Testing the modifiability of the state sport confidence inventory.

Completing the State Sport Confidence Inventory requires participants to compare themselves with the most confident athlete they know. Inconsistency is inherent in participants' choices that may result in unsystematic variance across such responses. This study tested the modifiability of the inventory by removing the "comparison" component from the instructions and from each of the 13 questions. Using data from 31 female figure skaters (M age=16.1 yr., SD=1.0), the original and modified versions were compared. Adequate reliabilities were found for both measures (both rs=.94), but a paired sample t test indicated significant differences between mean scores on the two versions. The total score and all of the item scores were higher for the modified version than the original, and coefficients of determination showed variance shared by the modified and original items ranged only from 9 to 59% per item. These results suggest that the modified version was not an adequate representation of the original inventory.