#1: Reporting measurements with unnecessary precision
#2: Dividing continuous data into ordinal categories without explaining why or how
#3: Reporting group means for paired data without reporting within-pair changes
#4: Using descriptive statistics incorrectly
#5: Using the standard error of the mean (SEM) as a descriptive statistic or as a measure of precision for an estimate
#6: Reporting only P values for results
#7: Not confirming that the data met the assumptions of the statistical tests used to analyze them
#8: Using linear regression analysis without establishing that the relationship is, in fact, linear
#9: Not accounting for all data and all patients
#10: Not reporting whether or how adjustments were made for multiple hypothesis tests
#11: Unnecessarily reporting baseline statistical comparisons in randomized trials
#12: Not defining “normal” or “abnormal” when reporting diagnostic test results
#13: Not explaining how uncertain (equivocal) diagnostic test results were treated when calculating the test's characteristics (such as sensitivity and specificity)
#14: Using figures and tables only to “store” data, rather than to assist readers
#15: Using a chart or graph in which the visual message does not support the message of the data on which it is based
#16: Confusing the “units of observation” when reporting and interpreting results
#17: Interpreting studies with nonsignificant results and low statistical power as “negative,” when they are, in fact, inconclusive
#18: Not distinguishing between “pragmatic” (effectiveness) and “explanatory” (efficacy) studies when designing and interpreting biomedical research
#19: Not reporting results in clinically useful units
#20: Confusing statistical significance with clinical importance
