## Data from Homework 2: Blocking of Associative Learning

 Table 1': Mean Responses of Many Participants (3 = Probably A, 1 = Maybe A, -1 = Maybe B, -3 = Probably B. Individual mean computed first for each cell, then mean across 71 individuals is tabulated here.) Disease A Disease B Blocking Overshadowing 2.92SS/df=0.078N=71 -2.96SS/df=0.041N=71 -0.31SS/df=0.44N=71 0.51SS/df=0.51N=71

### Memory for Training Cases:

For both diseases, accuracy during the test phase was extremely good.

### Blocking Test Cases:

There was a blocking effect, on average, with people tending to prefer disease B for the blocking cases. That is, the mean of -0.31 is significantly less than zero, t(70)=3.95, p<.01.

There were notable individual differences, however. Below is a frequency distribution of individual means for the blocking cases. It can be seen that many people showed no blocking effect, and a few even the opposite of a blocking effect. But a significant number of people did exhibit blocking.

 ``` Cum Value Frequency Percent Percent -3.00 1 1.4 1.4 -2.00 1 1.4 2.8 -1.50 2 2.8 5.6 -1.00 11 15.5 21.1 -.50 18 25.4 46.5 .00 29 40.8 87.3 .50 6 8.5 95.8 1.00 3 4.2 100.0 ------- ------- Total 71 100.0 | -3.00 |** 1 | -2.00 |** 1 | -1.50 |*** 2 | -1.00 |****************** 11 | -.50 |****************************** 18 | .00 |************************************************ 29 | .50 |********** 6 | 1.00 |***** 3 | +---------+---------+---------+---------+---------+ 0 6 12 18 24 30 Frequency ```

There was an overshadowing effect, on average, with people tending to prefer disease A for the overshadowing cases. That is, the mean of 0.51 is significantly greater than zero, t(70)=6.04, p<.01.

There were notable individual differences, however. Below is a frequency distribution of individual means for the overshadowing cases. It can be seen that many people showed no overshadowing effect, and a few even the opposite of an overshadowing effect. But a significant number of people did exhibit overshadowing.

 ``` Cum Value Frequency Percent Percent -.50 5 7.0 7.0 .00 29 40.8 47.9 .50 11 15.5 63.4 1.00 18 25.4 88.7 1.50 4 5.6 94.4 2.00 2 2.8 97.2 2.50 1 1.4 98.6 3.00 1 1.4 100.0 ------- ------- Total 71 100.0 | -.50 |******** 5 | .00 |************************************************ 29 | .50 |****************** 11 | 1.00 |****************************** 18 | 1.50 |******* 4 | 2.00 |*** 2 | 2.50 |** 1 | 3.00 |** 1 | +---------+---------+---------+---------+---------+ 0 6 12 18 24 30 Frequency ```

### Correlation of Blocking with Overshadowing

It turns out that there was a significant correlation of blocking with overshadowing. That is, if a person showed overshadowing, then he or she tended also to show blocking. The correlation was highly significant, with r = -.612. A scatterplot of 71 individuals is shown below. Notice that many people (21) showed zero blocking and zero overshadowing, but many other people showed both, with covarying magnitude.

 ``` Blocking vs. Overshadowing ++----+----+----+----+----+----+----+----+----+----+----+----+----++ 1+ 1 1 1 + | | | | | | | | .5+ 1 4 1 + | | | | | | | | 0+ 2 (21) 3 2 1 + | | | | B | | l | | o -.5+ 1 2 7 8 + c | | k | | i | | n | | g -1+ 1 1 6 3 + | | M | | e | | a | | n -1.5+ 1 1 + | | | | | | | | -2+ 1 + | | | | | | | | -2.5+ + | | | | | | | | -3+ 1 + ++----+----+----+----+----+----+----+----+----+----+----+----+----++ -2 -1 0 1 2 3 4 Overshadowing Mean ```