Rethinking Consistency in Cronbach’s Alpha
Jerry J. Vaske
Carly C. Sponarski
Cronbach alpha estimates the internal consistency of responses in multi-item bipolar scales. This article examined four research questions (RQ): (a) To what extent do inconsistencies exist in data (e.g., responses of -2 -2 2 2)? (b) Do the number of scale items influence the amount of inconsistency? (c) Does Cronbach alpha mask inconsistencies? and (d) Does the pattern of findings generalize across different contexts? Data were obtained from 29 research projects (n=10,616). Each survey had place attachment questions composed of two concepts: place identity and place dependence. Respondents were classified as consistent or inconsistent based on their responses to the place attachment questions. Results demonstrated that: (a) inconsistent response patterns exited in the data (RQ1), (b) number of scale items influenced amount of inconsistency (RQ2), (c) Alpha masked these inconsistencies (RQ3), and (d) this pattern generalized across different resources and respondents (RQ4). Discussion focused on implications of these findings.
Manuscript Download pdf file
SAS syntax and data Download zip file
SPSS syntax and data Download zip file
Output Download pdf file