Table 6. Relation of spectral-pattern discrimination and speech-in-noise ability to working memory.
| Model | Dependent Variable | Estimate | SE | 95% CI (lower/upper) | β | p value | Squared Semi-Partial Correlation | Squared Structure Coefficient |
|---|---|---|---|---|---|---|---|---|
| 1 | Ripple phase | -.183 | .064 | -.309/-.057 | -.247 | .005 | .051 | .492 |
| 2 | FM SNR | -2.063 | .724 | -3.496/-.629 | -.256 | .005 | .055 | .530 |
| 3 | QuickSIN | -.741 | .468 | -1.668/.185 | -.139 | .116 | .016 | .011 |
From three separate linear regression models that controlled for age, gender, race, and education, table entries are the estimated coefficient, standard error (SE), the lower and upper 95% confidence interval (CI) for the estimate, standardized coefficient (β), p value, squared semi-partial correlation, and squared structure coefficient for the independent variable working memory. For model 1, F(5,123) = 8.88, p < .001; model 2, F(5,123) = 6.18, p < .001; and model 3, F(6,123) = 7.90, p < .001. R2 (adjusted R2) was .273 (.243), .208 (.174), and .251 (.219) for models 1, 2, and 3, respectively.