Methods to Test Visual Attention Online

Podsumowanie

To replicate laboratory settings, online data collection methods for visual tasks require tight control over stimulus presentation. We outline methods for the use of a web application to collect performance data on two tests of visual attention.

Streszczenie

Online data collection methods have particular appeal to behavioral scientists because they offer the promise of much larger and much more representative data samples than can typically be collected on college campuses. However, before such methods can be widely adopted, a number of technological challenges must be overcome – in particular in experiments where tight control over stimulus properties is necessary. Here we present methods for collecting performance data on two tests of visual attention. Both tests require control over the visual angle of the stimuli (which in turn requires knowledge of the viewing distance, monitor size, screen resolution, etc.) and the timing of the stimuli (as the tests involve either briefly flashed stimuli or stimuli that move at specific rates). Data collected on these tests from over 1,700 online participants were consistent with data collected in laboratory-based versions of the exact same tests. These results suggest that with proper care, timing/stimulus size dependent tasks can be deployed in web-based settings.

Wprowadzenie

Over the past five years there has been a surge of interest in the use of online behavioral data collection methods. While the vast majority of publications in the domain of psychology have utilized potentially non-representative subject populations¹ (i.e., primarily college undergraduates) and often reasonably small sample sizes as well (i.e., typically in the range of tens of subjects), online methods offer the promise of far more diverse and larger samples. For instance, Amazon’s Mechanical Turk service has been the subject of a number of recent studies, both describing the characteristics of the “worker” population and the use of this population in behavioral research^2-6.

However, one significant concern related to such methods is the relative lack of control over critical stimulus variables. For example, in most visual psychophysics tasks, stimuli are described in terms of visual angle. The calculation of visual angles requires precise measurements of viewing distance, screen size, and screen resolution. While these parameters are trivial to measure and control in a lab setting (where there is a known monitor and participants view stimuli while in a chin rest placed a known distance from the monitor), the same is not true of online data collection. In an online environment, not only will participants inevitably use a wide variety of monitors of different sizes with different software settings, they also may not have easy access to rulers/tape measures that would allow them to determine their monitor size or have the knowledge necessary to determine their software and hardware settings (e.g., refresh rate, resolution).

Here we describe a set of methods to collect data on two well-known tests of visual attention – the Useful Field of View (UFOV) paradigm⁷ and the multiple object tracking (MOT) task⁸ – while avoiding as much as possible the sources of variability that are inherent in online measurements. These tasks can be run by any participant with an internet connection and an HTML5 compatible browser. Participants who do not know their screen size are walked through a measurement process utilizing commonly available items of standard size (i.e., credit card/CD – see Figure 1).

Data on these two tasks were collected from over 1,700 participants in a Massive Online Open Course. Average performance of this online sample was highly consistent with results obtained in tightly controlled laboratory-based measures of the exact same tasks^9,10. Our results are thus consistent with the growing body of literature demonstrating the efficacy of online data collection methods, even in tasks that require specific control over viewing conditions.

Protokół

The protocol was approved by the institutional review board at the University of Wisconsin-Madison. The following steps have been written as a guide for programmers to replicate the automated process of the web application described.

1. Login Participant

1. Instruct the participant to use an internet-enabled computer and navigate to the web application using an HTML5 compatible browser: http://brainandlearning.org/jove. Have the participant sit in a quiet room free of distractions, with the computer at a comfortable height.
   NOTE: Since the entire experiment is hosted online, the tasks can also be performed remotely without the presence of a research assistant. All instructions for the participant are included in the web application.
2. Have the participant input a unique ID that will be associated with the data collected and stored in a MySQL database. Have the participant reuse this ID if the online tasks are not completed within the same session. Before logging in, obtain consent from the participant via a consent form linked on the page.
   NOTE: A participant’s progress is saved after each task in order to allow for completion of the 2 tasks at separate times if needed. Instruct the participant to always use the same ID in order to start where one left off.

2. Screen Calibration

NOTE: The web application guides the participant through the three steps outlined in the calibration page at: http://brainandlearning.org/jove/Calibration.

