Paradigm and Hypotheses

This document plots the results of a 30 min version of Auditory Ensembles. In this task, subjects performed an auditory memory task in which they remembered a single pure tone (300 ms duration) over a delay with distractors. Subjects remembered the first tone of the presentation, which was then followed by 4 pure tone distractors after a 2000 ms ISI. Each of these distractor tones had an ISI of 300 ms. The tones were either +/-1, 3, or 5 semitones from the remembered tone and were chosen such that their average pitch was either equal to, higher than, or lower than the to-be-remembered tone with equal probability. Following the final distractor tone, there was a 2 sec delay. Participants then reported which of two test tones they heard at the beginning of the trial (i.e. 2IFC) using the v (first) and n (second) keys. There was an ITI of 3000 ms. We designed the task to see if ensemble information biases individual auditory memory representations. We hypothesized that subjects would be more likely to pick foils when they were in the same direction from the remembered tone as the ensemble (i.e. less accurate in task), and less likely to pick foils when they were in the opposite direction from the remembered tone as the ensemble (i.e. more accurate).

In the 3_3 version, we found that subjects were not reliably above chance in performing the task despite lengthening the time between the target and distractors. In this version, we are attempting to make the task even easier by increasing the distance between the target and foil to +/- 4 from +/- 2.

Note: No error bars yet! Subjects who had a d’ of less than .1 after collapsing across all conditions were excluded, which is (much) lower than we typically like but was frankly necessary given the performance on the task.

Subjects’ (N = 30) overall performance, before exclusion

Subjects’ (N = 12) performance on task, after exclusion

Target discrimination d’

Target discrimination RT

Manipulation Checks

d’ with target first vs. second

Frequency of choosing first vs. second interval

## `stat_bindot()` using `bins = 30`. Pick better value with `binwidth`.

d’ with foil higher vs. lower

Statistics

d’ ANOVA

Effect DFn DFd SSn SSd F p p<.05 ges partial_eta_squared
(Intercept) 1 11 4.534 1.532 32.554 0.000 * 0.547 0.747
taskCondition 2 22 0.074 2.215 0.367 0.697 0.019 0.032


RT ANOVA

Effect DFn DFd SSn SSd F p p<.05 ges partial_eta_squared
(Intercept) 1 11 2927502155 23906049251 1.347 0.270 0.039 0.109
taskCondition 2 22 4369031172 47592333721 1.010 0.381 0.058 0.084