. 2016 Nov 5;13(11):1092. doi: 10.3390/ijerph13111092

Table 3.

Age relations in different domains applying a gradual versus applying a stage model.

Variable	Gradual Model			Stage Model
Variable	b_linear	b_quadratic	R²	Age Break	b_linear	b_agegroup	b_interaction	R²	ΔR²
Processing speed	2.35 ^ns	−0.02 *	0.14 ***	90	−3.13 ***	22.83 ***	1.76 *	0.14 ***	< 0.001 ^ns
Verbal abilities	−0.15 ^ns	−0.003 ^ns	0.03 ***	87	0.38 ^ns	9.99 ***	−0.88 *	0.03 ***	0.003 ** (8.1%)
General health status	−1.22 ^ns	0.005 ^ns	0.03 ***	87	0.02 ^ns	6.07 ***	−0.53 *	0.03 ***	0.001 * (4.0%)
Activities	0.36 **	−0.004 ***	0.29 ***	80	−0.27 ***	3.29 ***	0.12 ***	0.29 ***	0.001 * (0.4%)
Life satisfaction	−0.06 ^ns	0.0004 ^ns	0.0006 ^ns	90	−0.05 *	−0.11 ^ns	0.05 ^ns	0.003 ^ns	0.002 * (312.2%)

Age relations in cognitive abilities, general health status, activities, and life satisfaction. Left panel: Regression findings applying a gradual model, with scores in the respective variable regressed on chronological age (testing for linear and quadratic age terms). b_linear = regression coefficient of the linear age term. b_quadratic = regression coefficient of the quadratic age term. R² = multiple R², resulting total variance that is accounted for by the linear and the quadratic age term. Right panel: Regression findings applying a stage model, with scores in the respective variable regressed on chronological age (testing for a linear age term) plus an age group factor (young-old versus old-old) plus an interaction term of the linear age term and the age group factor. Age break = the age separating young-old from old-old age. b_linear = regression coefficient of the linear age term. b_agegroup = regression coefficient of the age group factor. b_interaction = regression coefficient of the interaction term. R² = multiple R², resulting total variance that is accounted for by the linear age term, the age group factor, and the interaction term. ΔR² represents the amount of increase in explained variance in the stage model compared to the gradual model (in parentheses, the percentage regarding the increase relative to the initial amount of explained variance by the gradual model is given). Higher values represented better performance across all variables. *** p < 0.001; ** p < 0.01; * p < 0.05; ^ns non-significant, p > 0.05; two-tailed.