The multiple-choice science exercises used by the National Assessment of Educational Progress include an "I Don't Know" (IDK) alternative to estimate more accurately knowledge of groups of respondents. Group percentages of IDK responses were examined and compared with correct responses to see if the IDK introduces bias. Variance common to IDK and correct responses was reasoned to reflect bias related in part to personality variables. Residuals from a regression analysis predicting correct response from IDK response percentages were analyzed as modified correct response effects. The modification generally reduces differences among groups and smoothes trends across the four age levels. (Author) Primary type of information provided by report: Procedures (Packaging) (Scoring).