Emodiversity - As U Live Data
2017-06-02
Supplementary Analyses
The emodiversity construct was put forth by Quoidbach and colleagues (2014) describing the variety and relative abundance of individuals’ discrete emotion experiences. The original study used two samples and asked individuals to provide a single occasion rating of a number of discrete emotions. Results indicated that global, positive, and negative emodiversity were related to a variety of psychological and physical health items after accounting for mean levels of positive and negative emotions. Building on this initial work, Benson and colleagues (2017) published a methodological investigation of emodiversity using 30 days of diary data. Most recently, Brown & Coyne (2017) published a critique of the intial emodiversity paper, including critiques involving hierarchical regression and suppression. The following are a set of supplementary analyses utilizing the same As U Live data (Sturgeon, Zautra, & Okun, 2014) presented in the paper by Benson and colleagues (2017). Individuals reading these supplementary analyses are also encouraged to read the original articles on emodiversity.
References
Quoidbach, J., Gruber, J., Mikolajczak, M., Kogan, A., Kotsou, I., & Norton, M. I. (2014). Emodiversity and the emotional ecosystem. Journal of experimental psychology: General, 143(6), 2057. DOI: doi:10.1037/a0038025
Benson, L., Ram, N., Almeida, D. M., Zautra, A. J., & Ong, A. D. (2017). Fusing Biodiversity Metrics into Investigations of Daily Life: Illustrations and Recommendations With Emodiversity. The Journals of Gerontology: Series B. DOI: https://doi.org/10.1093/geronb/gbx025
Brown, N. J. L., & Coyne, J. C. (2017). Emodiversity: Robust Predictor of Outcomes or Statistical Artifact? Journal of Experimental Psychology. General. DOI: 10.1037/xge0000330
Summary of Supplementary Analyses
First, emodiversity calculated using the diary emotion data was examined (utilizing just the Gini coefficient):
- model 1 hierarchical series investigating (a) mean positive emotion, (b) mean positive emotion and positive emodiversity, and (c) mean positive emotion, positive emodiversity, and the interaction. Results suggested no suppression was occuring. Model comparisons suggested that model1a with just mean positive emotion provided the best fit to these data.
- model 2 hierarchical series investigating (a) mean negative emotion, (b) mean negative emotion and negative emodiversity, and (c) mean negative emotion, negative emodiversity, and the interaction. Results suggested suppression was occuring in terms of the magnitude of the beta coefficient increasing for mean negative emotion and the sign flipping for negative emodiversity. Model comparisons suggested that model2c with all three terms provided the best fit to these data.
- model 3 hierarchical series investigating (a) mean positive and mean negative emotion, (b) mean positive emotion, mean negative emotion, positive emodiversity, and negative emodiversity, and (c) mean positive emotion, mean negative emotion, positive emodiversity, and negative emodiversity, and the interactions. Results suggested suppression was occuring in terms of the magnitude of the beta coefficient increasing for mean negative emotion and the sign flipping for negative emodiversity. Model comparisons suggested that model3c including all terms provided the best fit to these data.
Second, emodiversity calculated using the single occasion emotion data was examined (utilizing both the Gini coefficient and Shannon’s entropy):
- model 4 hierarchical series utilizing the Gini coefficient, investigating (a) mean positive emotion, (b) mean positive emotion and positive emodiversity, and (c) mean positive emotion, positive emodiversity, and the interaction. Results suggested no suppression was occuring. Model comparisons suggested that model4b with mean positive emotion and positive emodiversity provided the best fit to these data.
- model 4 hierarchical series utilizing Shannon’s entropy, investigating (d) mean positive emotion, (e) mean positive emotion and positive emodiversity, and (f) mean positive emotion, positive emodiversity, and the interaction. Results suggested no suppression was occuring. Model comparisons suggested that model4e with mean positive emotion and positive emodiversity provided the best fit to these data.
- model 5 hierarchical series utilizing the Gini coefficient, investigating (a) mean negative emotion, (b) mean negative emotion and negative emodiversity, and (c) mean negative emotion, negative emodiversity, and the interaction. Results suggested suppression was occuring in terms of the magnitude of the beta coefficient increasing for mean negative emotion and the sign flipping for negative emodiversity. Model comparisons suggested that model5a with only mean negative emotion, provided the best fit to these data.
- model 5 hierarchical series utilizing Shannon’s entropy, investigating (d) mean negative emotion, (e) mean negative emotion and negative emodiversity, and (f) mean negative emotion, negative emodiversity, and the interaction. Results suggested suppression was occuring in terms of the magnitude of the beta coefficient increasing for mean negative emotion and the sign flipping for negative emodiversity. Model comparisons suggested that model5d with only mean negative emotion, provided the best fit to these data.
- model 6 hierarchical series utilizing the Gini coefficient, investigating (a) mean positive and mean negative emotion, (b) mean positive emotion, mean negative emotion, positive emodiversity, and negative emodiversity, and (c) mean positive emotion, mean negative emotion, positive emodiversity, and negative emodiversity, and the interactions. Results suggested suppression was occuring in terms of the magnitude of the beta coefficient increasing for mean negative emotion and the sign flipping for negative emodiversity. Model comparisons suggested that model6a including only mean positive and mean negative emotion, provided the best fit to these data.
- model 6 hierarchical series utilizing Shannon’s entropy, investigating (d) mean positive and mean negative emotion, (e) mean positive emotion, mean negative emotion, positive emodiversity, and negative emodiversity, and (f) mean positive emotion, mean negative emotion, positive emodiversity, and negative emodiversity, and the interactions. Results suggested suppression was occuring in terms of the magnitude of the beta coefficient increasing for mean negative emotion and the sign flipping for negative emodiversity. Model comparisons suggested that model6d including only mean positive and mean negative emotion, provided the best fit to these data.
Set-up
Read-in Data
zautra_full <- read.csv("6_EmoDiversity_Emotion and Biomarker & Followup Data_Zautra_2016-12-09.csv",header=TRUE)Load packages:
library(psych)
library(ggplot2)
library(lavaan)
library(foreign)
library(car)
library(grid)
library(gridExtra)
library(RColorBrewer)
library(ineq)
library(reshape)Subset of individuals who have 6 or more days of data (for diary analyses)
zautra <- zautra_full[which(zautra_full$days >= 6),] Subsets for the single occasion data analyses
Reduce dataset so same number of observations used in all models (for model fit comparison purposes later on) - this is for single occasion data using just the positive emotion items:
zautra2 <- zautra_full[which(!is.na(zautra_full$panaspos_c_mean_4wk) &
!is.na(zautra_full$panaspos_c_gini_4wk)),]Reduce dataset so same number of observations used in all models (for model fit comparison purposes later on) - this is for single occasion data using just the negative emotion items:
zautra3 <- zautra_full[which(!is.na(zautra_full$panasneg_c_mean_4wk) &
!is.na(zautra_full$panasneg_c_gini_4wk)),]Reduce dataset so same number of observations used in all models (for model fit comparison purposes later on) - this is for single occasion data using both positive and negative emotion items:
zautra4 <- zautra_full[which(!is.na(zautra_full$panaspos_c_mean_4wk) &
!is.na(zautra_full$panaspos_c_gini_4wk) &
!is.na(zautra_full$panasneg_c_mean_4wk) &
!is.na(zautra_full$panasneg_c_gini_4wk)),]Regression Models
Please see Benson, Ram, Almeida, Zautra, & Ong (2017) for further details on the individual emotion items used to calculated positive and negative emodiversity, mean positive and negative emotion, and physical health.
