Replication of Nagler (1994) and Alvarez and Brehm (1995) with Special Attention to Numeric Accuracy

In the course of an ongoing research project testing numerical accuracy of statistical techniques and software frequently used by political scientists we replicate two studies, Nagler (1994) and Alvarez and Brehm (1995). We find numeric inaccuracies, errors in formulas, and data problems that reduce the strength of the substantive conclusions of the studies. Our replication analyses reveal there are a number of factors that we may take for granted in the normal course of our analysis that affect the ability to faithfully replicate. More attention to these details may improve replication of research in the discipline.