The authors examined how the grading scheme affects learning and students' reactions to the Immediate Feedback Assessment Technique (IFAT), an answer form providing immediate feedback on multiple-choice questions. Undergraduate students (N = 141) took a general-knowledge multiple-choice test of low, medium, or high difficulty. They used the IFAT with a number-correct (NC), partial-credit (PC), or correction-for-guessing (CG) grading scheme. After 1 week, they retook the test using a traditional response form with an NC scheme. When the authors used the NC or PC scheme for Test 1, the number of correct answers increased by more than 30% on Test 2. However, the increase was only about half as large with the CG scheme, suggesting that it interferes with the IFAT's learning benefits. Responses to questionnaire items provided no strong clues regarding the origin of this difference. Participants in all treatment conditions had positive attitudes toward the IFAT. [ABSTRACT FROM AUTHOR]