Center Variables
# Center diary data variables
zautra$pos_b_gini.c <- scale(zautra$pos_b_gini,center=TRUE, scale=FALSE)
zautra$neg_b_gini.c <- scale(zautra$neg_b_gini,center=TRUE, scale=FALSE)
zautra$pos_c_mean.c <- scale(zautra$pos_c_mean,center=TRUE, scale=FALSE)
zautra$neg_c_mean.c <- scale(zautra$neg_c_mean,center=TRUE, scale=FALSE)
# Center single occasion data variables - participant reflecting over past 4 weeks
zautra2$panaspos_c_gini_4wk.c <- scale(zautra2$panaspos_c_gini_4wk,center=TRUE, scale=FALSE)
zautra4$panaspos_c_gini_4wk.c <- scale(zautra4$panaspos_c_gini_4wk,center=TRUE, scale=FALSE)
zautra3$panasneg_c_gini_4wk.c <- scale(zautra3$panasneg_c_gini_4wk,center=TRUE, scale=FALSE)
zautra4$panasneg_c_gini_4wk.c <- scale(zautra4$panasneg_c_gini_4wk,center=TRUE, scale=FALSE)
zautra2$panaspos_c_shannon_4wk.c <- scale(zautra2$panaspos_c_shannon_4wk,center=TRUE, scale=FALSE)
zautra4$panaspos_c_shannon_4wk.c <- scale(zautra4$panaspos_c_shannon_4wk,center=TRUE, scale=FALSE)
zautra3$panasneg_c_shannon_4wk.c <- scale(zautra3$panasneg_c_shannon_4wk,center=TRUE, scale=FALSE)
zautra4$panasneg_c_shannon_4wk.c <- scale(zautra4$panasneg_c_shannon_4wk,center=TRUE, scale=FALSE)
zautra2$panaspos_c_mean_4wk.c <- scale(zautra2$panaspos_c_mean_4wk,center=TRUE, scale=FALSE)
zautra4$panaspos_c_mean_4wk.c <- scale(zautra4$panaspos_c_mean_4wk,center=TRUE, scale=FALSE)
zautra3$panasneg_c_mean_4wk.c <- scale(zautra3$panasneg_c_mean_4wk,center=TRUE, scale=FALSE)
zautra4$panasneg_c_mean_4wk.c <- scale(zautra4$panasneg_c_mean_4wk,center=TRUE, scale=FALSE)Mean Positive Emotion and Positive Emodiversity Predicting Physical Health - Diary Data
Descritpives:
describe(zautra[,c("physhealth","pos_c_mean","pos_b_gini")])## vars n mean sd median trimmed mad min max range
## physhealth 1 141 77.74 21.33 84.44 81.02 16.40 12.06 100.00 87.94
## pos_c_mean 2 175 2.09 0.77 2.10 2.09 0.69 0.20 3.96 3.76
## pos_b_gini 3 175 0.92 0.11 0.97 0.95 0.04 0.39 1.00 0.61
## skew kurtosis se
## physhealth -1.23 0.78 1.80
## pos_c_mean -0.04 -0.08 0.06
## pos_b_gini -2.37 5.97 0.01Plot bivariate associations:
pairs.panels(zautra[,c("physhealth","pos_c_mean","pos_b_gini")],
lm = TRUE, hist = "gray")
Fit regression model:
mod1a <- lm(physhealth ~ pos_c_mean.c,
data = zautra)
summary(mod1a)##
## Call:
## lm(formula = physhealth ~ pos_c_mean.c, data = zautra)
##
## Residuals:
## Min 1Q Median 3Q Max
## -60.543 -11.753 4.911 13.625 35.336
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 77.655 1.647 47.16 < 2e-16 ***
## pos_c_mean.c 11.302 2.153 5.25 5.56e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.55 on 139 degrees of freedom
## (34 observations deleted due to missingness)
## Multiple R-squared: 0.1655, Adjusted R-squared: 0.1595
## F-statistic: 27.57 on 1 and 139 DF, p-value: 5.557e-07mod1a_adjrsq <- summary(mod1a)$adj.r.squared
mod1a_beta1 <- summary(mod1a)$coefficients[2]Fit regression model:
mod1b <- lm(physhealth ~ pos_c_mean.c + pos_b_gini.c,
data = zautra)
summary(mod1b)##
## Call:
## lm(formula = physhealth ~ pos_c_mean.c + pos_b_gini.c, data = zautra)
##
## Residuals:
## Min 1Q Median 3Q Max
## -59.01 -11.62 4.58 12.80 44.28
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 77.641 1.644 47.22 <2e-16 ***
## pos_c_mean.c 8.619 3.101 2.78 0.0062 **
## pos_b_gini.c 25.322 21.098 1.20 0.2321
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.52 on 138 degrees of freedom
## (34 observations deleted due to missingness)
## Multiple R-squared: 0.1741, Adjusted R-squared: 0.1622
## F-statistic: 14.55 on 2 and 138 DF, p-value: 1.85e-06mod1b_adjrsq <- summary(mod1b)$adj.r.squared
mod1b_beta1 <- summary(mod1b)$coefficients[2]
mod1b_beta2 <- summary(mod1b)$coefficients[3]Fit regression model:
mod1c <- lm(physhealth ~ pos_c_mean.c + pos_b_gini.c + I(pos_c_mean.c*pos_b_gini.c),
data = zautra)
summary(mod1c)##
## Call:
## lm(formula = physhealth ~ pos_c_mean.c + pos_b_gini.c + I(pos_c_mean.c *
## pos_b_gini.c), data = zautra)
##
## Residuals:
## Min 1Q Median 3Q Max
## -61.718 -10.598 5.938 12.918 43.723
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 79.086 2.369 33.384 < 2e-16 ***
## pos_c_mean.c 10.030 3.521 2.848 0.00507 **
## pos_b_gini.c -2.527 39.037 -0.065 0.94848
## I(pos_c_mean.c * pos_b_gini.c) -22.882 26.975 -0.848 0.39777
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 19.54 on 137 degrees of freedom
## (34 observations deleted due to missingness)
## Multiple R-squared: 0.1784, Adjusted R-squared: 0.1605
## F-statistic: 9.919 on 3 and 137 DF, p-value: 5.82e-06mod1c_adjrsq <- summary(mod1c)$adj.r.squared
mod1c_beta1 <- summary(mod1c)$coefficients[2]
mod1c_beta2 <- summary(mod1c)$coefficients[3]Model fit comparisons:
anova(mod1a,mod1b); fit1 <- anova(mod1a,mod1b)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ pos_c_mean.c
## Model 2: physhealth ~ pos_c_mean.c + pos_b_gini.c
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 139 53146
## 2 138 52597 1 549.03 1.4405 0.2321anova(mod1b,mod1c); fit2 <- anova(mod1b,mod1c)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ pos_c_mean.c + pos_b_gini.c
## Model 2: physhealth ~ pos_c_mean.c + pos_b_gini.c + I(pos_c_mean.c * pos_b_gini.c)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 138 52597
## 2 137 52322 1 274.8 0.7195 0.3978Summary:
The adjusted \(R^2\) for
- mod1a = 0.1595
- mod1b = 0.1622
- mod1c = 0.1605
- Indicating an increase of 0.0027 from mod1a to mod1b
- And an additional increase of -0.0017 from mod1b to mod1c in the overall variance explained.
Model fit comparisons:
- The change in fit from mod1a to mod1b was 0.232
- The change in fit from mod1b to mod1c was 0.398
The beta coefficient for mean positive emotion was
- 11.3021 for mod1a
- 8.6193 for mod1b
- 10.03 for mod1c
- In this case, the beta coefficient did not increase from mod1a to mod1b
The beta coefficient for positive emodiversity was
- 25.3217 for mod1b
- -2.5272 for mod1c
- The association between positive emodiversity and physical health remained the same (in terms of sign direction) when comparing the bivariate correlation to the regression model results
Conclusions:
- Given there was no sign flip for positive emodiversity and no increase in the beta coefficient value for mean positive emotion, it does not seem that statistical suppression is occurring here. The model fit information also suggests that the mod1a (just mean positive emotion) provides the most best/ most parsimonious fit to these data.
