Statistics Anxiety and Graduate Students

STATISTICS ANXIETY and graduate students in the social sciences

Many graduate students in the social sciences need to take statistics as part of the academic training, but these students often do not necessarily have backgrounds in statistics or mathematics from their undergraduate degree or other graduate training. In the classrooms, statistics anxiety is noticeably prevalent among graduate students whose academic background has little statistical training. According to Onwuegbuzie. Slate, Paterson, Watson, and Schwartz (2000), 75% to 80% of graduate students appear to experience uncomfortable levels of statistics anxiety. As a result, conducting statistics is often rated as the lowest skill in terms of academic competence (Huntley, Schneider, and Aronson, 2000).

Statistics anxiety has been defined simply as anxiety that occurs as a result of encountering statistics in any form and at any level (Onwuegbuzie, DaRos, & Ryan, 1997), and has been found to negatively affect learning (Onwuegbuzie & Seaman, 1995). Many researchers (Lazar, 1990; Lalonde & Gardner, 1993; Onwuegbuzie, 2000b) suggested that learning statistics is as difficult as learning a foreign language. On the other hand, statistics anxiety sometimes is not necessarily due to the lack of training or insufficient skills, but due to the misperception about statistics and negative experiences in a statistical class. For instance, students often think they do not have enough mathematics training so that they cannot do well in statistical classes. With fear of failing the course, they delay enrolling in statistics courses as long as possible, which often leads to failure to complete their degree programs (Onwuegbuzie, 1997). The lack of self-efficacy and higher anxiety in statistics keep many students away from engaging in research work or further to pursue an academic career. Therefore, statistics becomes one of the most anxiety-inducing courses in their programs of study (Blalock, 1987; Caine, Centa, Doroff, Horowitz, & Wisenbaker, 1978; Schacht & Stewart, 1990; Zeidner, 1991).

A growing body of research has documented a consistent negative relationship between statistics anxiety and course performance (Zeidner, 1991; Elmore et al.,1993; Lalonde & Gardner 1993; Onwuegbuzie & Seaman 1995; Zanakis & Valenza1997). In fact, statistics anxiety has been found to be the best predictor of achievement in research methodology (Onwuegbuzie et al., 2000) and statistics courses (Fitzgerald et al., 1996). Most recently, Onwuegbuzie (in press b), using pathanalytic techniques, found that statistics anxiety and expectation play a central rolein his Anxiety-Expectation Mediation (AEM) model, being related bi-directionallyto statistics achievement and, at the same time, moderating the relationship betweenstatistics achievement and research anxiety, study habits, course load, and thenumber of statistics courses taken. The AEM model is presented in Figure 1.Onwuegbuzie (in press b) posited that the pivotal role of statistics anxiety in theAEM model suggests that Wine's (1980) Cognitive-Attentional-Interference theorycan be applied to the field of statistics, as it can be to the foreign language learningcontext. According to Onwuegbuzie, Wine's theory predicts that anxiety interferes with performance by impeding students' ability to receive, to concentrate on, and toencode statistical terminology, language, formulae and concepts. Moreover, Onwuegbuzie theorised that anxiety reduces the efficiency with which memory processes are utilised while attempting to understand and to learn new statisticalmaterial, making it difficult to solve statistical problems. Moreover, a causal link between statistics anxiety and course achievement has been established. In particular, Onwuegbuzie and Seaman (1995) found that graduatestudents with high levels of statistics test anxiety who were randomly assigned toa statistics examination that was administered under timed conditions tended tohave lower levels of performance than did their high-anxious counterparts who wereadministered the same test under untimed conditions. In a follow-up experimentalinvestigation among female college students, Onwuegbuzie (1995) reported asignificant interaction between statistics test anxiety and type of examination (i.e.

Specifically, three types of factors are identified: (a) situational factors, such as math experience (Betz, 1978; Roberts & Bilderback, 1980; Tomazic & Katz, 1988; Zeidner, 1991; Wilson, 1997; Balo_lu, 2001; Hong & Karstensson, 2002; Balo_lu, 2003), statistics experience (Sutarso, 1992), computer experience (Zimmer & Fuller, 1996), and research experience (Trimarco, 1997); (b) dispositional factors, such as math self-concept or self-esteem (Zeidner, 1991), scholastic competence or multiple intelligences (Birenbaum & Eylath, 1994; Daley & Onwuegbuzie, 1997; Onwuegbuzie, 2000c), perfectionism (Onwuegbuzie & Daley, 1999; Walsh & Ugumba-Agwunobi, 2002), hope (Onwuegbuzie. 1998b), and procrastination (Onwuegbuzie, 2000a; Walsh & Ugumba-Agwunobi, 2002); and - personal factors, such as learning style (Onwuegbuzie, 1998a; Wilson & Onwuegbuzie, 2001), age (Baloglu, 2003), gender (Betz, 1978; Roberts & Bilderback, 1980; Demaria-Mitton, 1987; Benson, 1989; Benson & Bandalos, 1989; Hong & Karstensson, 2002; Baloglu, 2003), and ethnicity (Bell, 1998; Onwuegbuzie. 1999).

Unfortunately, sparse studies have been found on how to reduce the anxiety in learning statistics for graduate students in the social sciences (Onwuegbuzie & Wilson, 2000). Schacht and Stewart (1990) reported incorporating humorous cartoon examples in statistics classes were perceived by the students to helpful in statistics anxiety reduction. Journal writing was suggested to be another effective tool in reducing levels of anxiety (Smith, Miller, & Robertson, 1992; Sgoutas-Emch & Johnson, 1998). However, theses studies did not find a statistically significant decrease in anxiety levels. Forte (1995) argued for an effective teaching approach that incorporates computer usage, real-world applications, humor, statistical language practice, and group-learning principles. Additionally, Dolinsky (2001) suggested creating a collaborative environment in which using active learning strategies as the primary method to teach statistics. These teaching methods reveal promise to reduce students' statistics anxiety, yet more innovations of teaching statistics are needed (Huntley, Schneider, & Aronson, 2000). More importantly, empirical studies are needed to examine the effectiveness of these methods in reducing statistics anxiety.