The effects of data model representation method on task performance

Data models are frequently created with little thought about their layout. Our research examined how data models organized in a columnar fashion compared to equivalent data models which did not, by improving novice designer's performance in error detection, comprehension, and query construction.Experimental results showed that the columnar organized data model improved performance in error detection and querying, but not in comprehension. Also, the columnar organized data model was perceived as being more useful. The results of our research have important implications for data model design, especially for novices who must interpret and use them.