A quantitative, correlational, survey design with anecdotal qualitative data was used to
investigate the relationships among mathematical anxiety, attitude toward learning math, gender,
ethnicity, and separate and connected ways of knowing within the context of the mathematics
classroom. Participants were 88 student volunteers enrolled in undergraduate mathematics
classes at an open admissions technical college in the southeastern United States. Survey data
consisted of demographic self-report items, Likert scale items, semantic differential scale items,
and one qualitative free-response question. Quantitative data were analyzed by use of either a
Spearman Rho, Pearson product moment correlation or an independent samples t-test of
significance. These data were supplemented by participants’ qualitative responses which were
categorized. The results indicated that there was a significant positive relationship between
attitudes toward math and separate knowing in that those who had more positive attitudes toward
math were more likely to be separate knowers. Results also indicated that gender is related to
one’s way of knowing in that connected knowing correlated strongly with the female gender, and
a significant difference existed between males and females with regard to connected knowing.
However, results indicated that males’ mean score on connected knowing was significantly
higher than their mean score on separate knowing. Furthermore, results indicated a significant
correlation between ethnicity and ways of knowing with historically underrepresented and
marginalized individuals more likely to be separate knowers. Finally, results indicated that the
mean scores for females differed significantly from those of males on two out of eight factors
related to mathematical ways of knowing as measured by the Mathematical Dialectics Measure
which was designed specifically for this study. The present findings indicate that relationships
do exist among attitudes, anxiety, gender, ethnicity and ways of knowing in mathematics. Since
this study was correlational, statements cannot be made about the causal effect of any of these
variables on one another. Further research should use an experimental or quasi-experimental
design to more thoroughly examine the impact of these variables on one another and on
mathematics achievement in particular.