Factor analyses were conducted using SPSS principal components analysis to

Factor analyses were conducted using spss principal

This preview shows page 344 - 346 out of 367 pages.

Factor analyses were conducted using SPSS principal components analysis to evaluate the underlying structure of the TTIS for two purposes: 1) determine if the proposed subscales are valid and reliable and 2) identify items that are not adequately contributing to the subscales. All factor analyses utilized the orthogonal varimax rotation that minimizes fac- tor complexity by maximizing factor variance. Factor Analysis Results An exploratory factor analysis of all items generated twelve components; however an evaluation of the scree plot, eigenvalues, and communalities indicated that eight components should be retained. These eight components or factors accounted for 67% of the original variance and confirmed the general structure of the six proposed subscales with the addition of two factors that further delineated teacher and student use. The change in subscales suggested by some of the panelists was not supported. Since items among similar subscales often overlapped in the exploratory analysis, three confirmatory factor analyses were conducted. Factors and item loadings were assessed to evaluate appropriate fit of items and factor reliability. Hair (1998) indicates that loadings greater than .6 are “high” while loadings less than .4 are low and would call into question the item placement. Stevens (1992) suggests that components with four or more loadings above .6 are reliable. After factor analyses were complete, internal reliability was evaluated by calculating Cronbach’s alpha for each of the generated factors. The factor analysis examining items that measured Risk-taking , Benefits, and Beliefs and Behaviors (items 1-20) cre- ated three factors that accounted for 71% of the original variance (see Table 1). The first factor, Risk-taking, confirmed items 1-9 as a subscale with all but one item loading above .6. Top items loaded included: 2) Learning new technolo- gies is confusing to me (reversed); 3) I get anxious when using new technologies because I don’t know what to do if something goes wrong (reversed). Reliability for this factor was quite high ( α = .850). The second factor generated was Benefits . Item 11 (Computer technology allows me to create materials that enhance my teaching) had the highest loading. Internal reliability for this factor was calculated at α = .849. The third factor generated was Beliefs and Behaviors . High- est loading items were: 16) Using technology in the classroom is a priority for me; 19) I regularly plan learning activities/ lessons in which students use technology. Each of these factors had multiple loadings greater than .6 and no loadings less than .4. In addition, internal reliability coefficients were greater than .8. The author concluded that these three factors were appropriate and reliable. When combining all 20 items to create a sub-scale of Risk-taking, Benefits, and Beliefs (items 1-20), the researcher found an internal reliability of .916.
Image of page 344
332 Vannatta and Banister Table 1 Factor Analysis Results for Risk-taking, Benefits, Beliefs & Behaviors Items M SD Loading Risk-taking (α = .854) 2 Learning new technologies is confusing for me. (reversed)
Image of page 345
Image of page 346

You've reached the end of your free preview.

Want to read all 367 pages?

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture