The association between two bogus items, demographics, and military characteristics in a 2019 cross-sectional survey of U.S. Army soldiers.

Careless responding to surveys has not been sufficiently characterized in military populations. The objective of the current study was to determine the proportion and characteristics of careless responding in a 2019 survey given to a large sample of U.S. Army soldiers at 1 installation (n = 4,892). Two bogus survey items were asked to assess careless responding. Nearly 96% of soldier respondents correctly answered both bogus items and 4.5% incorrectly answered at least 1 bogus question. In the adjusted multiple logistic regression model, race and marital status were associated with incorrect answers to bogus item questions after controlling for all other covariates. Specifically, the odds of Black respondents incorrectly answering the bogus items (adjusted odds ratio [AOR]: 2.53; 95% CI: 1.74-3.68) were more than 2.5 times those of White respondents. The recommendations that stem from the results of surveys can influence policy decisions. A large proportion of careless responses could inadvertently lead to results that are not representative of the population surveyed. Careless responding could be detected through the inclusion of bogus items in military surveys which would allow researchers to analyze how careless responses may impact outcomes of interest.