Skip to main content
International Journal of Child Care and Education Policy

Table 5 Income by birth cohort interaction models: linear and non-linear specifications

From: Do increased availability and reduced cost of early childhood care and education narrow social inequality gaps in utilization? Evidence from Norway

Predictor

OR

95% CI

Income interaction model 1

  

 Linear income levels

1.14***

(1.07 to 1.22)

 Birth cohort

1.27***

(1.23 to 1.31)

 Income × cohort

.97

(.94 to 1.01)

Income Interaction Model 2

  

 Log of income levels

1.79***

(1.37 to 2.34)

 Birth Cohort

1.27***

(1.23 to 1.31)

 Log of income × cohort

.93

(.81 to 1.06)

Income interaction model 3

  

 1st to 25th income percentile

.64***

(.54 to .76)

 26th to 50th income percentile

.78***

(.69 to .89)

 51st to 75th income percentile

.93

(.84 to 1.04)

 Birth cohort

1.25***

(1.20 to 1.30)

 1st to 25th percentile × cohort

1.03

(.96 to 1.10)

 26th to 50th percentile × cohort

1.05†

(1.00 to 1.11)

 51st to 75th percentile× cohort

1.01

(.97 to 1.06)

  1. All covariates displayed in Table 1 and linear parent education effects were controlled in these models. For interaction terms, significant odds ratios less than 1 indicate a narrowing selection gap between families with relatively high versus low levels of the covariate in question. † p ≤ .10; ***p < .001.
Back to article page