Mean Negative Emotion and Negative Emodiversity Predicting Physical Health - Diary Data
Descritpives:
describe(zautra[,c("physhealth","neg_c_mean","neg_b_gini")])## vars n mean sd median trimmed mad min max range
## physhealth 1 141 77.74 21.33 84.44 81.02 16.40 12.06 100.00 87.94
## neg_c_mean 2 173 0.38 0.38 0.26 0.31 0.21 0.01 2.51 2.50
## neg_b_gini 3 173 0.49 0.18 0.48 0.48 0.19 0.12 0.95 0.83
## skew kurtosis se
## physhealth -1.23 0.78 1.80
## neg_c_mean 2.69 9.66 0.03
## neg_b_gini 0.13 -0.50 0.01Plot bivariate associations:
pairs.panels(zautra[,c("physhealth","neg_c_mean","neg_b_gini")],
lm = TRUE, hist = "gray")
Fit regression model:
mod2a <- lm(physhealth ~ neg_c_mean.c,
data = zautra)
summary(mod2a)##
## Call:
## lm(formula = physhealth ~ neg_c_mean.c, data = zautra)
##
## Residuals:
## Min 1Q Median 3Q Max
## -63.753 -9.363 4.410 13.058 43.420
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 76.904 1.611 47.741 < 2e-16 ***
## neg_c_mean.c -28.870 4.729 -6.104 9.99e-09 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.97 on 137 degrees of freedom
## (36 observations deleted due to missingness)
## Multiple R-squared: 0.2138, Adjusted R-squared: 0.2081
## F-statistic: 37.26 on 1 and 137 DF, p-value: 9.992e-09mod2a_adjrsq <- summary(mod2a)$adj.r.squared
mod2a_beta1 <- summary(mod2a)$coefficients[2]Fit regression model:
mod2b <- lm(physhealth ~ neg_c_mean.c + neg_b_gini.c,
data = zautra)
summary(mod2b)##
## Call:
## lm(formula = physhealth ~ neg_c_mean.c + neg_b_gini.c, data = zautra)
##
## Residuals:
## Min 1Q Median 3Q Max
## -59.586 -9.264 4.763 13.388 41.518
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 76.904 1.592 48.293 < 2e-16 ***
## neg_c_mean.c -35.824 5.782 -6.196 6.45e-09 ***
## neg_b_gini.c 23.756 11.620 2.045 0.0428 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.75 on 136 degrees of freedom
## (36 observations deleted due to missingness)
## Multiple R-squared: 0.2373, Adjusted R-squared: 0.2261
## F-statistic: 21.15 on 2 and 136 DF, p-value: 1.002e-08mod2b_adjrsq <- summary(mod2b)$adj.r.squared
mod2b_beta1 <- summary(mod2b)$coefficients[2]
mod2b_beta2 <- summary(mod2b)$coefficients[3]Fit regression model:
mod2c <- lm(physhealth ~ neg_c_mean.c + neg_b_gini.c + I(neg_c_mean.c*neg_b_gini.c),
data = zautra)
summary(mod2c)##
## Call:
## lm(formula = physhealth ~ neg_c_mean.c + neg_b_gini.c + I(neg_c_mean.c *
## neg_b_gini.c), data = zautra)
##
## Residuals:
## Min 1Q Median 3Q Max
## -64.412 -6.820 5.419 12.731 34.472
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 73.712 1.807 40.788 < 2e-16 ***
## neg_c_mean.c -52.761 7.525 -7.011 1.03e-10 ***
## neg_b_gini.c 38.465 12.035 3.196 0.00173 **
## I(neg_c_mean.c * neg_b_gini.c) 87.384 26.075 3.351 0.00104 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.08 on 135 degrees of freedom
## (36 observations deleted due to missingness)
## Multiple R-squared: 0.2959, Adjusted R-squared: 0.2802
## F-statistic: 18.91 on 3 and 135 DF, p-value: 2.689e-10mod2c_adjrsq <- summary(mod2c)$adj.r.squared
mod2c_beta1 <- summary(mod2c)$coefficients[2]
mod2c_beta2 <- summary(mod2c)$coefficients[3]Model fit comparisons:
anova(mod2a,mod2b); fit1 <- anova(mod2a,mod2b)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ neg_c_mean.c
## Model 2: physhealth ~ neg_c_mean.c + neg_b_gini.c
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 137 49277
## 2 136 47808 1 1469.4 4.18 0.04283 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1anova(mod2b,mod2c); fit2 <- anova(mod2b,mod2c)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ neg_c_mean.c + neg_b_gini.c
## Model 2: physhealth ~ neg_c_mean.c + neg_b_gini.c + I(neg_c_mean.c * neg_b_gini.c)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 136 47808
## 2 135 44136 1 3671.8 11.231 0.001044 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1Summary:
The adjusted \(R^2\) for
- mod2a = 0.2081
- mod2b = 0.2261
- mod2c = 0.2802
- Indicating an increase of 0.018 from mod2a to mod2b
- And an additional increase of 0.0541 from mod2b to mod2c in the overall variance explained.
Model fit comparisons:
- The change in fit from mod2a to mod2b was 0.043
- The change in fit from mod2b to mod2c was 0.001
The beta coefficient for mean negative emotion was
- -28.8702 for mod2a
- -35.8243 for mod2b
- -52.7612 for mod2c
- Note an increase in the beta value in the jump from mod2a to mod2b
The beta coefficient for negative emodiversity was
- 23.7563 for mod2b
- 38.465 for mod2c
- Note that the association between negative emodiversity and physical health flipped such that the sign of the bivariate correlation value was negative, and the sign of the beta coefficient in the regressions were positive.
Conclusions:
- Given the sign flip for negative emodiversity and the increase in the beta coefficient value for mean negative emotion, it does seem that statistical suppression is occurring here. The model fit information also suggests that the mod2c (mean negative emotion, negative emodiversity, and the ixn) provides the most best/ most parsimonious fit to these data.
Mean Pos + Neg Emotion and Neg + Pos Emodiversity Predicting Physical Health - Diary Data
Descritpives:
describe(zautra[,c("physhealth","pos_c_mean","neg_c_mean","pos_b_gini","neg_b_gini")])## vars n mean sd median trimmed mad min max range
## physhealth 1 141 77.74 21.33 84.44 81.02 16.40 12.06 100.00 87.94
## pos_c_mean 2 175 2.09 0.77 2.10 2.09 0.69 0.20 3.96 3.76
## neg_c_mean 3 173 0.38 0.38 0.26 0.31 0.21 0.01 2.51 2.50
## pos_b_gini 4 175 0.92 0.11 0.97 0.95 0.04 0.39 1.00 0.61
## neg_b_gini 5 173 0.49 0.18 0.48 0.48 0.19 0.12 0.95 0.83
## skew kurtosis se
## physhealth -1.23 0.78 1.80
## pos_c_mean -0.04 -0.08 0.06
## neg_c_mean 2.69 9.66 0.03
## pos_b_gini -2.37 5.97 0.01
## neg_b_gini 0.13 -0.50 0.01Plot bivariate associations:
pairs.panels(zautra[,c("physhealth","pos_c_mean","neg_c_mean","pos_b_gini","neg_b_gini")],
lm = TRUE, hist = "gray")
Fit regression model:
mod3a <- lm(physhealth ~ pos_c_mean.c + neg_c_mean.c,
data = zautra)
summary(mod3a)##
## Call:
## lm(formula = physhealth ~ pos_c_mean.c + neg_c_mean.c, data = zautra)
##
## Residuals:
## Min 1Q Median 3Q Max
## -55.419 -8.238 4.194 13.433 38.185
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 77.035 1.542 49.964 < 2e-16 ***
## pos_c_mean.c 7.932 2.149 3.691 0.000323 ***
## neg_c_mean.c -23.208 4.779 -4.857 3.23e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.15 on 136 degrees of freedom
## (36 observations deleted due to missingness)
## Multiple R-squared: 0.2854, Adjusted R-squared: 0.2749
## F-statistic: 27.16 on 2 and 136 DF, p-value: 1.192e-10mod3a_adjrsq <- summary(mod3a)$adj.r.squared
mod3a_betapos <- summary(mod3a)$coefficients[2]
mod3a_betaneg <- summary(mod3a)$coefficients[3]Fit regression model:
mod3b <- lm(physhealth ~ pos_c_mean.c + neg_c_mean.c + pos_b_gini.c + neg_b_gini.c,
data = zautra)
summary(mod3b)##
## Call:
## lm(formula = physhealth ~ pos_c_mean.c + neg_c_mean.c + pos_b_gini.c +
## neg_b_gini.c, data = zautra)
##
## Residuals:
## Min 1Q Median 3Q Max
## -54.449 -9.379 4.123 12.769 35.714
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 77.036 1.524 50.544 < 2e-16 ***
## pos_c_mean.c 7.605 3.091 2.460 0.0152 *
## neg_c_mean.c -30.162 5.875 -5.134 9.78e-07 ***
## pos_b_gini.c 4.550 20.813 0.219 0.8273
## neg_b_gini.c 24.342 11.787 2.065 0.0408 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.94 on 134 degrees of freedom
## (36 observations deleted due to missingness)
## Multiple R-squared: 0.312, Adjusted R-squared: 0.2914
## F-statistic: 15.19 on 4 and 134 DF, p-value: 2.881e-10mod3b_adjrsq <- summary(mod3b)$adj.r.squared
mod3b_betapos <- summary(mod3b)$coefficients[2]
