The author brings up the gap in sociolinguistic research regarding linguistic data and social practice that is responsible for the data. Eckert claims that gender does not have a uniform effect on linguistic variables because it is reflective of complex social practice based on sex. She asserts that the correlation of sex with linguistic variables is due to the effects of gender on an individual"s behavior. She claims that we must find explanations for these effects by looking at social constructions. She studied adolescents due to the fact that they are the age group leading linguistic change, with their life stages changing relation to the linguistic market. Additionally, in the time when the article was written, the social constructions of sex were changing, as women were calling traditional gender roles into question, making these gender differences even more complex to study. She assets that gender differentiation is greatest when power is the scarcest.