ERIC_NO:
ED314439
TITLE:
Stability of Q-Factors across Two Data Collection Methods.
AUTHOR:
Daniel, Larry G.
PUBLICATION_DATE:
1989
ABSTRACT:
The purpose of the present study was to determine how two different data collection techniques would affect the Q-factors derived from several
factor analytic procedures. Faculty members (N=146) from seven middle
schools responded to 61 items taken from an instrument designed to measure
aspects of an idealized middle school culture; the instrument is the Middle
School Description Survey (MSDS). In three of the schools, the data were
collected using conventional Q-sort procedures. In the remaining four
schools, data were collected using an unmarked graphic scale. Separate
principal components Q-technique factor analyses were used with the data
collected in each school, and results were rotated to the varimax criterion.
For each of the seven Q-technique analyses, factor stability was assessed
by consulting the magnitude of the eigenvalue of the first factor extracted,
the mean eigenvalue of all extracted factors, the average squared communality
value, and the mean factor structure coefficient. Results indicate that
the data collected via the unmarked graphic scale produced more stable
and reliable Q-factors than did data collected via traditional Q-sort
procedures. The 61 items included in the MSDS are listed. (Author/TJH)
MAJOR_DESCRIPTORS:
Data Collection; Factor Analysis; Reliability; Research Methodology;
MINOR DESCRIPTORS:
Comparative Analysis; Junior High Schools; Methods Research; Middle Schools; Rating Scales; Secondary School Teachers;
IDENTIFIERS:
Middle School Description Survey; Principal Components Analysis; *Q Statistic; Varimax Procedures
PUBLICATION_TYPE:
143; 150