Examining the Effectiveness of Cognitive Training

This study examines the extent to which working memory training may give rise to cognitive benefits and the extent to which these effects can be explained by, or augmented by, placebos.

It is predicted that:

• Significant placebo effects will be found in cognitive training, both at the immediate post-test and the second post-test after participants were debriefed about the placebo effects.
• Adaptive N-back training will produce larger gains than adaptive Knowledge training.

Participants will be assigned to either an online or in-person version of the study depending upon the guidelines for participant interactions that are in force at the time (if only one version is allowable under current guidelines participants will be assigned to the allowable version).

Pre-Test Cognitive Assessments: For each domain, two separate measures will be utilized to ensure that any observed results are not specific to a given tasks, but are more generally representative for the six cognitive domains below.

• Working Memory: The two measures will be a non-trained version of the N-back task and the O-Span task. In the N-back task, participants view a series of letters and are asked to indicate whether the current letter is the same as the letter "N" items back. The N-back levels will vary between 1- and 4-back. In the O-Span task, participants alternate between completing simple math problems (e.g., 3+(2x2)) and viewing single letters. At the end of a given sequence, the participants are asked to recall all of the letters that they had seen in the correct order. The number of letters to be recalled will range between 3 and 7.
• Cognitive Flexibility: The two measures will be a standard task-switching paradigm and countermanding task. In the task switching task, on each trial, participants view a letter and a digit. The location of the pair instructs the participants to either categorize the letter as a consonant or vowel or else categorize the digit as even or odd. Trials can be either "non-switch trials" - in which the participant is asked to perform the same task as on the previous trial - or "switch trials" - in which the participant is asked to perform the opposite task as on the previous trial. In the countermanding task, participants are presented with two types of stimuli (e.g., dogs or monkeys) on either the right or the left side of the screen and are required to tap on one of two buttons on either the same side as the stimulus (e.g., dog; congruent trials) or on the opposite side of the stimulus (e.g., monkey; incongruent trials).
• Visual Selective Attention: The two measures will be the Useful Field of View (UFOV) task and the Attentional Network Test (ANT; executive control trials only). In the UFOV, participants are briefly presented with a display consisting of 24 items (3 on each of the four radial spokes and the four obliques; evenly spaced). One of the items is a target (a sheriff's badge), while the remaining items are distractors (open boxes). The participants' task is to indicate upon which of the 8 spokes the target appeared. In the ANT, on each trial participants view a center arrow that faces either left or right. The arrow can be flanked on either side by either response compatible arrows (i.e., arrows facing in the same direction as the center arrow) or response incompatible arrows (i.e., facing the other direction). The participants' task is to indicate the direction that the center arrow is pointing.
• Spatial Cognition: The two measures will be a standard mental rotation test and paper folding test. In the mental rotation task, participants are shown two images side by side. The images are either identical, but with one rotated relative to the other, or else are mirror-reversed and rotated copies of one another. The participants' task is to indicate whether the two items are identical or are mirror-reversed copies. In the paper-folding task, the participant is shown a depiction of a piece of paper being folded various ways before a hole is punched in the paper. The participants' task is to indicate what the paper will look like when it is unfolded.
• Fluid Intelligence: The two measures of fluid intelligence will be the Raven's Advanced Progressive Matrices and a new matrix task (UCMRT) developed and implemented by PIs Jaeggi and Seitz. In each test item, the participant is presented with a grid of elements with one of the elements missing and is asked to identify the missing element that completes the grid pattern.
• Vocabulary (control): Vocabulary will be assessed using the Mill-Hill vocabulary and the Shipley Institute of Living scales. These require participants to select the appropriate synonym for a target word among several alternatives.

Pre-Test Individual Differences Surveys: At Pre-test, participants will complete a series of surveys to assess demographic variables (e.g., age, gender, race/ethnicity, SES) as well as other individual difference factors that may be predictive of placebo-responsiveness. These include: Big 5 Personality Traits, Work and Family Orientation Scale (WOFO), Grit, Behavioral Avoidance/Inhibition scales (BIS/BAS), theories of intelligence, Schutte Self-Report Emotional Intelligence Test, and Meta-Cognitive Skills Scale.

Training Conditions: Participants will be assigned to either an experimental training condition (N-back training, which previous research has shown is associated with enhancements in general cognitive function) or a control training condition (explicit knowledge training, which previous research has shown is not associated with enhancements in general cognitive function). The cognitive training task will be an adaptive version of a visual N-back task used previously. Participants will be presented with a series of objects and will be asked to indicate whether the current stimulus matches the one presented n positions back in the sequence. An active knowledge training task (control) will be used in which participants will be asked to answer Graduate Record Examination-type (GRE) general knowledge, vocabulary, and trivia questions selected from a pool of approximately 5,000 questions. For every question, participants are required to pick one of four answer alternatives presented below the question and after their response are given the correct answer.

