Feedlot veterinarians' moral and instrumental beliefs regarding antimicrobial use in feedlot cattle

Abstract

This cross-sectional survey research study examined the role moral beliefs play in predicting behavioural beliefs and attitudes and the role that subjective norms play in predicting moral beliefs. Using a self-administered questionnaire, one hundred and three feedlot veterinarians completed measures of behavioural beliefs, referent others, perceived constraints and moral beliefs regarding recommendations to use antimicrobials in four situations (i.e. acutely sick cattle, chronically sick cattle, at-risk cattle and high-risk cattle). Regression analysis and F-tests indicate moral beliefs as contributing significant increases in R 2 to models predicting behavioural beliefs regarding antimicrobial use in each situation. In addition, subjective norms contribute a significant increase in R 2 in models predicting moral beliefs in each of the four situations. The results indicate the effects of moral beliefs on behavioural beliefs are somewhat contingent on the condition; that is the level of risk associated with treating cattle with antimicrobials, the level of risk of not doing so, and the effectiveness of the antimicrobial in situations such as acute illness or being at-risk of illness.

Ordinary Least Squares Regression of Ordered Categorical Data: Inferential Implications for Practice

“…Using simulation studies, I evaluate empirical Type I error (α) and statistical power for inference on ordered categorical responses under a variety of scenarios. These simulation scenarios are motivated by subject-matter applications encountered in the scientific literature (Abreu et al, 2008;De Groote et al, 2010;Hernandez et al, 2005;Javaras & Ripley, 2007;McIntosh et al, 2009;Raubertas & Shook, 1982;Russell & Bobko, 1992;Xu et al, 2007;Yi et al, 2007;U.S., n.d. ).…”

“…According to Javaras and Ripley (2007), ordered categorical data that are generated as human reactions to a survey tend to be influenced by response styles and are frequently asymmetrical. An exemplification of this may be seen in outcomes from quality of life scales (Abreu et al, 2008) and in responses to antimicrobial use in feedlot cattle (McIntosh et al, 2009 It should be noted that, in contrast to continuous variables which have an infinite number of possible levels that they may assume, ordered categorical responses may only assume a finite number of levels. For illustrative purposes, I may examine previously mentioned examples in relation to their number of levels.…”

“…Before being analyzed by Yu as ordered categories, the previously mentioned phenotype example was analyzed with interval level techniques by Rocha, Eisen, Seiwerdt, Vleck and Pomp (2004). McIntosh et al (2009) analyzed the numerical scores assigned to the categories "always", "often", "sometimes", "rarely" and "never" as though they were realizations of a continuous variable. Russel and Boboko (1992) set up a scenario based on willingness to revise a manuscript as measured by a 5-point scale ranging from "very unmotivated" to "unmotivated," where not only did they wish to employ linear regression, but to investigate interaction effects as well.…”

“…(Liu & Agresti, 2005). Examples of interest in agriculture include quality assessments, such as for soil (De Groote et al, 2010) or food products (Hernandez et al, 2005), evaluation of lesion severity, such as teat ends status in dairy cattle (Raubertas & Shook , 1982), and antimicrobial use in feedlot cattle (McIntosh et al, 2009 (Liu & Agresti, 2005).…”

“…To demonstrate the application of OLSLR to a real-world problem, data used by McIntosh et al (2009) were analyzed. In that study, the researchers evaluated use of antimicrobials in feedlot cattle as it relates to Icek Ajzen's theory of planned behavior (Ajzen, n.d.).…”