Evaluating the Effects of Visualizing Missing Values on Data Exploration

People must often perform analysis on data that contains missing values. We conducted a preliminary empirical study to understand the effect of visualizing those missing values on participants’ visual data exploration and decision-making. Study participants purchased a hypothetical portfolio of stocks based on a data set where some stocks had missing attribute values. The stock data was shown in scatter plots. For one group of participants, stocks with missing values simply were not shown, while the second group saw such stocks depicted with estimated values as points with error bars as annotations to indicate missing values and to communicate estimated values. We measured participants’ awareness of missing values in the data set and their cognitive load in decision-making. Our results indicate that when missing values were visualized, participants reported a higher awareness level of the missing values, considered a higher number of individual data items, and their decision-making workflow was different.