Understanding the hidden magnetic forces between your questions and the underlying concepts they measure.
Factor loadings are essentially correlation coefficients between observed items (the specific questions you asked) and the underlying latent factors (the abstract concept).
Adjust the slider to see how "Loading Strength" affects the relationship.
Very weak connection.
| Item Content | Factor 1 (Engagement) |
Factor 2 (Confidence) |
|---|---|---|
| I enjoy online lessons | 0.72 | 0.12 |
| I feel anxious using EdTech | 0.10 | 0.65 |
| I find Zoom effective | 0.68 | 0.18 |
| I feel confident with LMS | 0.15 | 0.70 |
Items 1 & 3 load strongly here.
Items 2 & 4 load strongly here.
How strong is "strong"? Use these rule-of-thumb thresholds.
The item is a quintessential measure of the factor.
Solid relationship. Most researchers are happy with this.
Acceptable minimum is often 0.40. Anything lower is questionable.
Usually ignored or removed from the analysis.
When an item has high loadings (e.g., > 0.4) on two or more factors. It interprets ambiguously.
When an item doesn't reach 0.30 or 0.40 on any factor. It contributes nothing to the solution.
"A principal axis factoring with Promax rotation revealed a two-factor solution. Items loaded on factors based on a criterion of ≥ 0.40, with no significant cross-loadings. Factor 1 reflected 'Digital Engagement' and Factor 2 reflected 'Confidence with Technology.'"