Abstract
Articles on best practices in research usually focus on collecting and analysing data. However, an important ethical and practical issue is often ignored: false precision. Researchers, reviewers, and editors often ignore the precision of instruments and the concept of significant digits, familiar from introductory courses in many sciences. The result is that findings are presented so that they appear to be more precise or accurate than they actually are. Imprecision is also ignored ( and precision implied) when results are presented without margins of error ( confidence intervals). Other practices also increase or mask imprecision. It is not widely appreciated that imprecision is inflated when scale scores are calculated by summing items, a common practice for clinical instruments. The use of global scores can mask the complex, multidimensional nature of constructs such as stress, resilience, and depression. Although these practices are not intended to mislead or deceive, that is their effect when presented to policymakers, clients- and ourselves. Improvements are obvious: Reported results should reflect the precision of measurements; margins of error should be reported; scale scores should be calculated by averaging, not summing, items; and unidimensional scales should be used in research articles instead of global scores.