ERIC Number: ED312299
Record Type: Non-Journal
Publication Date: 1989-Nov
Pages: 36
Abstractor: N/A
ISBN: N/A
ISSN: N/A
EISSN: N/A
Second-Order Factor Analysis as a Validity Assessment Tool: A Case Study Example Involving Perceptions of Stereotypic Love.
Borrello, Gloria M.; Thompson, Bruce
The calculation of second-order results in the validity assessment of measures and some useful interpretation aids are presented. First-order and second-order results give different and informative pictures of data dynamics. Several aspects of good practice in interpretation of second-order results are presented using data from 487 subjects responding to the Love Relationships Scale (LRS) of B. Thompson and G. M. Borrello (1987, 1989). Fifty-one of the subjects participated in this study, and the rest of the subjects participated in previous studies of the LRS. It is suggested that interpreting only the rotated second-order factor matrix is not good practice. It is also suggested that interpretation can be aided by consulting both the unrotated and the rotated S(VxS) product matrix obtained by post-multiplying the first-order factor pattern matrix by the second-order factor pattern matrix. It is further suggested that orthogonalizing the first-order factors using the algorithm of J. Schmid and J. Leiman (1957) is helpful in interpreting second-order results. Seven tables present product matrices and listings of items from product matrices, one figure illustrates culturally-defined stereotypic love, and the LRS is appended. (Author/SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers; Tests/Questionnaires
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Grant or Contract Numbers: N/A