Skip to main content
. 2021 Jul 8;11:14163. doi: 10.1038/s41598-021-93652-4

Table 1.

Results from model fitting based on Akaike Information Criterion (AIC).

Model K AIC ΔAIC wi
Outcome = hours spent helping parents
 Covariates 27 4534.02 0.90 0.39
 Fecundity status + covariates 29 4533.12 0.00 0.61
Outcome = hours spent caring for grandchildren
 Covariates 27 7184.35 6.59 0.04
 Fecundity status + covariates 29 717,776 0.00 0.96

The lowest AIC value is deemed to best fit the data, with ΔAIC referring to the difference in AIC value from the best fitting model (shown in italics). A ΔAIC value of more than two demonstrates a significantly poorer model fit, and wi indicates model probability. Covariates include number of grandchildren, age the woman left education, employment status, marital status, and health. Where the outcome is hours spent helping parents, hours spent caring for grandchildren is included in the covariates, and vice versa.