Explicit Expectation Induction: Participants will be assigned to either positive or negative expectations induction, which will be delivered via a slideshow. All instructions and descriptions of the to-be-trained task in the Positive Expectations groups will be designed to maximize participant belief in the effectiveness of the given training in inducing positive change in cognitive abilities or minimize participant belief in the Negative Expectations groups.

Associative Learning Mid-Tests: For the associative learning mid-tests, the same basic cognitive tasks employed at Pre-test will be altered to either provide evidence to the participant of "improved cognitive performance" (positive associative learning) or "diminished cognitive performance" (negative associative learning).

• Working Memory Mid-Test Associative Learning Manipulations:

O-Span: In the O-Span Associative Learning Mid-Test, task difficulty will be manipulated by either lessening the difficulty of the interleaved math problems (e.g., by using problems that only contain small numbers) or increasing the difficulty of the interleaved math problems.

• Cognitive Flexibility Mid-Test Associative Learning Manipulation:

Task-switching: In the task-switching test, several factors make the test more difficult. First is the number of task switches - fewer switch trials results in overall better performance. Second are the digits themselves. For instance, in the high/low task, digits closer to 5 are responded to more slowly than digits further from 5. Thus, in the Task-Switching Associative Learning Mid-Tests, the switch rate will be reduced (easier) or increased (harder) relative to the true assessments, and the distribution of digits that are displayed will include a greater or lesser percentage of easier items.

Countermanding: In the Countermanding test, difficulty is most strongly associated with the presence of incongruent trials as well as switch trials. Thus, the Countermanding Mid-Test will have decreased or increased proportions of incongruent and switch trials from 50% to 30%/70%, respectively.

• Visual Selective Attention Mid-Test Associative Learning Manipulation:

UFOV: In the case of the UFOV, the simplest method to decrease difficulty is to increase presentation duration. Thus, in the UFOV Associative Learning Mid-tests, the proportion of long presentation duration trials will be increased (easier) or decreased (more difficult).

ANT: In the case of the ANT, the Associative Learning Mid-tests will be made easier by increasing the proportion of neutral trials or made more difficult by decreasing the proportion of neutral trials and increasing the proportion of incompatible (most difficult) trials.

• Spatial Cognition Mid-Test Associative Learning Manipulation:

Mental Rotation: Difficulty in mental rotation is monotonically related to the degree of rotation between the two items (larger rotations = more difficult). The Mental Rotation Associative Learning Mid-tests will thus be constructed to contain a disproportionate number of items with either small (easier) or large (more difficult) orientation differences.

Paper Folding: In the paper folding task, more difficult trials involve not only more total folds, but more partial folds and more angled partial folds. The Paper Folding Associative Learning Mid-tests will thus contain a disproportionate number of low total folds/low partial folds trials (easier) or high total folds/high partial folds trials (more difficult).

• Fluid Intelligence Mid-Test Associative Learning Manipulation: All matrix items that will be utilized have a known level of difficulty. Thus, the Fluid Intelligence Associative Learning Mid-test sets will be created by mixing in a disproportionate number of easier/more difficult items for both Raven's and new matrix task created by PIs Seitz and Jaeggi.
• Vocabulary Mid-Test Associative Learning Manipulation: As with the fluid intelligence test, the items have known levels of difficulty, and thus, the Associative Learning Mid-test will have a disproportionate number of easier/more difficult items for both the Mill-Hill vocabulary and the Shipley Institute of Living scales.

Knowledge of the Hypothesis Scales: Prior to Post-Test1, participants will be given two measures - the Perceived Awareness of the Research Hypothesis scale and the Expectation Assessment scale. Both measures are designed to assess the extent to which participants have internalized/believed the expectations that were meant to be induced.

Unblinding: Participants in all groups will be given a thorough debriefing after Post-Test1 and before Post-Test2 as to the goals of the study, the conditions, and any deception that they may have been subject to.

Post-Test1 and 2 will be identical to the Pre-test cognitive tests (new trials/items will be used in all cases).

Inclusion criteria

• Self-reported normal or corrected-to-normal vision
• No known neurological impairments (this includes any neurological impairments that would negatively impact participants' ability to perform perceptual or cognitive tasks or to complete long-term cognitive training; this could include neurological damage due to stroke in various brain areas; seizure conditions that would preclude the ability to view flashing images; motor control issues that preclude the ability to respond via button presses; etc.).
• Access to the internet, a computer, and a hand-held device, such as a cell phone or tablet (online version only)

Exclusion criteria

• Self-reported non-normal or non-corrected-to-normal vision
• Neurological impairments
• No access to the internet, a computer, or a hand-held device (online version only)