mod3b_betaneg <- summary(mod3b)$coefficients[3]Fit regression model:
mod3c <- lm(physhealth ~ pos_c_mean.c + neg_c_mean.c + pos_b_gini.c + neg_b_gini.c +
I(pos_c_mean.c*pos_b_gini.c) + I(neg_c_mean.c*neg_b_gini.c),
data = zautra)
summary(mod3c)##
## Call:
## lm(formula = physhealth ~ pos_c_mean.c + neg_c_mean.c + pos_b_gini.c +
## neg_b_gini.c + I(pos_c_mean.c * pos_b_gini.c) + I(neg_c_mean.c *
## neg_b_gini.c), data = zautra)
##
## Residuals:
## Min 1Q Median 3Q Max
## -50.514 -8.639 4.206 12.829 31.045
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 77.030 2.321 33.185 < 2e-16 ***
## pos_c_mean.c 7.941 3.357 2.366 0.01946 *
## neg_c_mean.c -44.794 7.824 -5.725 6.63e-08 ***
## pos_b_gini.c -39.980 35.659 -1.121 0.26424
## neg_b_gini.c 33.283 12.184 2.732 0.00716 **
## I(pos_c_mean.c * pos_b_gini.c) -40.619 24.533 -1.656 0.10016
## I(neg_c_mean.c * neg_b_gini.c) 68.532 26.103 2.625 0.00968 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.46 on 132 degrees of freedom
## (36 observations deleted due to missingness)
## Multiple R-squared: 0.358, Adjusted R-squared: 0.3288
## F-statistic: 12.27 on 6 and 132 DF, p-value: 6.131e-11mod3c_adjrsq <- summary(mod3c)$adj.r.squared
mod3c_betapos <- summary(mod3c)$coefficients[2]
mod3c_betaneg <- summary(mod3c)$coefficients[3]Model fit comparisons:
anova(mod3a,mod3b)## Analysis of Variance Table
##
## Model 1: physhealth ~ pos_c_mean.c + neg_c_mean.c
## Model 2: physhealth ~ pos_c_mean.c + neg_c_mean.c + pos_b_gini.c + neg_b_gini.c
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 136 44791
## 2 134 43125 2 1666.2 2.5886 0.07888 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1anova(mod3a,mod3c)## Analysis of Variance Table
##
## Model 1: physhealth ~ pos_c_mean.c + neg_c_mean.c
## Model 2: physhealth ~ pos_c_mean.c + neg_c_mean.c + pos_b_gini.c + neg_b_gini.c +
## I(pos_c_mean.c * pos_b_gini.c) + I(neg_c_mean.c * neg_b_gini.c)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 136 44791
## 2 132 40243 4 4547.9 3.7293 0.006571 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1anova(mod3b,mod3c)## Analysis of Variance Table
##
## Model 1: physhealth ~ pos_c_mean.c + neg_c_mean.c + pos_b_gini.c + neg_b_gini.c
## Model 2: physhealth ~ pos_c_mean.c + neg_c_mean.c + pos_b_gini.c + neg_b_gini.c +
## I(pos_c_mean.c * pos_b_gini.c) + I(neg_c_mean.c * neg_b_gini.c)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 134 43125
## 2 132 40243 2 2881.7 4.7261 0.01041 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1Summary and conclusions: These results echo what was reported in the model 1 series (a, b, c) and model 2 series (a, b, c). This additional set of analyses is given to more closely match what was reported in Benson, Ram, Almeida, Zautra, & Ong, 2017. The model fit comparisons also suggest that mod3c provides the best/ most parsimonious fit to these data.
Mean Positive Emotion and Positive Emodiversity Predicting Physical Health - Single Occasion Data
Note that in these analyses where emodiversity was calculated with single occasion data, I utilize the Quoidbach method of retaining the 0-4 scale for the emotion items.
Using the Gini Coefficient
Descritpives:
describe(zautra2[,c("physhealth","panaspos_c_mean_4wk","panaspos_c_gini_4wk")])## vars n mean sd median trimmed mad min max
## physhealth 1 148 78.05 20.94 85.09 81.21 15.38 12.06 100.0
## panaspos_c_mean_4wk 2 148 3.49 0.73 3.70 3.54 0.59 1.30 4.9
## panaspos_c_gini_4wk 3 148 0.84 0.15 0.88 0.86 0.09 0.17 1.0
## range skew kurtosis se
## physhealth 87.94 -1.27 0.96 1.72
## panaspos_c_mean_4wk 3.60 -0.69 0.37 0.06
## panaspos_c_gini_4wk 0.83 -1.97 4.41 0.01Plot bivariate associations:
pairs.panels(zautra2[,c("physhealth","panaspos_c_mean_4wk","panaspos_c_gini_4wk")],
lm = TRUE, hist = "gray")
Fit regression model:
mod4a <- lm(physhealth ~ panaspos_c_mean_4wk.c,
data = zautra2)
summary(mod4a)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c, data = zautra2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -57.694 -6.087 3.937 11.964 36.737
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 78.051 1.506 51.82 < 2e-16 ***
## panaspos_c_mean_4wk.c 14.094 2.079 6.78 2.77e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.33 on 146 degrees of freedom
## Multiple R-squared: 0.2395, Adjusted R-squared: 0.2342
## F-statistic: 45.97 on 1 and 146 DF, p-value: 2.771e-10mod4a_adjrsq <- summary(mod4a)$adj.r.squared
mod4a_beta1 <- summary(mod4a)$coefficients[2]Fit regression model:
mod4b <- lm(physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_gini_4wk.c,
data = zautra2)
summary(mod4b)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_gini_4wk.c,
## data = zautra2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -56.863 -6.958 4.286 10.447 42.590
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 78.051 1.479 52.775 <2e-16 ***
## panaspos_c_mean_4wk.c 7.617 3.266 2.333 0.0210 *
## panaspos_c_gini_4wk.c 41.111 16.183 2.540 0.0121 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.99 on 145 degrees of freedom
## Multiple R-squared: 0.2719, Adjusted R-squared: 0.2618
## F-statistic: 27.07 on 2 and 145 DF, p-value: 1.025e-10mod4b_adjrsq <- summary(mod4b)$adj.r.squared
mod4b_beta1 <- summary(mod4b)$coefficients[2]
mod4b_beta2 <- summary(mod4b)$coefficients[3]Fit regression model:
mod4c <- lm(physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_gini_4wk.c + I(panaspos_c_mean_4wk.c*panaspos_c_gini_4wk.c),
data = zautra2)
summary(mod4c)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_gini_4wk.c +
## I(panaspos_c_mean_4wk.c * panaspos_c_gini_4wk.c), data = zautra2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -57.129 -7.097 4.388 10.547 43.363
##
## Coefficients:
## Estimate Std. Error
## (Intercept) 78.299 1.807
## panaspos_c_mean_4wk.c 7.840 3.405
## panaspos_c_gini_4wk.c 37.218 22.939
## I(panaspos_c_mean_4wk.c * panaspos_c_gini_4wk.c) -2.995 12.464
## t value Pr(>|t|)
## (Intercept) 43.335 <2e-16 ***
## panaspos_c_mean_4wk.c 2.303 0.0227 *
## panaspos_c_gini_4wk.c 1.622 0.1069
## I(panaspos_c_mean_4wk.c * panaspos_c_gini_4wk.c) -0.240 0.8105
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.05 on 144 degrees of freedom
## Multiple R-squared: 0.2722, Adjusted R-squared: 0.257
## F-statistic: 17.95 on 3 and 144 DF, p-value: 5.962e-10mod4c_adjrsq <- summary(mod4c)$adj.r.squared
mod4c_beta1 <- summary(mod4c)$coefficients[2]
mod4c_beta2 <- summary(mod4c)$coefficients[3]Model fit comparisons:
anova(mod4a,mod4b); fit1 <- anova(mod4a,mod4b)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panaspos_c_mean_4wk.c
## Model 2: physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_gini_4wk.c
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 146 49028
## 2 145 46939 1 2089.1 6.4534 0.01213 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1anova(mod4b,mod4c); fit2 <- anova(mod4b,mod4c)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_gini_4wk.c
## Model 2: physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_gini_4wk.c +
## I(panaspos_c_mean_4wk.c * panaspos_c_gini_4wk.c)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 145 46939
## 2 144 46921 1 18.809 0.0577 0.8105Summary:
The adjusted \(R^2\) for
- mod4a = 0.2342
- mod4b = 0.2618
- mod4c = 0.257
- Indicating an increase of 0.0276 from mod4a to mod4b
- And an additional increase of -0.0048 from mod4b to mod4c in the overall variance explained.
