TABLE 6.
Experiment 2 Model 3 results from the mixed effects model with proficiency and interactions between proficiency and the distribution factors.
|
Fixed effects |
Random effects |
||||||
| By subject4 | By item | ||||||
| Parameters | Estimate | SE | z | p | SD | SD | |
| Intercept | –6.366 | 1.939 | –3.28 | 0.001** | 0.82 | 1.00 | |
| Morpheme Type1 | Plural -s | –0.888 | 0.235 | –3.78 | 0.000*** | ||
| Third-person -s | 0.746 | 0.231 | 3.23 | 0.001** | |||
| Progressive -ing | 1.242 | 0.242 | 5.14 | 0.000*** | |||
| Morpheme reliability | 3.946 | 1.142 | 3.46 | 0.001*** | |||
| Morpheme availability2 | 0.591 | 0.433 | 1.36 | 0.172 | |||
| Formulaicity (Low)3 | –1.437 | 0.614 | –2.34 | 0.019* | |||
| Proficiency | 1.234 | 0.378 | 3.26 | 0.001** | |||
| Proficiency*Morpheme reliability | –0.455 | 0.220 | –2.06 | 0.039* | |||
| Proficiency*Morpheme availability | –0.050 | 0.085 | –0.58 | 0.559 | |||
| Proficiency*Formulaicity | 0.165 | 0.120 | 1.38 | 0.168 | |||
1Past-tense -ed is the reference level. 2Word-form frequency was logarithmically transformed. 3High formulaicity was the reference level. Model formula: accuracy ∼ morpheme + proficiency ∗ (reliability + availability + formulaicity) + (1| subject) + (1| item). 4Random slopes by subject are not included in this model because adding them did not significantly improve model fit [χ2(df = 5) = 10.30, p = 0.07]. ***p = 0; **p < 0.001; *p < 0.01; p < 0.1.