------------------- GENERAL INFORMATION ------------------- 1. Title of Dataset AMD Home-based training experiment - Vertical reading 2. Author Information Principal Investigator Contact Information Name: Aurelie Calabrese Institution: University of Minnesota Address: 75 East River Road, Minneapolis MN 55455, USA Email: acalabre@umn.edu Associate or Co-investigator Contact Information Name: Gordon E. Legge Institution: University of Minnesota Address: 75 East River Road, Minneapolis MN 55455, USA Email: legge@umn.edu 3. Date of data collection (single date, range, approximate date) 2010 - 2014 4. Geographic location of data collection (where was data collected?): Minnesota, USA 5. Information about funding sources that supported the collection of the data: This research was supported by NIH grant EY002934. --------------------- DATA & FILE OVERVIEW --------------------- 1. File List A. Filename: pre-post.csv Short description: contains pre- and post-training data for all subjects B. Filename: training.csv Short description: contains training data for all subjects 2. Relationship between files: N/A 3. Additional related data collected that was not included in the current data package: N/A 4. Are there multiple versions of the dataset? NO -------------------------- METHODOLOGICAL INFORMATION -------------------------- 1. Description of methods used for collection/generation of data: Each participant was trained using a home-based training protocol, designed in our laboratory. All the pre- and post-training measurements, as well as the training sessions, were conducted remotely, with the participant at home and the experimenter in the lab. Live communication was established through Skype. Remote control software and 3G wireless network connection were used to run the tests remotely. Six sessions (1 session per day; spread over the course of two weeks) were necessary to complete the experiment. Each session lasted 60 to 90 minutes (including rest breaks). The actual training took place from Day 2 to Day 5. Prior to training, the baseline performance of each participant was assessed during a 'pre-training' session (Day 1). After training, a 'post-training' session (Day 6) was run to estimate training-related improvement. 'Pre-training' and 'post-training' measurements included: 1) RSVP reading speed, 2) flashcard reading speed (to check for transfer to page reading, involving eye movements), and 3) visual span size (to see if training would also produce an increase in the size of the visual span). Participants were randomly allocated to one of the three groups for which the training differed by the text orientation. In Group 'V', participants were trained with vertical text display (text rotated 90o clockwise). In Group 'VH', participants were first trained with vertical text display; a few months later, they also completed training with horizontal text. In Group 'H', participants were trained with horizontal text display only. Whether vertical or horizontal, participants were all trained with rapid serial visual presentation (RSVP) reading. This task has been shown to yield better reading speed improvement than other types of training with letters and words. Visual stimuli were presented on a 17" Dell laptop (Dell Precision M6500). Stimuli were generated and presented using MATLAB 7.10 software (MathWorks, Inc.) with Psychophysics Toolbox extensions. All sentences and letters were displayed in Courier font with high-contrast black letters on a white screen (luminance of 190 cd/m2). Print size was adjusted for each participant to twice their MNREAD critical print size. Sentences (average 11.5 words each) presented during the RSVP and flashcard tasks were drawn from two different pools of sentences, with no sentence presented more than once to a given participant. RSVP reading - Sentences were presented on the screen one word at a time, either oriented horizontally or vertically (rotated 90o clockwise). For a given orientation, words were all displayed at the same location (aligned on the first character), requiring minimal eye movements. Participants were asked to fixate at all times using their preferred retinal locus (PRL) and read the sentences aloud. Sentences were presented during blocks of 18 trials including 6 different exposure durations. The range of word exposure durations was chosen to yield an overall percent correct located between 10 and 100%. This range was adjusted over the course of the experiment to match the participant's change in performance. Flashcard reading - Sentences were displayed as a whole on the screen, either oriented horizontally (first word starting at the left top corner) or vertically (first word starting at the right top corner). Sentences were displayed over four lines to simulate page reading, which requires saccadic eye movements. Data collection and analysis were similar to RSVP testing. Blocks of 18 sentences were presented to the participant at 6 different exposure durations. The range of sentence durations was chosen to encompass reading performance from 10 to 100% correct and was modified if needed during the experiment, according to the participant's performance change. Visual span - Participants were asked to fixate between two dots at the center of the screen while random strings of three letters (called trigrams) were presented at various eccentricities. Participants were asked to report all three letters from first to third while maintaining fixation. For horizontal measurements, we used a standard display: two dots, vertically arranged around the horizontal midline, were used as a fixation target. Trigrams of horizontal letters were randomly presented at various distances left and right of fixation along the midline. For vertical visual span, the same display was generated, but it was presented 90-degree clock-wise rotated. 2. Methods for processing the data: RSVP reading - In each block of trials, the proportion of words read correctly was calculated for each exposure duration. A cumulative Gaussian function was then fitted to the data. The exposure time yielding 80% correct word recognition was estimated from the fitted function and converted to reading speed in words per minute (wpm). During the training, we aimed to present 10 blocks of 18 sentences each during each session, but this number varied slightly depending on the participant's reading speed. Flashcard reading - For each block, a cumulative Gaussian function was fitted according to the proportion of words correctly read. From this function, we estimated the exposure time yielding 80% correct word recognition and converted this value to reading speed in words per minute (wpm). Visual span - Visual span size was calculated with a standard method (see He, Legge, & Yu, 2013), based on the recognition accuracy for nine letter positions centered at fixation. 3. Instrument- or software-specific information needed to interpret the data: N/A 4. Standards and calibration information, if appropriate: N/A 5. Environmental/experimental conditions: N/A 6. Describe any quality-assurance procedures performed on the data: N/A 7. People involved with sample collection, processing, analysis and/or submission: N/A ----------------------------------------- DATA-SPECIFIC INFORMATION FOR: [FILENAME] ----------------------------------------- ------ pre-post.csv ------ 1. Number of variables: 8 2. Number of cases/rows: 30 3. Missing data codes: NA 4. Variable List A. Name: id Description: subject identifier B. Name: Orientation Description: Text orientation in testing C. Name: testDay Description: when was the measure recorded: before (pre) or after (post) training D. Name: visual_span Description: measure of visual span (number of letters identified in single fixation) E. Name: rsvp Description: measure of rsvp reading (in words per minute) F. Name: flashcard Description: measure of flashcard reading (in words per minute) G. Name: diagnosis Description: type of macular degeneration: age-related (AMD) or juvenile (JMD) H. Name: group Description: what group of training was the subject in: Horizontal (H), Vertical (V) or both (HV) ------ training.csv ------ 1. Number of variables: 7 2. Number of cases/rows: 78 3. Missing data codes: NA 4. Variable List A. Name: id Description: subject identifier B. Name: Orientation Description: Text orientation in testing C. Name: testType Description: when was the measure recorded: during training (train) D. Name: testDay Description: what day of training was the measure recorded t0: pre-training t1: day 1 t2: day 2 t3: day 3 t4; day 4 t5: post-training E. Name: readingSpeed Description: measure of rsvp reading (in words per minute) F. Name: diagnosis Description: type of macular degeneration: age-related (AMD) or juvenile (JMD) G. Name: group Description: what group of training was the subject in: Horizontal (H), Vertical (V) or both (HV) ------ References ------ He, Y., Legge, G. E., & Yu, D. (2013) Sensory and cognitive influences on the training-related improvement of reading speed in peripheral vision. J Vis 13, 14.