Model fit comparisons:
- The change in fit from mod4a to mod4b was 0.012
- The change in fit from mod4b to mod4c was 0.81
The beta coefficient for mean positive emotion was
- 14.0936 for mod4a
- 7.6173 for mod4b
- 7.8399 for mod4c
- In this case, the beta coefficient did not increase from mod4a to mod4b
The beta coefficient for positive emodiversity was
- 41.1112 for mod4b
- 37.2179 for mod4c
- The association between positive emodiversity and physical health remained the same (in terms of sign direction) when comparing the bivariate correlation to the regression model results
Conclusions:
- Given there was no sign flip for positive emodiversity and no increase in the beta coefficient value for mean positive emotion, it does not seem that statistical suppression is occurring here. The model fit information also suggests that the mod4b (mean positive emotion and positive emodiversity) provides the most best/ most parsimonious fit to these data.
Using Shannon’s Entropy
Descritpives:
describe(zautra2[,c("physhealth","panaspos_c_mean_4wk","panaspos_c_shannon_4wk")])## vars n mean sd median trimmed mad min
## physhealth 1 148 78.05 20.94 85.09 81.21 15.38 12.06
## panaspos_c_mean_4wk 2 148 3.49 0.73 3.70 3.54 0.59 1.30
## panaspos_c_shannon_4wk 3 148 2.19 0.22 2.27 2.24 0.04 0.64
## max range skew kurtosis se
## physhealth 100.0 87.94 -1.27 0.96 1.72
## panaspos_c_mean_4wk 4.9 3.60 -0.69 0.37 0.06
## panaspos_c_shannon_4wk 2.3 1.67 -3.99 19.65 0.02Plot bivariate associations:
pairs.panels(zautra2[,c("physhealth","panaspos_c_mean_4wk","panaspos_c_shannon_4wk")],
lm = TRUE, hist = "gray")
Fit regression model:
mod4d <- lm(physhealth ~ panaspos_c_mean_4wk.c,
data = zautra2)
summary(mod4d)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c, data = zautra2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -57.694 -6.087 3.937 11.964 36.737
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 78.051 1.506 51.82 < 2e-16 ***
## panaspos_c_mean_4wk.c 14.094 2.079 6.78 2.77e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.33 on 146 degrees of freedom
## Multiple R-squared: 0.2395, Adjusted R-squared: 0.2342
## F-statistic: 45.97 on 1 and 146 DF, p-value: 2.771e-10mod4d_adjrsq <- summary(mod4d)$adj.r.squared
mod4d_beta1 <- summary(mod4d)$coefficients[2]Fit regression model:
mod4e <- lm(physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_shannon_4wk.c,
data = zautra2)
summary(mod4e)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_shannon_4wk.c,
## data = zautra2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -59.597 -7.619 3.644 11.314 42.756
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 78.051 1.480 52.744 < 2e-16 ***
## panaspos_c_mean_4wk.c 9.132 2.845 3.210 0.00163 **
## panaspos_c_shannon_4wk.c 23.238 9.276 2.505 0.01335 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18 on 145 degrees of freedom
## Multiple R-squared: 0.271, Adjusted R-squared: 0.261
## F-statistic: 26.95 on 2 and 145 DF, p-value: 1.116e-10mod4e_adjrsq <- summary(mod4e)$adj.r.squared
mod4e_beta1 <- summary(mod4e)$coefficients[2]
mod4e_beta2 <- summary(mod4e)$coefficients[3]Fit regression model:
mod4f <- lm(physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_shannon_4wk.c + I(panaspos_c_mean_4wk.c*panaspos_c_shannon_4wk.c),
data = zautra2)
summary(mod4f)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_shannon_4wk.c +
## I(panaspos_c_mean_4wk.c * panaspos_c_shannon_4wk.c), data = zautra2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -59.552 -7.430 3.672 11.083 42.548
##
## Coefficients:
## Estimate Std. Error
## (Intercept) 77.511 1.987
## panaspos_c_mean_4wk.c 8.382 3.392
## panaspos_c_shannon_4wk.c 32.315 24.074
## I(panaspos_c_mean_4wk.c * panaspos_c_shannon_4wk.c) 4.818 11.787
## t value Pr(>|t|)
## (Intercept) 39.007 <2e-16 ***
## panaspos_c_mean_4wk.c 2.471 0.0146 *
## panaspos_c_shannon_4wk.c 1.342 0.1816
## I(panaspos_c_mean_4wk.c * panaspos_c_shannon_4wk.c) 0.409 0.6833
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.05 on 144 degrees of freedom
## Multiple R-squared: 0.2719, Adjusted R-squared: 0.2567
## F-statistic: 17.92 on 3 and 144 DF, p-value: 6.14e-10mod4f_adjrsq <- summary(mod4f)$adj.r.squared
mod4f_beta1 <- summary(mod4f)$coefficients[2]
mod4f_beta2 <- summary(mod4f)$coefficients[3]Model fit comparisons:
anova(mod4d,mod4e); fit1 <- anova(mod4d,mod4e)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panaspos_c_mean_4wk.c
## Model 2: physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_shannon_4wk.c
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 146 49028
## 2 145 46995 1 2034 6.2758 0.01335 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1anova(mod4e,mod4f); fit2 <- anova(mod4e,mod4f)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_shannon_4wk.c
## Model 2: physhealth ~ panaspos_c_mean_4wk.c + panaspos_c_shannon_4wk.c +
## I(panaspos_c_mean_4wk.c * panaspos_c_shannon_4wk.c)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 145 46995
## 2 144 46940 1 54.473 0.1671 0.6833Summary:
The adjusted \(R^2\) for
- mod4d = 0.2342
- mod4e = 0.261
- mod4f = 0.2567
- Indicating an increase of 0.0267 from mod4d to mod4e
- And an additional increase of -0.0043 from mod4e to mod4f in the overall variance explained.
Model fit comparisons:
- The change in fit from mod4d to mod4e was 0.013
- The change in fit from mod4e to mod4f was 0.683
The beta coefficient for mean positive emotion was
- 14.0936 for mod4d
- 9.1323 for mod4e
- 8.3822 for mod4f
- In this case, the beta coefficient does not increase from mod4d to mod4e
The beta coefficient for positive emodiversity was
- 23.2381 for mod4e
- 32.3148 for mod4f
- The association between positive emodiversity and physical health remains the same (in terms of sign direction) when comparing the bivariate correlation to the regression model results.
Conclusions:
- Given there was no sign flip for positive emodiversity and no increase in the beta coefficient value for mean positive emotion, it seems that statistical suppression is not occurring here. The model fit information also suggests that the mod4e (mean positive emotion and positive emodiversity) provides the most best/ most parsimonious fit to these data.
