Table 3 Multilevel model with fixed and random effects for delayed recall

From: Profiles of social isolation and loneliness as moderators of the longitudinal association between uncorrected hearing impairment and cognitive aging

 

Delayed recall

SE

t (df)

CI

p

B (β)

Fixed effects

 (Intercept)

6.65 (0.04)

0.08

87.03 (35,770)

6.52 to 6.82

<0.001

Retest effects

 Wave 2

0.16 (0.02)

0.02

7.50 (111,000)

0.12 to 0.20

<0.001

 Wave 4

0.36 (0.04)

0.02

19.09 (114,300)

0.33 to 0.40

<0.001

 Wave 5

0.34 (0.05)

0.02

20.54 (118,800)

0.30 to 0.37

<0.001

 Wave 6

0.35 (0.07)

0.02

20.88 (124,000)

0.32 to 0.38

<0.001

 Wave 7

0.29 (0.03)

0.03

10.95 (129,200)

0.23 to 0.34

<0.001

 Wave 8

0.26 (0.04)

0.03

10.00 (135,000)

0.21 to 0.31

<0.001

 Wave 9

0.26 (0.04)

0.03

8.86 (135,700)

0.21 to 0.32

<0.001

Between-subjects’ effects

 Age (M)

−0.06 (−0.23)

0.001

−60.68 (34,700)

−0.06 to −0.06

<0.001

 Sex (Female respondents)

0.44 (0.10)

0.02

27.75 (32,900)

0.41 to 0.47

<0.001

 Education

0.39 (0.27)

0.01

69.58 (32,890)

0.38 to 0.41

<0.001

 Chronic conditions (M)

−0.07 (−0.04)

0.01

−11.22 (33,950)

−0.09 to −0.06

<0.001

 Hearing impairment (M)

−0.17 (−0.06)

0.01

−15.74 (33,860)

−0.19 to −0.15

<0.001

Profiles of social isolation and loneliness (ref: non-isolated and low loneliness)

 Non-isolated and high loneliness

−0.23 (−0.11)

0.02

−14.12 (32,940)

−0.26 to −0.20

<0.001

 Isolated and low loneliness

−0.24 (−0.11)

0.08

−3.07 (34,250)

−0.39 to −0.09

0.002

 Isolated and high loneliness

−0.45 (−0.21)

0.06

−6.97 (33,950)

−0.57 to −0.32

<0.001

Within-subjects’ effects

 Age (C)

−0.04 (−0.07)

0.002

−15.92 (134,200)

−0.04 to −0.03

<0.001

 Chronic conditions (C)

−0.01 (−0.01)

0.01

−2.52 (103,800)

−0.02 to −0.001

0.012

 Hearing impairment (C)

−0.05 (−0.01)

0.01

−5.19 (16,880)

−0.07 to −0.03

<0.001

 Hearing impairment (C2)

−0.06 (−0.02)

0.01

−7.39 (4863)

−0.07 to −0.04

<0.001

Interaction effects

 Hearing impairment (C) * Non-isolated and high loneliness

−0.04 (−0.01)

0.01

−3.03 (16,080)

−0.07 to −0.01

0.002

 Hearing impairment (C) * Isolated and low loneliness

0.02 (0.01)

0.07

0.26 (16,300)

−0.12 to 0.15

0.797

 Hearing impairment (C) * Isolated and high loneliness

−0.08 (−0.02)

0.06

−1.42 (15,810)

−0.19 to 0.03

0.156

 

Estimates

SD

CI

Random effects

 Residual variance

2.01 (1.42)

1.42

1.20 to 2.43

 Intercept (variance)

1.48 (1.22)

1.22

1.20 to 1.23

 Hearing Impairment (C) slope (variance)

0.12 (0.34)

0.34

0.31 to 0.37

 Hearing Impairment (C2) slope (variance)

0.02 (0.14)

0.14

0.09 to 0.19

 Intercept*Hearing impairment (C) slope (covariance)

0.13

–

0.08 to 0.19

 Intercept*Hearing impairment (C2) slope (covariance)

−0.34

–

−0.68 to −0.09

 Hearing impairment (C) slope* Hearing impairment (C2) slope (covariance)

−0.32

–

−0.51 to −0.22

ICC

0.43

N

33,725

Observations

137,039

Marginal R2/Conditional R2

0.207/0.545

  1. The CIs in the random effects’ variances correspond to their Standard Deviations, while in the random effects’ covariances they correspond to the actual estimates.
  2. M person-mean variable (between-subjects differences), C person-mean centred variable indicating the linear change, C2 person-mean centred variable indicating the quadratic change, SD standard deviation, CI 95% confidence intervals, Marginal R2 variance explained by fixed effects only, Conditional R2 variance explained by fixed and random effects.
Back to article page