Figure 1.
Association strength (AS) of all 800 word-pairs that were used.
The blue shaded area contains 100 word-pairs with an AS of zero, meaning the words are completely unrelated. The red shaded area contains the 100 word-pairs with the highest AS in the association norm data. Both words are four to six letters in length. The AS of the remaining word-pairs follows a log scale.
Table 1.
Descriptive statistics of the bins across all conditions.
Figure 2.
Boxplots of button responses made by the subjects.
Each boxplot corresponds to a bin (see table 1). Whiskers extend to the inner quartile range. Outliers are plotted as semi-transparent dots, with some random jitter applied to the x () and y (
) position to reduce overlap. Bin 1 contains trials with a high AS and subsequent bins contain trials with lower AS. Bin 8 contains mostly trials with
(the trials in this bin with
are mostly outliers in the corresponding boxplot). The y-axis shows the number of subjects that rated the two words of the word-pair as being related, divided by the total number of subjects (10). A: responses during the speeded condition. B: responses during the delayed condition.
Figure 3.
Stimulus-locked ERPs and scalp topographies for both experimental conditions.
Vertical lines are plotted to show the mean RT of the trials belonging to each bin (Table 1), using the same color code as the ERPs. Scalp topographies are drawn for components of interest, using red for positive and blue for negative values. Note that the y-axes have different scales and each scalp topography uses its own normalized scale. A: Stimulus-locked ERP of the speeded condition; trials sorted by decreasing AS. The scalp topography shows the difference between bins 1 and 8 at 400 ms. B: Stimulus-locked ERP of the speeded condition; trials sorted by increasing RT. Scalp topography shows the difference between bins 1 and 8 at 400 ms. C: Stimulus-locked ERP of the delayed condition; trials sorted by decreasing AS. A vertical dotted line indicates the moment the target word turned yellow, which cued the response of the subject. Two scalp topographies are drawn, one showing the difference between bins 1 and 8 at [400]ms, the other showing the voltage at 1300 ms relative to zero.
Figure 4.
Response-locked ERPs and scalp topographies for both experimental conditions.
Shown are the grand response-locked ERPs across all bins and all subjects for the speeded condition (thick line) and delayed condition (thin line). Scalp topographies are given showing the voltage at 0 s, using red for positive and blue for negative values. Note that both scalp topographies use their own normalized scale.
Figure 5.
Two-dimensional histogram of single trial component latency versus RT.
A hexagonal grid shows histograms of component latency versus RT for all non-rejected trials of all subjects. The values along the x and y axes indicate seconds since the onset of the target stimulus. Component latency was determined through iterative template matching (see the results section). A: P3 latency during the speeded condition; superimposed as a red line is . B: P3 latency during the delayed condition, versus the mean RT of each word-pair collected during the speeded condition.
Figure 6.
Grand average ERPs demonstrating the effect of differences in P3-latency. A
: Speeded condition, low AS () and short P3-latency (
ms, thick line) versus high AS (
) and long P3-latency (
ms, thin line). Vertical lines indicate the mean P3-latency of each group. B: Delayed condition, using the same subject/word-pair combinations used to create A. C: Speeded condition, low AS (
, thick line) versus high AS (
, thin line) using a fixed P3-latency and RT (around
ms). Vertical lines indicate the mean P3-latency of each group, which overlap in this case.