Skip to main content
. Author manuscript; available in PMC: 2012 May 27.
Published in final edited form as: Cognition. 2010 Nov 27;118(3):377–397. doi: 10.1016/j.cognition.2010.10.008

Table 1.

Experiment 1 Mean Semantic Integration Ratings, Mean Attachment Preferences, and Response Counts by Condition

Condition Semantic Integration Rating
%N1 Attachment Response Count
N1-N2 N1-N3 N2-N3 Error Correct Uninflected Misc
Descending
SSS 4.13 (.94) 5.23 (.49) 4.46 (.83) 4.7 (7.01) 1 (0) 145 (19) 37 (5) 33
SPS 4.34 (.79) 5.08 (.47) 4.22 (.72) 5.6 (5.57) 15 (2) 129 (13) 29 (5) 43
SSP 4.02 (.77) 5.21 (.45) 4.20 (.53) 4.1 (5.52) 2 (0) 124 (12) 31 (5) 59
M 4.16 (.83) 5.17 (.46) 4.29 (.70) 4.8 (5.93)

Flat
SSS 4.21 (1.38) 5.24 (.98) 2.81 (.94) 93.6 (5.68) 1 (0) 148 (20) 38 (4) 29
SPS 4.11 (1.36) 4.96 (.99) 2.62 (.91) 91.9 (6.68) 19 (2) 114 (10) 31 (5) 52
SSP 3.97 (1.23) 5.13 (.97) 2.64 (.75) 94.4 (5.57) 5 (2) 128 (10) 33 (2) 49
M 4.10 (1.29) 5.11 (.96) 2.69 (.85) 93.3 (5.92)

Grand Mean 4.13 (1.08) 5.14 (.75) 3.49 (1.12) 49.0 (45.0)

Note. Conditions are indicated by noun number (N1, N2, N3), with S = singular and P = plural. Semantic integration rating scale was 1 (loosely linked) to 7 (tightly linked). For semantic integration ratings and attachment, means are computed by-items and standard deviations are in parentheses. For response counts, dysfluency counts are in parentheses. Misc = Miscellaneous.