achievement, and .30, p<.01 for standardized tests when no sophisticated control
variables were included in the statistical procedure. When sophisticated controls were
included in the procedure, the coefficients for the effect of communication on overall
academic achievement dropped to .15, and the coefficient for the effect of
communication on standardized tests dropped to .14. Controls for SES, race, gender, or
previous achievement reduced these coefficients by half which is noteworthy, but it is
also important to realize the reduced coefficients were not statistically significant, and
therefore communication likely has no effect on academic achievement (p.95).
Socioeconomic Status — SES
In the Jeynes (2005) study being replicated here, SES is made up of five different
variables including, mother’s education, father’s education, father’s occupation, mother’s
occupation, and family income (p.105). Even though this particular construct of SES is
relatively common, it is important to develop a basic appreciation of the numerous factors
that are often overlooked in many evaluations of SES. A brief discussion of a few
examples of the less common factors that make up SES in other studies will also
illuminate the difficulties which arise in making comparisons between studies with
heterogeneous variable constructs.
Two other studies, Crane (1996) and Attewell and Battle (1999) operationalized a
longer list of variables to comprise SES, but there were several common components.
Crane conducted an evaluation of the relationship between students’ home environment,
SES, maternal test scores and academic achievement as measured by mathematics scores.