Table 2 Mixed effects location scale model results for each cognitive task

From: Ecological momentary assessments of cognitive performance are more variable in breast cancer survivors

 

Processing speed

Working memory updating

Spatial working memory

 

Estimate

SE

CI Lower

CI Upper

PD

Estimate

SE

CI Lower

CI Upper

PD

Estimate

SE

CI Lower

CI Upper

PD

Comparison: intercept (alphaMu)

17.45

0.58

16.48

18.40

100%

3.81

0.20

3.49

4.13

100%

1.56

0.03

1.51

1.61

100%

Survivor: intercept (alphaMu)

21.89

0.73

20.45

23.30

100%

4.30

0.26

3.80

4.81

100%

1.49

0.05

1.39

1.58

100%

Difference: intercept (alphaMu)

4.44

0.81

3.10

5.77

100%

0.49

0.28

0.04

0.94

96%

−0.07

0.06

−0.16

0.02

90%

Comparison: observation

-

-

-

-

-

-

-

-

-

-

−0.02

0.00

−0.03

−0.01

100%

Survivor: observation

-

-

-

-

-

-

-

-

-

-

−0.04

0.01

−0.06

−0.03

100%

Difference: observation

-

-

-

-

-

-

-

-

-

-

−0.02

0.01

−0.04

−0.01

98%

Comparison: initial slope (time = 0)

2.49

0.43

1.74

3.42

100%

1.45

0.15

1.17

1.78

100%

-

-

-

-

-

Survivor: initial slope (time = 0)

2.04

0.82

0.90

4.04

100%

1.51

0.17

1.21

1.86

100%

-

-

-

-

-

Difference: initial slope (time = 0)

−0.45

0.92

−1.97

1.66

72%

0.06

0.23

−0.38

0.51

61%

-

-

-

-

-

Comparison: decay (gammaMu)

−1.22

0.22

−1.60

−0.88

100%

−1.41

0.12

−1.61

−1.22

100%

-

-

-

-

-

Survivor: decay (gammaMu)

−1.91

0.57

−3.11

−0.90

100%

−1.45

0.15

−1.77

−1.18

100%

-

-

-

-

-

Difference: decay (gammaMu)

−0.69

0.61

−1.75

0.26

87%

−0.04

0.19

−0.36

0.27

57%

-

-

-

-

-

Comparison: asymptote (betaMu)

25.83

1.18

24.07

27.90

100%

9.74

0.41

9.09

10.43

100%

-

-

-

-

-

Survivor: asymptote (betaMu)

35.21

4.15

29.42

45.39

100%

10.74

0.72

9.47

12.28

100%

-

-

-

-

-

Difference: asymptote (betaMu)

9.38

4.29

3.76

17.42

100%

0.99

0.80

−0.26

2.36

90%

-

-

-

-

-

Age

−0.11

0.06

−0.20

−0.01

97%

−0.06

0.02

−0.09

−0.03

100%

0.00

0.00

−0.01

0.00

93%

Education

1.24

0.66

0.15

2.32

97%

0.34

0.25

−0.07

0.74

92%

−0.03

0.02

−0.07

0.01

92%

Within-Person variance model

 Comparison: intercept

1.25

0.03

1.20

1.31

100%

0.17

0.04

0.11

0.23

100%

2.20

0.06

2.08

2.32

100%

 Survivor: intercept

1.42

0.04

1.33

1.51

100%

0.32

0.05

0.22

0.43

100%

2.31

0.11

2.10

2.52

100%

 Difference: Intercept

0.17

0.06

0.07

0.26

100%

0.15

0.06

0.05

0.26

99%

0.10

0.12

−0.14

0.35

80%

 Comparison: observation

0.04

0.01

0.03

0.05

100%

0.09

0.01

0.08

0.10

100%

−0.01

0.01

−0.04

0.02

71%

 Survivor: observation

0.03

0.01

0.01

0.04

100%

0.09

0.01

0.07

0.11

100%

−0.07

0.02

−0.11

−0.02

100%

 Difference: observation

−0.01

0.01

−0.03

0.01

83%

0.00

0.01

−0.02

0.02

52%

−0.06

0.03

−0.11

0.00

98%

Between-Person variance model

 Comparison: intercept (alphaMu)

4.58

0.41

3.94

5.28

100%

1.43

0.13

1.23

1.66

100%

0.27

0.02

0.23

0.31

100%

 Survivor: intercept (alphaMu)

3.68

0.51

2.92

4.59

100%

1.38

0.17

1.11

1.68

100%

0.28

0.04

0.22

0.35

100%

 Difference: intercept (alphaMu)

−0.89

0.65

−2.14

0.43

91%

−0.06

0.22

−0.47

0.39

62%

0.01

0.05

−0.07

0.11

61%

 Comparison: asymptote (betaMu)

7.04

0.72

6.00

8.32

100%

2.82

0.25

2.43

3.26

100%

-

-

-

-

-

 Survivor: asymptote (betaMu)

8.10

2.06

5.27

11.58

100%

3.13

0.40

2.54

3.85

100%

-

-

-

-

-

 Difference: asymptote (betaMu)

1.05

2.18

−3.27

5.71

71%

0.31

0.47

−0.56

1.30

74%

-

-

-

-

-

Hurdle model

 Comparison: intercept

-

-

-

-

-

-

-

-

-

-

0.05

0.01

0.03

0.07

100%

 Survivor: intercept

-

-

-

-

-

-

-

-

-

-

0.13

0.03

0.08

0.19

100%

 Difference: intercept

-

-

-

-

-

-

-

-

-

-

0.08

0.03

0.03

0.14

100%

  1. Results from mixed-effects location scale (MELS) models with the random scale parameter conducted using the R brms package. Estimates for observation (time) represent the change for every 10 observations. Higher scores on processing speed and working memory updating tasks indicate better performance. Lower scores on the spatial working memory task indicate better performance. All models include random intercepts and slopes in the mean and within-person variance models. In the processing speed and working memory updating models, covariates are only included at the asymptote (betaMu). The processing speed model is based on 9358 observations from 152 participants, the working memory updating model is based on 9312 observations from 152 participants, and the spatial working memory model is based on 9332 observations from 152 participants. Italicized estimates (within-person variance model for processing speed and working memory updating; mean model for spatial working memory) are on the log-linear scale. The hurdle model estimates the likelihood of a 0 value; the model estimates have been transformed from logit to probability to facilitate interpretation. Greater (i.e., more positive) values of gammaMu indicate a more rapid decline from the value of the intercept towards the asymptote. SE standard error. CI = 95% credibility interval. PD probability of direction.
Back to article page