An empirical study on counterfactual visualization to support visual causal inference explores how counterfactuals can enhance users' understanding of causal relationships in data visualizations. The study investigates the impact of counterfactual visualizations on users' ability to interpret datasets, focusing on four levels of comprehension: recognizing, understanding, analyzing, and recalling. The research involved an empirical study with 32 participants, who were shown different combinations of chart sets displaying varying levels of counterfactual information. Participants were asked to answer questions related to three design objectives: recognizing correlations, making predictions, and identifying causal relationships. The study found that counterfactuals significantly improved participants' ability to understand and draw inferences from datasets, while also improving recall. However, counterfactuals required longer response times for answering questions. The study also suggests that counterfactuals do not impair users' ability to read charts. Based on these findings, the study proposes a set of design heuristics to guide the integration of counterfactuals into data visualizations. The study's results support the hypothesis that counterfactuals can help users better understand and analyze causal relationships in data. The study also highlights the importance of considering the cognitive processes involved in visual data communication when designing counterfactual visualizations. The study's findings contribute to the understanding of how counterfactuals can be effectively used in data visualizations to support causal inference.An empirical study on counterfactual visualization to support visual causal inference explores how counterfactuals can enhance users' understanding of causal relationships in data visualizations. The study investigates the impact of counterfactual visualizations on users' ability to interpret datasets, focusing on four levels of comprehension: recognizing, understanding, analyzing, and recalling. The research involved an empirical study with 32 participants, who were shown different combinations of chart sets displaying varying levels of counterfactual information. Participants were asked to answer questions related to three design objectives: recognizing correlations, making predictions, and identifying causal relationships. The study found that counterfactuals significantly improved participants' ability to understand and draw inferences from datasets, while also improving recall. However, counterfactuals required longer response times for answering questions. The study also suggests that counterfactuals do not impair users' ability to read charts. Based on these findings, the study proposes a set of design heuristics to guide the integration of counterfactuals into data visualizations. The study's results support the hypothesis that counterfactuals can help users better understand and analyze causal relationships in data. The study also highlights the importance of considering the cognitive processes involved in visual data communication when designing counterfactual visualizations. The study's findings contribute to the understanding of how counterfactuals can be effectively used in data visualizations to support causal inference.