Search

This project will evaluate 1) the sensitivity of four cognitive constructs (working memory, long-term memory, attention control, and fluid intelligence) to a motivational manipulation and 2) changes in the correlational and factor structure of cognitive abilities based on the motivational manipulation. Hypotheses: 1) The motivational incentive will improve cognitive performance across domains. This hypothesis will be tested with a task- and factor-level comparison of cognitive performance in each domain. Factor scores will be extracted from confirmatory factor analyses and compared across conditions. 2) The factor structure of cognitive abilities will not be different based on conditions. That is, the factor loadings and interfactor correlations among attention control, working memory, long-term memory, and fluid intelligence will not differ between the experimental (motivated) and control condition. Demographic data - Participants will self-report their age, gender identity, race, ethnicity, native speaking language, English proficiency, any vision issues, and any color-blindness issues - Participants with low English proficiency or outside our target age range (17 - 40) will be excluded from the analysis Procedure - Participants will complete the measures during a single 2-hour laboratory session. The sessions will have 4 to 8 participants per session. The measures will be completed on desktop computers in a group run room, each separated by cubicle-style dividers. A research assistant will be present to answer questions and provide clarification if necessary - Informed consent will be gathered from participants before any demographic or behavioral data are collected - Data will be collected at two sites: Arizona State University (ASU) and the University of Texas at Arlington (UTA) Condition assignment - Participants are assigned to conditions based on their sequential four-digit subject ID numbers. Participants with even subject ID numbers will be assigned to the control condition, and odd subject numbers will be assigned to the experimental condition. To distinguish between participants who come from ASU and UTA, ASU will use subject ID numbers in the 2000s and UTA will use subject ID numbers in the 3000s. The following tasks will be used to measure attention control: Antisaccade. Each trial starts with a warning cue (*** in size 28 Courier New bold font) that appears for either 1,000 or 2,000 ms. A white flashing cue (*) then appears on either the right or left size of the screen for 350 ms. A target (O or Q in size 28 Courier New black font) then appears on the opposite side of the screen for 100 ms, followed by a backward mask (##) until the participant makes a response. Participants use the ‘O’ and ‘Q’ keys on the keyboard to identify the target. There are eight slow-paced practice trials in which the target stays on-screen for 500 ms, then 16 practice trials of the fast-paced version with a 100-ms target duration, then 75 experimental trials, split into 5 blocks. In the experimental condition, feedback will be provided after every trial and an accuracy percentage will be provided after every 15-trial block. Sustained Attention to Response Task (SART). Each trial starts presents a single digit (1 to 9) for 300 ms, followed by blank 900-ms interstimulus interval. The participant's task isto press the spacebar whenever they see any digit other than 3. When they seethe digit 3, they are to withhold their response. There are 30 practice trials followed by 5 blocks of 90 experimental trials. The dependent measure is the standard deviation of reaction times (RT SD) on ‘go’ trials. Psychomotor vigilance task (PVT). Each trial starts with a row of zeros appeared at the center of the screen (00.000). After an unpredictable amount of time ranging from 2 to 8 seconds, the zeros begin counting forward like a stopwatch. Participants are instructed to press the spacebar as quickly as possible to stop the numbers from scrolling. In the experimental condition, upon response, the timer stops, revealing their reaction time (e.g., 00.324) for 1 second. In the control condition, the timer switches to XX.XXX. After a 1-second blank intertrial interval, the next trial begins. In both conditions, participants completed 5 practice trials followed by 5 blocks of 15 experimental trials. The dependent variable is each participant’s slowest 20% of reaction times. Degraded-mask vigilance (DMV). Each trial presents a grey digit/letter (0, D, or backward D) against a patterned masking background for 40 ms, followed by just the masking background for 1,000 ms. The participant's task is to press the spacebar whenever they see the 0, but not when they see the D or backward D. In the experimental condition, feedback is given after each of 5 blocks lasting 120 trials. The dependent variable is discrimination (hit rate - false alarm rate). The following tasks will be used to measure working memory capacity: Color change-detection. Each trial starts with a black fixation cross against a gray background for 1,000 ms. Then, the target items appear for 250 ms. The items are 4, 6, or 8 colored squares, sampled without replacement from a set of 9 possible colors. After a 1,000-ms blank delay screen, the items reappear with one square circled by a black ring. The participants’ task is to decide whether this square isthe same color or a different color than its initial presentation. Participants made their responses by pressing keys marked ‘S’ for ‘same’ and ‘D’ for ‘different’ (the ‘F’ and ‘J’ keys on the keyboard). After a 1,000-ms blank delay screen, the next trial began. There is one practice block of 8 trials followed by 5 blocks of 12 trials each. The dependent variable is k (set size*[hit rate - false alarm rate]), averaged across the three