The Relationships Among Survey Page Length, Progress Indicators, and Item Completion Rates

• In Event: Roundtable Session 97
  In Roundtable Session: Item Features and Their Role in Validity

Abstract

This study investigates the effect of several web survey design techniques on item completion rates. In 2009, the NSSE staff created six NSSE web surveys interfaces that included various numbers of pages (4 and 8 pages) and progress indicators (a progress bar and page numbers). Experimental survey versions were offered to eight of the 617 participating schools, with the total respondent sample across all schools equaling 333,923. The multi-level model analysis revealed that all three 8-page surveys were performing worse than the standard NSSE version. Only 2% of the variance in the outcome was found between institutions. Student-level predictors were not significant in predicting the number of completed items. Analyses run with level-2 predictors are expected to produce different results.

Authors

• Malika Tukibayeva, Indiana University

• Shimon Sarraf, Indiana University