female the start of the trial. Then began the sequential presentation of the target item displayed on a gray background, until the participant generated a spoken response and pressed the space bar on the keyboard. Immediately after the response, a blank screen was displayed for 100 ms, followed by the drift check that signaled the start of the next trial. We presented target items in two blocks of trials: One block involved the presentation of 48 randomly displayed numeral trials, and the other block involved the presentation of 24 randomly displayed word trials followed by 24 randomly displayed pseudoword trials. We counterbalanced the order of the two blocks across participants. Before each category of items, we asked participants to read the items out loud as accurate and quickly as possible. For the pseudoword category, we indicated that the items had no meaningfulness.