1. Ask the participant to input the diagonal size of the screen in inches in the labeled textbox.
   1. However, if the participant does not know this information, have the participant find a CD or credit card as a calibration object (Figure 1). When one is selected, prompt the participant to place the object against the screen and align it with a representative image of the object displayed on the screen.
   2. Prompt the participant to adjust the size of the screen image to match the size of the physical object. Based on the measurements of a physical CD (diameter of 4.7”) or a credit card (width of 3.2”) in addition to the pixel size of the representative image, determine the ratio of pixels to inches for the screen.
   3. Retrieve the pixel resolution of the monitor via JavaScript’s screen.width and screen.height properties to then calculate the diagonal size of the screen in pixels. Knowing this value and the previously estimated pixel-to-inch ratio (see step 2.1.2), convert the diagonal size to inches. Have the participant confirm this value through a dialog box.
2. Prompt the participant to adjust the screen brightness settings until all 12 bands in a black-to-white gradient displayed on the screen are clearly distinguishable. Brightness setting controls vary by computer.
3. Ask the participant to sit an arm’s length away from the monitor in a comfortable position and then set the browser window to full screen mode. The browser window must be in full screen mode to maximize the visual space used by the tasks and to remove any visual distractions, such as the browser toolbar and desktop taskbars.
4. Knowing the resolution of the participant’s screen and the diagonal size of the monitor, use the web application to automatically calculate the pixels/degree conversion value, based on a 50 cm viewing distance. Resize the dimensions of the stimuli in the tasks using this value. All visual angle dimensions reported below are based on this assumed mean distance value from the monitor.
5. Once calibration is complete, ask the participant to complete the two tasks described below. Choose the order of the tasks or randomly assign the order via the web application.

3. Multiple Object Tracking Task (MOT) – Figure 2

1. Introduce and familiarize the participant with the MOT stimuli through a self-guided tutorial, seen at: http://brainandlearning.org/jove/MOT/practice.php. Ask the participant to read step-by-step instructions that demonstrate how the trials will work. Once the participant finishes reading the instructions, prompt the participant to go through the practice trials.
   1. Setup the practice stimuli to consist of 8 dots at 0.8° with a movement speed of 2°/sec. Use the HTML5 requestAnimationFrame API to optimize browser animation at a frame rate of 60 Hz in order to control this stimulus motion.
   2. Ensure the dots move within the boundaries of a circle of 2° eccentricity and a circle no larger than the height of the participant’s screen, without the instructions obscured.
   3. Set the dots to move in a random trajectory, where at each frame a dot has a 60% chance of changing direction by a maximum angle of 0.2°. If a dot collides with another dot or the inner or outer radial limits, move the dot in the opposite direction.
   4. Prompt the participant to track the blue dots (varying between 1 and 2 dots per practice trial), with the yellow dots acting as distractors.
   5. After 2 sec, change the blue dots to yellow dots and continue to move them amongst the original yellow dots for another 4 sec. At the end of each trial, stop the dots and highlight one.
   6. Prompt the participant to respond via key press whether the highlighted dot was a tracked dot or a distractor dot. Next, prompt the participant to press the space bar to continue onto the next trial.
   7. After 3 consecutive correct trials, or a maximum of 6 trials, move the participant onto the full task.
2. Start the full MOT task for the participant. An example of the task can be found at: http://brainandlearning.org/jove/MOT.
   1. Setup the full task with 16 dots that move at 5°/sec within the space between 2° eccentricity and 10° eccentricity. If the participant’s screen cannot fit a circle of 10° eccentricity, use the maximum size the screen can contain instead.
   2. Have the participant complete a total of 45 trials: a mixture of 5 trials consisting of 1 tracked dot and 10 trials each consisting of 2 - 5 tracked dots. Match all other parameters to the practice trials (see steps 3.1.3 - 3.1.6).
   3. Record the participant’s response and response time once the dot is highlighted.
   4. For every 15 trials, suggest a break to the participant. At these breaks, display the participant’s performance (percent of correct trials) within the block on the screen.

4. Moving from One Task to Another (Optional Step)

1. Allow the participant to take a break between the two tasks. However, repeat steps 1 and 2 if the tasks are not completed during the same login session.

