Incorrect responses

The current analysis counted reaction times from responses on the lexical decision task regardless of whether they were correct. Other lexical decision task studies (e.g., Matthews and Southall) count only reaction times for correct decisions. This technique may eliminate guessing biases. Thus, the same analyses as discussed above were re-run using only correct responses. When incorrect responses are eliminated from the data set, the Depression by Task by Valence interaction is still statistically significant, and the effect size increases considerably from the condition in which incorrect responses were not eliminated, F(2,68)=9.23, p<0.0005,t#tex2html_wrap_inline1496#=0.214. Additionally, when incorrect responses are eliminated, differences between the SD's are more strongly represented, and the depression by SD by Valence interaction is statistically significant, F(4,66)=3.62, p=0.01,t#tex2html_wrap_inline1498#=0.18. When means are compared to the condition in which incorrect responses are not removed the differences are very slight. There is a two millisecond greater difference between negative and neutral reaction times in the 150ms condition for depressed people, though there is a slightly smaller delay shown by depressed people with respect to nondepressed people, since some of the long erroneous reaction times were omitted. As in the original analysis, simple effects analyses of this condition do not reveal a statistically significant effect of valence for depressed people on the lexical decision task, F(2,28)=1.4, p=.261, t#tex2html_wrap_inline1500#=.091. Differences in means are reported in Table 4, p. 78.

Lexical decision task, nondepressed people

 

Table: Lexical Decision Task, Nondepressed and Depressed People

condition	all responses	correct only
50ms
positive	.64	.61
negative	.62	.59
neutral	.64	.61
100ms
positive	.53	.52
negative	.54	.52
neutral	.56	.53
150ms
positive	.48	.46
negative	.46	.45
neutral	.50	.48

Lexical decision task, depressed people

condition	all responses	correct only
50ms
positive	.67	.60
negative	.66	.60
neutral	.69	.64
100ms
positive	.57	.51
negative	.61	.56
neutral	.57	.54
150ms
positive	.51	.48
negative	.54	.52
neutral	.51	.47