Repulsion depends on task demands
Next we asked wether the repulsion of similar color values depends on the task demand to separate the objects or if it is a byproduct of any learning. In the previous experiments, the associative memory task required participants to distinguish between the two similar colored objects. Here, the experiment was run identically to Experiment 1, expect we altered the task so that participants had to generalize across the two similar colored objects and tested if the repulsion bias would remain, go away, or possibly flip so that there would be an attraction of the color values towards one another. Specifically, during the test, one of the learned faces appeared and participants had to select the face (from a set of 4 faces) that was paired with the same object. With this subtle task change, the color of the objects was no longer relevant to optimal performance on the associative task. Rather, the object identity was the important feature to link between the two faces in memory.
Participants successfully learned to infer the face pairings as they chose the target face at rates well above chance in all conditions (ps <0.00001). Color error also decreased over learning.
After 8 rounds of learning color error varied by similarity condition (p = 0.003). Follow up t-tests revealed that there was larger color errors for objects in the low similarity condition compared to the both the high (p = 0.007) and the moderate (p=0.025) similarity conditions. Color error did not differ between the high and moderate similarity conditions (p=0.11). Turning to our main question of interest we next investigated if the change in task demand altered the bias in color estimates. As in Experiment 1 both measures of color bias varied by similarity condition (percent away: p=0.015; signed error: p = 0.002), however, in this Experiment there was no color bias in the high similarity condition (percent away: p = 0.56, signed error: p = 0.29) while both the moderate (percent away: p=0.06; signed error: p = 0.007) and high (percent away: p=0.00004; signed error: p = 0.00009) similarity condition showed positive bias towards the competing color. Direct t-test comparisons across the experiments revealed a significant difference in color bias for the high similarity (percent away: p = 0.001; signed error: p=0.009)and low similarity (percent away: p = 0.001; signed error: p = 0.002) conditions but not for the moderate similarity condition (percent away: p = 0.11; signed error: p = 0.38). Thus, changing the task demand eliminated the repulsion bias in the high similarity condition and created an attraction bias in the low similarity condition.
Methods
Experiment 1
Subjects
Materials
Stimuli consisted of 18 object images and 36 face images. The object images were selected from a set of images designed to be color-rotated(Brady et al. 2013). Objects were chosen that had no strong association with a particular color or set of colors. To alter the color of each object, the hue of an image was rotated through a circular colorspace ranging from 0-360 degrees. Colors were altered by independently rotating every pixel through an equiluminant circle in L*a*b* space. Face images were pictures of white males. For each subject, object images were randomly assigned to each of the three color difference conditions (24 degrees, 48 degree, 72 degree). Each object was then assigned two colors separated by the hue angle degree difference of their respective condition. To do so, the colorspace was sub-selected to include 45 colors each separated by 8 degrees. Colors were then assigned to objects with the following constraints 1) within a condition, there was an even distribution of colors across thirds (120 degrees) of the colorspace, and 2) each color was only assigned once across all objects. Each colored object was then assigned a unique face associate.
Procedure
The experiment consisted of an associative learning task. Participants learned the object-face pairs through 8 rounds of iterative study and test. During a study block, the face and object appeared on screen side by side and subjects had 2.5 seconds to study the pairing. All 36 pairs were presented in a round. Next participants were tested on their color memory for each object. On each trial a studied face was presented along with its paired object presented in greyscale. Color memory was assessed using the method of adjustment. A trial started with a cursor in the center of the object image. Once participants moved the mouse the object would appear in color. The hue of the object was determined by the angle between the mouse and the center of the object image. A line marker moved around a ring surrounding the object image to indicate the current hue angle. Once participants rotated to the desired color they clicked the mouse to finalize their choice. The color wheel was randomly rotated across trials so there was no correspondence between spatial position and color. Following the color memory test, participants were tested on their associative memory for all object-face pairs. On each trial, a colored object was presented at the top of the screen. Participants were instructed to select the face paired with the image from a choice of four faces presented below. The target face was always presented along with the competitor face paired with the object's pairmate (i.e the same object in a different color). Thus participants had to discriminate between the objects' colors in order to select the correct face. The other two non-competitor foils were two randomly selected faces. Each colored object served as a non-competitor foil on exactly 2 trials. Participants made a selection using a computer mouse. Participants had unlimited time to answer. Feedback was given indicating whether their choice was correct along with the correct colored object-face pairing displayed on screen for 1 second.
Following the 8 rounds of learning participants completed a post test which consisted of an extended color memory test. Trials in this color memory test were identical to those in the learning phase, however in this test, each object was tested 5 times. The order of trials was randomized across the entire test with the constraint that an object and its pairmate were not tested on successive trials. Memory estimates were averaged across the 5 trials to reduce the noise associated with individual trials. Participants returned 24 hours later for a second post test. This test was identical to the previous day's test except the order of trials was re-randomized.
Analysis
The error in color memory was computed by taking the absolute value difference between a participant's color estimate and the true color. Bias in color memory was then assessed by measuring the error in color estimates in relation to the object's competing pairmate color. Specifically, a color bias measure was computed by signing the color error: color estimates positioned towards the competing color were coded as the positive color error, and color estimates positioned away from the competing color were coded as the negative color error. In addition to this continuous measure of color bias, we also computed a complimentary measure of bias that was agnostic to the magnitude of the color error. Each color estimate was coded as towards or away from the competitor color and then we computed the percentage of estimates across all trials that fell away from the competing color.
Experiment 3 (Integration)
Subjects
Materials
The stimuli used were identical to those used during Experiment 1.
Procedure
The procedure was the same as Experiment 1 except for a few changes. The critical change from the previous experiments was the associative memory test following each round. On each trial in this test a learned face appeared at the top of the screen. Participants were instructed to select the face paired with the same object using a computer mouse. The target face was always presented along with 3 other learned faces. All the faces were tested once per round and each face served as a non-target lure on exactly 3 trials. Feedback was given indicating whether their choice was correct along with the correct face-face pairing displayed on screen for 1 second. In addition to this change, a few other minor improvements were made to the previous study. First, to address the problem of participants not finishing the experiment within the allotted time, a trial time limit of 10s was imposed for the associative learning and color memory tests. Second, since we did not find any differences in color memory after a 24 h delay, we excluded the delayed color memory test from this experiment.
Experiment 4 (Ultrasim)
Subjects