What is Implicit About Implicit Category Learning?

RESULTS 
Behavioral Results 
Accuracy 
Figure 7 displays average accuracy by block for the RB (n = 6) and II (n = 5) conditions. 
A two-way, mixed-factor ANOVA showed neither a significant effect of block, F(2, 18) = 2.103, 
p = .151, nor condition, F(1, 9) = 0.90, p = .771. 
Figure 7: Average participant accuracy over blocks 
 
Before analyzing attribution responses, we sorted participants into two categories 
(“guessers” and “non-guessers”) to determine whether we could predict task accuracy by the 
extent to which participants reported the “guess” criterion. This designation was based on a 
median split using the proportion of guess responses. “Guessers” reported guess at a proportion 
0.500
0.550
0.600
0.650
0.700
0.750
0.800
0.850
0.900
0.950
1.000
1
2
3
Ac
cura
cy
Block
Average Accuracy by Block
RB average
II average

---

14 
0.500
0.600
0.700
0.800
0.900
1.000
1
2
3
Ac
cura
cy
Block
II Accuracy Sorted by Guess
Non-
guessers
Guessers
of 0.5 or higher over blocks while “non-guessers” reported that they were guessing less than half 
the time, i.e., reported an average of the three other criteria (intuition, memory and rule) more 
than half of the time. In the RB condition (n = 6), there were three “guessers” and three “non-
guessers. In the II condition (n = 5), there were two “guessers” and three “non-guessers.” In both 
conditions, we predicted that “non-guessers” would perform with higher accuracy than 
“guessers,” which is plotted in Figure 8. 
Figure 8: Accuracy for “guessers” and “non-guessers” by condition 
“Non-guessers” performed with higher accuracy in both the RB and II conditions across blocks.