Spreadsheets have been used by organizations for decades. Errors in spreadsheets are commonly found in laboratory and field findings. In recent years, many exciting new visualization techniques have been developed to help users understand spreadsheet models and to check for errors. Two visualization tools were tested in an experiment for their effects on error correction. The first is a simple arrow tool which shows dependencies among cells. The second shows the inputprocess- output function of cells in addition to the dependency arrows. The experiment shows significantly better error detection with the arrow method than for the plain method (without visualization tools). Wrong data errors took more time to correct than missing data errors.