Mean Negative Emotion and Negative Emodiversity Predicting Physical Health - Single Occasion Data
Using the Gini Coefficient
Descritpives:
describe(zautra3[,c("physhealth","panasneg_c_mean_4wk","panasneg_c_gini_4wk")])## vars n mean sd median trimmed mad min max
## physhealth 1 118 75.87 21.06 82.16 78.74 17.19 12.06 100.0
## panasneg_c_mean_4wk 2 118 1.79 0.65 1.60 1.69 0.59 1.10 3.9
## panasneg_c_gini_4wk 3 118 0.45 0.24 0.44 0.44 0.26 0.10 1.0
## range skew kurtosis se
## physhealth 87.94 -1.16 0.68 1.94
## panasneg_c_mean_4wk 2.80 1.38 1.60 0.06
## panasneg_c_gini_4wk 0.90 0.30 -0.74 0.02Plot bivariate associations:
pairs.panels(zautra3[,c("physhealth","panasneg_c_mean_4wk","panasneg_c_gini_4wk")],
lm = TRUE, hist = "gray")
Fit regression model:
mod5a <- lm(physhealth ~ panasneg_c_mean_4wk.c,
data = zautra3)
summary(mod5a)##
## Call:
## lm(formula = physhealth ~ panasneg_c_mean_4wk.c, data = zautra3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -55.215 -11.680 3.735 12.938 37.955
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 75.870 1.721 44.092 < 2e-16 ***
## panasneg_c_mean_4wk.c -15.232 2.671 -5.703 9.15e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.69 on 116 degrees of freedom
## Multiple R-squared: 0.219, Adjusted R-squared: 0.2122
## F-statistic: 32.52 on 1 and 116 DF, p-value: 9.153e-08mod5a_adjrsq <- summary(mod5a)$adj.r.squared
mod5a_beta1 <- summary(mod5a)$coefficients[2]Fit regression model:
mod5b <- lm(physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_gini_4wk.c,
data = zautra3)
summary(mod5b)##
## Call:
## lm(formula = physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_gini_4wk.c,
## data = zautra3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -53.681 -10.956 3.567 14.142 40.706
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 75.870 1.705 44.505 < 2e-16 ***
## panasneg_c_mean_4wk.c -20.848 4.112 -5.070 1.54e-06 ***
## panasneg_c_gini_4wk.c 19.547 10.955 1.784 0.077 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.52 on 115 degrees of freedom
## Multiple R-squared: 0.24, Adjusted R-squared: 0.2268
## F-statistic: 18.16 on 2 and 115 DF, p-value: 1.403e-07mod5b_adjrsq <- summary(mod5b)$adj.r.squared
mod5b_beta1 <- summary(mod5b)$coefficients[2]
mod5b_beta2 <- summary(mod5b)$coefficients[3]Fit regression model:
mod5c <- lm(physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_gini_4wk.c + I(panasneg_c_mean_4wk.c*panasneg_c_gini_4wk.c),
data = zautra3)
summary(mod5c)##
## Call:
## lm(formula = physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_gini_4wk.c +
## I(panasneg_c_mean_4wk.c * panasneg_c_gini_4wk.c), data = zautra3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -55.125 -10.431 4.401 13.224 39.074
##
## Coefficients:
## Estimate Std. Error
## (Intercept) 76.978 2.359
## panasneg_c_mean_4wk.c -18.346 5.519
## panasneg_c_gini_4wk.c 15.537 12.458
## I(panasneg_c_mean_4wk.c * panasneg_c_gini_4wk.c) -9.299 13.646
## t value Pr(>|t|)
## (Intercept) 32.629 < 2e-16 ***
## panasneg_c_mean_4wk.c -3.324 0.00119 **
## panasneg_c_gini_4wk.c 1.247 0.21492
## I(panasneg_c_mean_4wk.c * panasneg_c_gini_4wk.c) -0.681 0.49696
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.56 on 114 degrees of freedom
## Multiple R-squared: 0.2431, Adjusted R-squared: 0.2232
## F-statistic: 12.2 on 3 and 114 DF, p-value: 5.536e-07mod5c_adjrsq <- summary(mod5c)$adj.r.squared
mod5c_beta1 <- summary(mod5c)$coefficients[2]
mod5c_beta2 <- summary(mod5c)$coefficients[3]Model fit comparisons:
anova(mod5a,mod5b); fit1 <- anova(mod5a,mod5b)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panasneg_c_mean_4wk.c
## Model 2: physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_gini_4wk.c
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 116 40528
## 2 115 39437 1 1091.8 3.1839 0.077 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1anova(mod5b,mod5c); fit2 <- anova(mod5b,mod5c)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_gini_4wk.c
## Model 2: physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_gini_4wk.c +
## I(panasneg_c_mean_4wk.c * panasneg_c_gini_4wk.c)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 115 39437
## 2 114 39277 1 160 0.4644 0.497Summary:
The adjusted \(R^2\) for
- mod5a = 0.2122
- mod5b = 0.2268
- mod5c = 0.2232
- Indicating an increase of 0.0146 from mod5a to mod5b
- And an additional increase of -0.0036 from mod5b to mod5c in the overall variance explained.
Model fit comparisons:
- The change in fit from mod5a to mod5b was 0.077
- The change in fit from mod5b to mod5c was 0.497
The beta coefficient for mean negative emotion was
- -15.2322 for mod5a
- -20.8476 for mod5b
- -18.346 for mod5c
- Note an increase in the beta value in the jump from mod5a to mod5b
The beta coefficient for negative emodiversity was
- 19.5474 for mod5b
- 15.5367 for mod5c
- Note that the association between negative emodiversity and physical health flipped such that the sign of the bivariate correlation value was negative, and the sign of the beta coefficient in the regressions were positive.
Conclusions:
- Given the sign flip for negative emodiversity and the increase in the beta coefficient value for mean negative emotion, it does seem that statistical suppression is occurring here. The model fit information also suggests that the mod5a (just mean negative emotion) provides the most best/ most parsimonious fit to these data.
