Information Technology Reference
In-Depth Information
Online survey . For the second part of the data collection, on-line surveys related
to academic motivation and the use of learning strategies were completed from the
students' own computer or in computer labs in the college.
Data Analysis
The first part of the analysis focused on evaluating the psychometric characteris-
tics of the measures. We examined internal consistency of scales by calculating
Cronbach's alpha coefficients. Also, in order to evaluate responses to the same
instrument by different groups (in this case Latino and White students) separately,
the second part of analysis examines whether the constructs were the same across
White and Latino/a groups. Therefore, multiple group confirmatory factor analysis
(MG-CFA) was employed to test factorial invariance between those groups.
MG-CFA test traditionally posits a series of nested models in which the end point
is the least restrictive model with no invariance constraints and the most restric-
tive (total invariance) model with all parameters constrained to be equal across all
groups. Thus, we first examined the extent to which the theoretical structure of the
instrument was invariant to ensure the structural equivalence of underlying con-
structs between White and Latino/a groups. We also investigated the extent to which
the factor patterns and weighing of loadings were invariant, in order to ensure the
equivalent interpretation of item content across two different groups. Finally, we
aimed to determine how demographic and background information related to moti-
vation/learning strategies and academic achievement for White and Latino/a college
students. We initially developed a model for the entire sample ( n
3,257) to test
the theoretical SEM and then we estimated two single-group SEMs using White stu-
dents and Latino students to compare the significant path coefficients between two
groups.
=
Results
Reliabilities, Descriptive Statistics, and Intercorrelations
Reliabilities . Cronbach's alpha reliability coefficients were computed for nine scales
on motivation/learning strategies items. Scale reliabilities for all scales were gener-
ally satisfactory (see Table 7.1). For the total sample of 3,527 participants, reliability
coefficients varied from 0.71 to 0.93 ( M
0.84). This average reliability and the
reliabilities of each of the nine scales are higher than at least 0.70, suggesting
that the measures used in this study appear to meet basic psychometric criteria.
Descriptive analyses were conducted to provide overall information of the study
variables. Table 7.2 presents means and standard deviations of primary variables.
=
Search WWH ::




Custom Search