Table 1
Rotated Component Matrix (extraction method: principal component analysis; rotation method: promax).
| COMPONENTS | ||
|---|---|---|
| AUTOMATIZATION | READING ACCURACY | |
| Word reading speed | 0.829 | 0.114 |
| Word reading accuracy | –0.017 | 0.902 |
| Pseudoword reading speed | 0.740 | 0.097 |
| Pseudoword reading accuracy | –0.050 | 0.918 |
| WM | 0.479 | 0.377 |
| RAN colors | –0.950 | 0.073 |
| RAN objects | –0.959 | 0.181 |
[i] WM = Working memory; RAN = Rapid automatized naming.
Table 2
Results of the best-fitting model resulting from the model selection procedure.
| FIXED EFFECT | F-VALUE | NumDF, DenDF | p-VALUE | b |
|---|---|---|---|---|
| OSC | 17.55 | 1,94 | < .001 | –.19 |
| word frequency | 30.73 | 1,94 | < .001 | –.02 |
| length | 75.43 | 1,94 | < .001 | .04 |
| automatization | 1.40 | 1,137 | .23 | –.03 |
| reading accuracy | 2.03 | 1,133 | .15 | –.03 |
| automatization : length | 85.88 | 1,10049 | < .001 | –.01 |
| reading accuracy : word frequency | 8.44 | 1,10049 | .003 | .003 |
Figure 1
Results of the best-fitting model on participants’ RTs. RTs decreased as a function of OSC (A); the interactions automatization by length and reading accuracy by word frequency indicated that expert readers (in warmer colors) are less sensitive to the length effect (B) and to the word frequency effect (C).
Figure 2
Contour plots of the interactions automatization by length (A) and reading accuracy by frequency (B). The individual-level predictor is on the Y-axis, the word-level predictor is on the X-axis, while the timing of the RTs is indicated by the color populating both plots, with warmer colors indicating slower RTs. The lower sensitivity to length and frequency effects showed by more expert readers can be understood by comparing the colors spectrum from the upper (low variability) to the lower (high variability) parts of the plots.