Using Shannon’s Entropy
Descritpives:
describe(zautra3[,c("physhealth","panasneg_c_mean_4wk","panasneg_c_shannon_4wk")])## vars n mean sd median trimmed mad min
## physhealth 1 118 75.87 21.06 82.16 78.74 17.19 12.06
## panasneg_c_mean_4wk 2 118 1.79 0.65 1.60 1.69 0.59 1.10
## panasneg_c_shannon_4wk 3 118 1.38 0.70 1.59 1.44 0.66 0.00
## max range skew kurtosis se
## physhealth 100.0 87.94 -1.16 0.68 1.94
## panasneg_c_mean_4wk 3.9 2.80 1.38 1.60 0.06
## panasneg_c_shannon_4wk 2.3 2.30 -0.76 -0.51 0.06Plot bivariate associations:
pairs.panels(zautra3[,c("physhealth","panasneg_c_mean_4wk","panasneg_c_shannon_4wk")],
lm = TRUE, hist = "gray")
Fit regression model:
mod5d <- lm(physhealth ~ panasneg_c_mean_4wk.c,
data = zautra3)
summary(mod5d)##
## Call:
## lm(formula = physhealth ~ panasneg_c_mean_4wk.c, data = zautra3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -55.215 -11.680 3.735 12.938 37.955
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 75.870 1.721 44.092 < 2e-16 ***
## panasneg_c_mean_4wk.c -15.232 2.671 -5.703 9.15e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.69 on 116 degrees of freedom
## Multiple R-squared: 0.219, Adjusted R-squared: 0.2122
## F-statistic: 32.52 on 1 and 116 DF, p-value: 9.153e-08mod5d_adjrsq <- summary(mod5d)$adj.r.squared
mod5d_beta1 <- summary(mod5d)$coefficients[2]Fit regression model:
mod5e <- lm(physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_shannon_4wk.c,
data = zautra3)
summary(mod5e)##
## Call:
## lm(formula = physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_shannon_4wk.c,
## data = zautra3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -55.751 -11.301 4.211 13.493 40.011
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 75.870 1.714 44.270 < 2e-16 ***
## panasneg_c_mean_4wk.c -19.339 3.973 -4.867 3.64e-06 ***
## panasneg_c_shannon_4wk.c 5.089 3.657 1.392 0.167
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.62 on 115 degrees of freedom
## Multiple R-squared: 0.2319, Adjusted R-squared: 0.2185
## F-statistic: 17.36 on 2 and 115 DF, p-value: 2.582e-07mod5e_adjrsq <- summary(mod5e)$adj.r.squared
mod5e_beta1 <- summary(mod5e)$coefficients[2]
mod5e_beta2 <- summary(mod5e)$coefficients[3]Fit regression model:
mod5f <- lm(physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_shannon_4wk.c +
I(panasneg_c_mean_4wk.c*panasneg_c_shannon_4wk.c),
data = zautra3)
summary(mod5f)##
## Call:
## lm(formula = physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_shannon_4wk.c +
## I(panasneg_c_mean_4wk.c * panasneg_c_shannon_4wk.c), data = zautra3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -56.247 -10.998 3.984 13.435 39.047
##
## Coefficients:
## Estimate Std. Error
## (Intercept) 76.516 3.475
## panasneg_c_mean_4wk.c -17.640 8.888
## panasneg_c_shannon_4wk.c 3.636 7.721
## I(panasneg_c_mean_4wk.c * panasneg_c_shannon_4wk.c) -1.929 9.016
## t value Pr(>|t|)
## (Intercept) 22.017 <2e-16 ***
## panasneg_c_mean_4wk.c -1.985 0.0496 *
## panasneg_c_shannon_4wk.c 0.471 0.6386
## I(panasneg_c_mean_4wk.c * panasneg_c_shannon_4wk.c) -0.214 0.8310
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 18.69 on 114 degrees of freedom
## Multiple R-squared: 0.2322, Adjusted R-squared: 0.212
## F-statistic: 11.49 on 3 and 114 DF, p-value: 1.223e-06mod5f_adjrsq <- summary(mod5f)$adj.r.squared
mod5f_beta1 <- summary(mod5f)$coefficients[2]
mod5f_beta2 <- summary(mod5f)$coefficients[3]Model fit comparisons:
anova(mod5d,mod5e); fit1 <- anova(mod5d,mod5e)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panasneg_c_mean_4wk.c
## Model 2: physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_shannon_4wk.c
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 116 40528
## 2 115 39857 1 671.15 1.9365 0.1667anova(mod5e,mod5f); fit2 <- anova(mod5e,mod5f)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_shannon_4wk.c
## Model 2: physhealth ~ panasneg_c_mean_4wk.c + panasneg_c_shannon_4wk.c +
## I(panasneg_c_mean_4wk.c * panasneg_c_shannon_4wk.c)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 115 39857
## 2 114 39841 1 15.999 0.0458 0.831Summary:
The adjusted \(R^2\) for
- mod5d = 0.2122
- mod5e = 0.2185
- mod5f = 0.212
- Indicating an increase of 0.0063 from mod5d to mod5e
- And an additional increase of -0.0065 from mod5e to mod5f in the overall variance explained.
Model fit comparisons:
- The change in fit from mod5d to mod5e was 0.167
- The change in fit from mod5e to mod5f was 0.831
The beta coefficient for mean negative emotion was
- -15.2322 for mod5d
- -19.3391 for mod5e
- -17.6399 for mod5f
- Note an increase in the beta value in the jump from mod5d to mod5e
The beta coefficient for negative emodiversity was
- 5.0895 for mod5e
- 3.6363 for mod5f
- Note that the association between negative emodiversity and physical health flipped such that the sign of the bivariate correlation value was negative, and the sign of the beta coefficient in the regressions were positive.
Conclusions:
- Given the sign flip for negative emodiversity and the increase in the beta coefficient value for mean negative emotion, it does seem that statistical suppression is occurring here. The model fit information also suggests that the mod5d (just mean negative emotion) provides the most best/ most parsimonious fit to these data.
Mean Pos + Neg Emotion and Neg + Pos Emodiversity Predicting Physical Health - Single Occasion Data
Using the Gini Coefficient
Descritpives:
describe(zautra4[,c("physhealth","panaspos_c_mean_4wk","panasneg_c_mean_4wk",
"panaspos_c_gini_4wk","panasneg_c_gini_4wk")])## vars n mean sd median trimmed mad min max
## physhealth 1 117 75.69 21.06 82.06 78.55 16.96 12.06 100.0
## panaspos_c_mean_4wk 2 117 3.38 0.72 3.60 3.44 0.59 1.30 4.8
## panasneg_c_mean_4wk 3 117 1.79 0.65 1.60 1.69 0.59 1.10 3.9
## panaspos_c_gini_4wk 4 117 0.82 0.15 0.88 0.85 0.08 0.17 1.0
## panasneg_c_gini_4wk 5 117 0.45 0.24 0.43 0.44 0.27 0.10 1.0
## range skew kurtosis se
## physhealth 87.94 -1.15 0.67 1.95
## panaspos_c_mean_4wk 3.50 -0.75 0.27 0.07
## panasneg_c_mean_4wk 2.80 1.38 1.58 0.06
## panaspos_c_gini_4wk 0.83 -1.94 4.17 0.01
## panasneg_c_gini_4wk 0.90 0.28 -0.77 0.02Plot bivariate associations:
pairs.panels(zautra4[,c("physhealth","panaspos_c_mean_4wk","panasneg_c_mean_4wk",
"panaspos_c_gini_4wk","panasneg_c_gini_4wk")],
lm = TRUE, hist = "gray")
Fit regression model:
mod6a <- lm(physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c,
data = zautra4)
summary(mod6a)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c,
## data = zautra4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -56.205 -8.728 3.563 11.781 38.357
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 75.691 1.641 46.120 < 2e-16 ***
## panaspos_c_mean_4wk.c 9.429 2.641 3.571 0.000522 ***
## panasneg_c_mean_4wk.c -10.218 2.915 -3.505 0.000653 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.75 on 114 degrees of freedom
## Multiple R-squared: 0.3018, Adjusted R-squared: 0.2895
## F-statistic: 24.63 on 2 and 114 DF, p-value: 1.284e-09mod6a_adjrsq <- summary(mod6a)$adj.r.squared
mod6a_beta1 <- summary(mod6a)$coefficients[2]
mod6a_beta2 <- summary(mod6a)$coefficients[3]Fit regression model:
mod6b <- lm(physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c + panaspos_c_gini_4wk.c +
panasneg_c_gini_4wk.c,
data = zautra4)
summary(mod6b)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c +
## panaspos_c_gini_4wk.c + panasneg_c_gini_4wk.c, data = zautra4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -54.011 -7.490 3.037 11.217 43.832
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 75.691 1.617 46.806 < 2e-16 ***
## panaspos_c_mean_4wk.c 3.801 3.793 1.002 0.31854
## panasneg_c_mean_4wk.c -12.700 4.381 -2.899 0.00451 **
## panaspos_c_gini_4wk.c 34.038 17.461 1.949 0.05375 .
## panasneg_c_gini_4wk.c 8.959 11.128 0.805 0.42246
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.49 on 112 degrees of freedom
## Multiple R-squared: 0.3339, Adjusted R-squared: 0.3102
## F-statistic: 14.04 on 4 and 112 DF, p-value: 2.576e-09mod6b_adjrsq <- summary(mod6b)$adj.r.squared
mod6b_beta1 <- summary(mod6b)$coefficients[2]
mod6b_beta2 <- summary(mod6b)$coefficients[3]
mod6b_beta3 <- summary(mod6b)$coefficients[4]
mod6b_beta4 <- summary(mod6b)$coefficients[5]Fit regression model:
mod6c <- lm(physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c + panaspos_c_gini_4wk.c +
panasneg_c_gini_4wk.c + I(panaspos_c_mean_4wk.c*panaspos_c_gini_4wk.c) +
I(panasneg_c_mean_4wk.c*panasneg_c_gini_4wk.c),
data = zautra4)
