Statistics anxiety has been defined as "feelings of anxiety encountered when taking a statistics course or doing statistical analyses; that is gathering, processing and interpreting data" (Cruise et. al., 1985). Statistics anxiety is negatively related to statistics achievement (Baloglu & Zelhart, 2003). Previous studies concerning international students have dealt with statistics anxiety in Production and Operations Management courses (Bell, 1998, 2001, 2003). The mathematical preparation of international students is superior to the preparation of the domestic students (Madison & Hart, 1990). Mathematics background plays an integral role in statistics anxiety. Students with an enhanced calculus preparation (a five semester hour calculus course or high school calculus) significantly lowered their anxiety levels during a business statistics course (Bell, 2003). Will the enhanced mathematical preparation cause the international student to exhibit less statistics anxiety? Another factor to consider is the format of the course. Students who took statistics in an accelerated time frame encountered higher levels of statistics anxiety (Bell, 2001,2005). Students who took a beginning course entirely via the computer encountered lower levels of statistics anxiety than those students who opted for a traditional course (Bell & Weller). Both the lack of positive feedback and the lack of encouragement by student instructors were found to lead to a negative student perception about the course and thus increased levels of statistics anxiety (Onwuegbuzie, 1997). Other studies have found that females tend to flourish in same-sex classrooms (Campbell & Evans, 1997) and women experienced higher levels of statistics anxiety than men (Onwuegbuzie, 1993). Statistics levels have been significantly lowered by removing time constraints when testing (Onwuegbuzie, 1995). However, relatively few studies exist that consider statistics anxiety concerning international students in the beginning statistics course.
In the spirit of Corey's "Action Research" of the 1950s, the Statistics Anxiety Rating Scale (STARS) was administered at the beginning of a night course (one night/week, three hours/night, fifteen weeks) in business statistics. The night class was chosen for two reasons: students taking statistics in a night class suffered higher levels of statistics anxiety than those who opted for the traditional MWF or TR format and the night course seemed more popular with international students. What is "Action Research"? Corey (1954) defined "Action Research" as "deliberate, solution-oriented investigation which is designed, conducted, and implemented by teachers themselves to improve teaching in the classroom." The subjects in this experiment were sixty-six students in a beginning statistics class, twelve international students and fifty-four domestic students. Due to the small number of international students and the ordinal level of measurement of the Likert scale in STARS, the nonparametric Mann-Whitney-Wilcoxon T test was employed. The parametric test for independent samples requires that both populations are normal and have equal variances whereas the Mann-Whitney-Wilcoxon T test only requires at least ordinal data. Since both sample sizes are at least ten, the normal approximation to T can be used (Anderson, et. al. 1996).
The Statistics Anxiety Rating Scale (STARS) consists of two parts. The first part presents twenty-eight situations often associated with statistics anxiety. These items are scored on a Likert-type scale from one to five, with a "one" indicating no anxiety with that situation while a "five" indicates considerable anxiety. The second part consists of twenty-eight statements dealing with statistics, with responses recorded on a Likert-type scale from one (no anxiety) to five (considerable anxiety). Hence, lower scores indicate lower levels of statistics anxiety. …