set sizes. Orientation change-detection. Each trial starts with a black fixation cross against a gray background for 1,000 ms. Then, the target items appear for 250 ms. The items are 4, 6, or 8 oriented bars, sampled with replacement from a set of 4 possible orientations (horizontal, vertical, slanted 45 degrees right or slanted 45 degrees left). After a 1,000-ms blank delay screen, the items reappear with one bar marked by a white dot. The participants’ task is to decide whether this bar is the same orientation or a different orientation than its initial presentation. Participants make their responses by pressing keys marked ‘S’ for ‘same’ and ‘D’ for ‘different’ (the ‘F’ and ‘J’ keys on the keyboard). After a 1,000-ms blank delay screen, the next trial begins. There is one practice block of 8 trials followed by 5 blocks of 12 trials each. The dependent variable is k (set size*[hit rate - false alarm rate]), averaged across the three set sizes. Letter change-detection. Each trial starts with a black fixation cross against a gray background for 1,000 ms. Then, the target items appear for 250 ms. The items are 4, 6, or 8 oriented bars, sampled without replacement from the set of English consonants. After a 1,000-ms blank delay screen, the items reappear with one letter marked by a black square. The participants’ task is to decide whether this letter is the same letter or a different letter than its initial presentation. Participants make their responses by pressing keys marked ‘S’ for ‘same’ and ‘D’ for ‘different’ (the ‘F’ and ‘J’ keys on the keyboard). After a 1,000-ms blank delay screen, the next trial begins. There is one practice block of 8 trials followed by 5 blocks of 12 trials each. The dependent variable is k (set size*[hit rate - false alarm rate]), averaged across the three set sizes. Long-term memory will be measured with the following tasks: Immediate free recall. Participants are presented with lists of 10 words, one word at a time for 2 seconds, followed by a 500-ms interword interval. Then, at the end of the list, participants are given 30 seconds to recall as many words from the list as possible. Recall is completed by typing the words into a text box on the screen, and pressing 'enter' after each word to submit it as a response. There are six lists total. The dependent variable is proportion of correctly recalled words. Picture source-recognition. Participants are presented with a sequence of 30 images. Each image appears in one of the four quadrants of the screen (top-left, top-right, bottom-left, bottom-right). Each image appears for 3 s. Then, during the test phase, participants are presented with 60 images, one at a time at the center of the screen. Participants must decided whether an image is 'old' (i.e., on the study list) or 'new.' If the image is 'old' participants press a key corresponding to the quadrant in which it was studied (1 = top-left, 2 = top-right, 3 = bottom-left, 4 = bottom-right). If the image is 'new', they press the 5 key. The dependent variable is the proportion of correct responses. Cued recall. Participants are presented with lists of cue-target word pairs (e.g., moon - hole). The cue-target pairs are presented one pair at a time for 3 seconds, followed by a 500-ms interpair interval. At the end of the list, participants are presented with the cue words, one at a time, in a different random order (e.g, moon - ???), and the participants' task is to recall the target word that was paired with that cue. Participants have 5 seconds to type their response. The dependent variable is proportion of correctly-recalled target words. Fluid intelligence will be measured with the following tasks: Raven Advanced Progressive Matrices. On each trial a 3 x 3 grid appears with patterned shapes in each cell. The bottom-right piece of the grid is missing, and the participants’ task is to select the piece that best completes the pattern in the grid from a set of eight options. Participants had 10 minutes to complete the 18 odd-numbered items. The dependent variable is the number of correctly-reported items. Number series. On each trial, a sequence of numbers appear, and the participant’s task is to select from a set of five possible options the number that best continues the sequence. Participants have four and a half minutes to complete as many trials as possible, with a maximum possible score of 15. Letter sets. On each trial, a set of four different four-letter sets appears. Among the sets, three of the four sets follow an implicit rule, and one of the four sets violates this rule. The participant’s task is to select the set of letters that violated the rule. Participants have five minutes to complete as many trials as possible, with a maximum possible score of 20. The motivation manipulation will be delivered as follows: - In all tasks, participants in the experimental condition will receive both trial-by-trial and block by block feedback about their performance. Specifically, on the color change-detection, orientation change-detection, letter change-detection, antisaccade, immediate free recall, picture source recognition, cued recall, Raven, number series, and letter sets tasks, each response will be followed by a brief (500-ms) screen that says "correct" in blue font or "incorrect" in red font. Then, at the end of each block (or list), participants will see how many trials they got correct (e.g., "That is the end of the block. You correctly responded on 12 out of 15 trials. The next block will begin in __ seconds" with a 10 second countdown). Then, at the end of each task, they will see their overall performance, and given a percentile ranking based on historic performance on these tasks. The benchmarks for performance have been selected from prior work (Robison & Brewer, 2020; 2022; Robison