5. Useful Field of View Task (UFOV) – Figure 3

1. Introduce and familiarize the participant with the UFOV stimuli through a self-guided tutorial, seen at: http://brainandlearning.org/jove/UFOV/practice.php. Ask the participant to go through 4 stages of step-by-step instructions that demonstrate the two target stimuli that must be attended to during the task.
   1. Set the central target stimulus as a 1° smiley that flashes at the center of the screen with either long or short hair. Randomize the smiley’s hair length across trials.
   2. Set the peripheral target stimulus as a 1° star that flashes at 4° eccentricity at one of 8 locations around the circle (0°, 45°, 90°, 135°, 180°, 225°, 270°, and 315°). Randomize the location of the star across trials.
   3. Control stimulus duration via number of frames used for presentation time. Optimize frame refresh at about 17 msec per frame by using the HTML5 requestAnimationFrame API.
   4. To check if the expected presentation time was achieved, use JavaScript’s getTime() method to obtain the stimulus duration’s start time and end time based on the participant’s system clock. Calculate the measured presentation time from these two values and use this value for data analysis.
   5. For each practice trial, wait 500 msec before displaying the stimuli for approximately 200 msec (about 12 frames).
   6. Follow stimulus presentation with a noise mask composed of a randomly generated grayscale dot array for 320 msec (about 19 frames).
   7. For stage 1, only display the central target and then prompt the participant to respond via key press which hair length was displayed.
   8. For stage 2, only display the peripheral target and then prompt the participant to click on one of 8 radial lines, representing the 8 possible target locations, to indicate where the star appeared.
   9. For stage 3, display both central and peripheral target stimuli and then prompt the participant to provide responses to both the type of smiley and the location of the star.
      NOTE: Participants can freely choose the order of these two responses.
   10. For stage 4, display both target stimuli in addition to peripheral distractors, and then prompt the participant to respond to both target stimuli. For the distractors, display 1° squares presented at the remaining 7 locations at 4° eccentricity, in addition to 8 more squares at 2° eccentricity.
   11. After the participant’s response, show the participant feedback (a green checkmark for a correct answer or a red cross for an incorrect answer) for each target response after each trial.
   12. Move the participant onto the next practice stage after getting 3 consecutive correct trials. After stage 4, move the participant onto the full task.
2. Prompt the participant to start the full UFOV task. An example of the task can be found at: http://brainandlearning.org/jove/UFOV.
   1. Present the same central stimulus as in the practice session (see step 5.1.1). Display the peripheral target at 7° eccentricity at one of the previously mentioned 8 locations (see step 5.1.2). 24 distractor squares are also displayed at 3° eccentricity, 5° eccentricity, and the remaining 7° eccentricity locations.
   2. Use a 3-down, 1-up staircase procedure to determine the presentation time of the stimuli: decrease the duration of stimuli after 3 consecutive correct trials and increase after each error trial.
   3. Before the first 3 reversals in the staircase, use a step size of 2 frames (approximately every 33 msec). After 3 reversals, use a step size of 1 frame. Vary the delay before the stimulus onset between 1 frame and 99 frames per trial, and keep the noise mask duration at 320 msec (about 19 frames).
      NOTE: Reversals are the points at which the duration changes either from increasing to decreasing, or decreasing to increasing.
   4. End the task when one of three conditions is met: the staircase procedure reaches 8 reversals; the participant completes 10 consecutive trials at either the ceiling duration (99 frames) or the floor duration (1 frame); or the participant reaches a maximum of 72 trials.
   5. Record the participant’s response and response time for both the central stimulus and the peripheral stimulus.

Wyniki

Outlier Removal

A total of 1,779 participants completed the UFOV task. Of those, 32 participants had UFOV thresholds that were greater than 3 standard deviations from the mean, suggesting that they were unable to perform the task as instructed. As such, the UFOV data from these participants were removed from the final analysis, leaving a total of 1,747 participants.

Data were obtained from 1,746 participants for the MOT task. Two participants had mean accuracy scores t...

Dyskusje

Online data collection has a number of advantages over standard laboratory-based data collection. These include the potential to sample far more representative populations than the typical college undergraduate pool utilized in the field, and the ability to obtain far greater sample sizes in less time than it takes to obtain sample sizes that are an order of magnitude smaller in the lab^1-6 (e.g., the data points collected from 1,700+ participants in the current paper were obtained in less than one wee...

Materiały

Name	Company	Catalog Number	Comments
Computer/tablet			It must have an internet connection and an HTML5 compatible browser
CD or credit card			May not be needed if participant already knows the monitor size