Confirmatory Factor Analysis of Two Versions of the Statistical Anxiety Scale in a Sample of American Students
Saved in:
| Title: | Confirmatory Factor Analysis of Two Versions of the Statistical Anxiety Scale in a Sample of American Students |
|---|---|
| Language: | English |
| Authors: | Keston G. Lindsay (ORCID |
| Source: | International Journal of Research in Education and Science. 2026 12(1):105-115. |
| Availability: | International Society for Technology, Education, and Science. e-mail: ijresoffice@gmail.com; Web site: https://www.ijres.net/index.php/ijres |
| Peer Reviewed: | Y |
| Page Count: | 11 |
| Publication Date: | 2026 |
| Document Type: | Journal Articles Reports - Research |
| Education Level: | Higher Education Postsecondary Education |
| Descriptors: | Undergraduate Students, Anxiety, Statistics, Affective Measures, Factor Structure, Social Desirability, Test Items, Goodness of Fit, Construct Validity, Test Reliability |
| Geographic Terms: | Illinois, Texas, Colorado |
| ISSN: | 2148-9955 |
| Abstract: | The objective of this study was to confirm the factor structure for two versions of the Statistics Anxiety Scale (SAS) created for American undergraduate statistics students. To address the inconsistent factor structure of SAS across different populations, Lorenzo-Seva et al. (2022) revised the SAS. This version reduced the number of questions and added a 4th factor called social desirability. They reported a good fit without the need for introducing correlations between error terms, which were outcomes found in different versions of the SAS including this study. This revised SAS considers a 4th dimension called social desirability, suggesting that this dimension could account for much of the error variance found in previous studies. That dimension was not included in this study. A modified three factor model explained construct validity in the original version of the SAS. A modified five factor model explained construct validity in a version of the SAS with additional items. Both versions of the SAS and their factors displayed acceptable levels of fit, were found to have reliability coefficients above 0.70, and generally shared moderate negative relationships with Wise's Attitude Toward Statistics scale. |
| Abstractor: | As Provided |
| Entry Date: | 2026 |
| Accession Number: | EJ1494346 |
| Database: | ERIC |
| Abstract: | The objective of this study was to confirm the factor structure for two versions of the Statistics Anxiety Scale (SAS) created for American undergraduate statistics students. To address the inconsistent factor structure of SAS across different populations, Lorenzo-Seva et al. (2022) revised the SAS. This version reduced the number of questions and added a 4th factor called social desirability. They reported a good fit without the need for introducing correlations between error terms, which were outcomes found in different versions of the SAS including this study. This revised SAS considers a 4th dimension called social desirability, suggesting that this dimension could account for much of the error variance found in previous studies. That dimension was not included in this study. A modified three factor model explained construct validity in the original version of the SAS. A modified five factor model explained construct validity in a version of the SAS with additional items. Both versions of the SAS and their factors displayed acceptable levels of fit, were found to have reliability coefficients above 0.70, and generally shared moderate negative relationships with Wise's Attitude Toward Statistics scale. |
|---|---|
| ISSN: | 2148-9955 |