et al., under review; Unsworth et al., 2011). For example, a participant will see a screen that says, "That is the end of the task! Overall, you were accurate on 78% of trials. That means you performed better than 82% of people on this task." Below that statement, there will be a black bar with 5 equally-spaced tick marks with labels for 'below average', 'average', and 'above average' spaced across the bottom. A triangular red marker will point to where in the distribution their performance fell. - In the control conditions, participants will receive neither trial-by-trial, block-by-block, nor percentile feedback about their performance. Each 500-ms feedback screen will simply be a blank intertrial interval; each blockwise feedback screen will be a screen that says "That is the end of the block. The next block will begin in ___ seconds." with a countdown from 10. At the end of the task, the screen will say "That is the end of the task. Please get the experimenter." Between the practice trials and experimental trials of every task, participants will rate their motivation on a 9-point scale (1 = not at all motivated, 9 = extremely motivated). The statement will say, "Please rate how motivated you feel to perform well on the task." Participants will use the keyboard to mark their response. The data will be screened for outliers and inauthentic attempts at the tasks with the following criteria: - In the color change-detection, orientation change-detection, and letter change-detection tasks, participants with k values less than 0.5 will be excluded from the analysis. - In the antisaccade task, participants who have an average accuracy of 40% or lower will be excluded - In the SART, participants with a hit rate less than 75% or an RT SD > 300 ms will be excluded from the analysis. - In the DGV, participants with a sensitivity of less than 0 will be excluded - In the PVT, trials with RTs faster than 200 ms or slower than 3,000 ms will be excluded. Then, participants with fewer than 60 valid trials will be excluded. - In the Raven, number series, and letter sets tasks, participants with median response times of less than 3 seconds will be excluded. - In the immediate free recall, cued recall, and picture source recognition task, participants with lower than 10% accuracy will be excluded from the analysis. - After the above exclusions, the data will be standardized and any values outside +/- 3 standard deviations of the mean will be removed - Participants must have valid data for at least 6 of the 12 tasks to be included in the final analysis Between-group comparisons - The dependent variables from each task will be submitted to independent-samples t-tests. Because there are 12 dependent variables and therefore 12 comparisons, we will use an adjusted p-value of 0.004 (0.05/12) for the t-tests. We will also report effect size (Cohen's d) and the 95% confidence interval around the effect size. Factor analysis - Confirmatory factor analyses with factors for attention control, working memory, long-term memory, and fluid intelligence will be specified with the lavaan package in R - Antisaccade, PVT, SART, and DMV will be set to load onto the attention control factor; color change-detection, orientation change-detection, and letter change-detection will be set to load on the working memory factor; free recall, cued recall, and picture source recognition will be set to load on the long-term memory factor; and Raven, number series, and letter sets will be set to load on the fluid intelligence factor. The factors will all be allowed to correlate. - Missing values will be handled with the missing = 'ml' argument of the CFA function - We will report several fit statistics for the models including the chi-squared test, the comparative fit index (CFI), the Tucker-Lewis index (TLI), the standardized root mean residual (SRMR), and the root mean squared error of approximation (RMSEA). We will consider CFI and TLIs > 0.90 acceptable and SRMR and RMSEAs of < 0.08 acceptable. - The standardized factor loadings and interfactor correlations will be reported. Measurement invariance - We will perform several tests of measurement (in)variance across the two conditions - First, we will specify the same CFA on each sample (experimental and control conditions) and allow all factor loadings and correlations to be estimated freely for each sample - Then, we will specify a model in which the factor loadings will be set to be equal - Then, we will specify a model in which the interfactor correlations are set to be equal - Finally, we will specify a model in which the factor means (intercepts) are set to be equal - These models will be sequentially compared using the chi-square test, allowing us to examine whether the experimental condition systematically affects the underlying factor structure of the data Sample size determination - We estimated the sample size necessary to detect an effect size of 0.30 with 80% power and an alpha 0.004 (one-tailed) using G*Power 3.1. This analysis yielded a required sample size of 274 participants per condition. - We then referred to a simulation study by Kretzschmar and Gignac (2019), which estimated the sample size at which latent variable correlations stabilize. Interfactor correlations between attention control, working memory, long term memory, and fluid intelligence are usually quite high (0.50 - 0.80). Based on the simulations, we would be able to estimate the factor correlations with a reasonable about of precision (corridor of stability = 0.10) and level of confidence (0.80). - Given these considerations, we set our minimum target sample size at 274 participants per condition. - We will collect data until we reach the end of an academic semester or we reach 350 participants per condition, whichever happens first.