summary(mod6c)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c +
## panaspos_c_gini_4wk.c + panasneg_c_gini_4wk.c + I(panaspos_c_mean_4wk.c *
## panaspos_c_gini_4wk.c) + I(panasneg_c_mean_4wk.c * panasneg_c_gini_4wk.c),
## data = zautra4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -55.898 -7.282 2.564 11.063 39.298
##
## Coefficients:
## Estimate Std. Error
## (Intercept) 77.7140 2.4694
## panaspos_c_mean_4wk.c 3.1556 3.9143
## panasneg_c_mean_4wk.c -6.9998 5.9154
## panaspos_c_gini_4wk.c 42.1182 24.5351
## panasneg_c_gini_4wk.c -0.4403 12.9290
## I(panaspos_c_mean_4wk.c * panaspos_c_gini_4wk.c) 3.0240 13.6168
## I(panasneg_c_mean_4wk.c * panasneg_c_gini_4wk.c) -19.0626 13.2344
## t value Pr(>|t|)
## (Intercept) 31.471 <2e-16 ***
## panaspos_c_mean_4wk.c 0.806 0.4219
## panasneg_c_mean_4wk.c -1.183 0.2392
## panaspos_c_gini_4wk.c 1.717 0.0889 .
## panasneg_c_gini_4wk.c -0.034 0.9729
## I(panaspos_c_mean_4wk.c * panaspos_c_gini_4wk.c) 0.222 0.8247
## I(panasneg_c_mean_4wk.c * panasneg_c_gini_4wk.c) -1.440 0.1526
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.49 on 110 degrees of freedom
## Multiple R-squared: 0.3464, Adjusted R-squared: 0.3107
## F-statistic: 9.714 on 6 and 110 DF, p-value: 1.43e-08mod6c_adjrsq <- summary(mod6c)$adj.r.squared
mod6c_beta1 <- summary(mod6c)$coefficients[2]
mod6c_beta2 <- summary(mod6c)$coefficients[3]
mod6c_beta3 <- summary(mod6c)$coefficients[4]
mod6c_beta4 <- summary(mod6c)$coefficients[5]Model fit comparisons:
anova(mod6a,mod6b); fit1 <- anova(mod6a,mod6b)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c
## Model 2: physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c +
## panaspos_c_gini_4wk.c + panasneg_c_gini_4wk.c
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 114 35925
## 2 112 34268 2 1656.5 2.707 0.07111 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1anova(mod6b,mod6c); fit2 <- anova(mod6b,mod6c)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c +
## panaspos_c_gini_4wk.c + panasneg_c_gini_4wk.c
## Model 2: physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c +
## panaspos_c_gini_4wk.c + panasneg_c_gini_4wk.c + I(panaspos_c_mean_4wk.c *
## panaspos_c_gini_4wk.c) + I(panasneg_c_mean_4wk.c * panasneg_c_gini_4wk.c)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 112 34268
## 2 110 33630 2 638.04 1.0435 0.3557Summary and conclusions: This additional set of analyses is given to more closely match what was reported in Benson, Ram, Almeida, Zautra, & Ong, 2017 (with the exception here that the Benson et al., 2017 paper did not report any regression results using the single occassion emotion item data). These results echo what was reported in the model 4 series (a, b, c) and model 5 series (a, b, c). The model fit comparisons suggests that mod6a (just mean positive emotion and mean negative emotion) provides the best/ most parsimonious fit to these data.
Using Shannon’s Entropy
Descritpives:
describe(zautra4[,c("physhealth","panaspos_c_mean_4wk","panasneg_c_mean_4wk",
"panaspos_c_shannon_4wk","panasneg_c_shannon_4wk")])## vars n mean sd median trimmed mad min
## physhealth 1 117 75.69 21.06 82.06 78.55 16.96 12.06
## panaspos_c_mean_4wk 2 117 3.38 0.72 3.60 3.44 0.59 1.30
## panasneg_c_mean_4wk 3 117 1.79 0.65 1.60 1.69 0.59 1.10
## panaspos_c_shannon_4wk 4 117 2.18 0.24 2.26 2.24 0.04 0.64
## panasneg_c_shannon_4wk 5 117 1.37 0.70 1.56 1.43 0.69 0.00
## max range skew kurtosis se
## physhealth 100.0 87.94 -1.15 0.67 1.95
## panaspos_c_mean_4wk 4.8 3.50 -0.75 0.27 0.07
## panasneg_c_mean_4wk 3.9 2.80 1.38 1.58 0.06
## panaspos_c_shannon_4wk 2.3 1.67 -3.85 17.84 0.02
## panasneg_c_shannon_4wk 2.3 2.30 -0.76 -0.51 0.06Plot bivariate associations:
pairs.panels(zautra4[,c("physhealth","panaspos_c_mean_4wk","panasneg_c_mean_4wk",
"panaspos_c_shannon_4wk","panasneg_c_shannon_4wk")],
lm = TRUE, hist = "gray")
Fit regression model:
mod6d <- lm(physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c,
data = zautra4)
summary(mod6d)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c,
## data = zautra4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -56.205 -8.728 3.563 11.781 38.357
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 75.691 1.641 46.120 < 2e-16 ***
## panaspos_c_mean_4wk.c 9.429 2.641 3.571 0.000522 ***
## panasneg_c_mean_4wk.c -10.218 2.915 -3.505 0.000653 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.75 on 114 degrees of freedom
## Multiple R-squared: 0.3018, Adjusted R-squared: 0.2895
## F-statistic: 24.63 on 2 and 114 DF, p-value: 1.284e-09mod6d_adjrsq <- summary(mod6d)$adj.r.squared
mod6d_beta1 <- summary(mod6d)$coefficients[2]
mod6d_beta2 <- summary(mod6d)$coefficients[3]Fit regression model:
mod6e <- lm(physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c + panaspos_c_shannon_4wk.c + panasneg_c_shannon_4wk.c,
data = zautra4)
summary(mod6e)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c +
## panaspos_c_shannon_4wk.c + panasneg_c_shannon_4wk.c, data = zautra4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -55.141 -7.729 3.210 11.557 42.881
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 75.691 1.620 46.730 < 2e-16 ***
## panaspos_c_mean_4wk.c 4.966 3.407 1.458 0.14774
## panasneg_c_mean_4wk.c -12.070 4.124 -2.927 0.00414 **
## panaspos_c_shannon_4wk.c 19.138 9.740 1.965 0.05191 .
## panasneg_c_shannon_4wk.c 2.521 3.544 0.711 0.47843
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.52 on 112 degrees of freedom
## Multiple R-squared: 0.3318, Adjusted R-squared: 0.3079
## F-statistic: 13.9 on 4 and 112 DF, p-value: 3.068e-09mod6e_adjrsq <- summary(mod6e)$adj.r.squared
mod6e_beta1 <- summary(mod6e)$coefficients[2]
mod6e_beta2 <- summary(mod6e)$coefficients[3]
mod6e_beta3 <- summary(mod6e)$coefficients[4]
mod6e_beta4 <- summary(mod6e)$coefficients[5]Fit regression model:
mod6f <- lm(physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c + panaspos_c_shannon_4wk.c +
panasneg_c_shannon_4wk.c + I(panaspos_c_mean_4wk.c*panaspos_c_shannon_4wk.c) +
I(panasneg_c_mean_4wk.c*panasneg_c_shannon_4wk.c),
data = zautra4)
summary(mod6f)##
## Call:
## lm(formula = physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c +
## panaspos_c_shannon_4wk.c + panasneg_c_shannon_4wk.c + I(panaspos_c_mean_4wk.c *
## panaspos_c_shannon_4wk.c) + I(panasneg_c_mean_4wk.c * panasneg_c_shannon_4wk.c),
## data = zautra4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -56.893 -7.408 2.158 10.979 40.148
##
## Coefficients:
## Estimate Std. Error
## (Intercept) 77.208 3.652
## panaspos_c_mean_4wk.c 3.257 4.009
## panasneg_c_mean_4wk.c -4.582 8.986
## panaspos_c_shannon_4wk.c 38.204 25.694
## panasneg_c_shannon_4wk.c -3.892 7.625
## I(panaspos_c_mean_4wk.c * panaspos_c_shannon_4wk.c) 9.733 12.885
## I(panasneg_c_mean_4wk.c * panasneg_c_shannon_4wk.c) -7.952 8.702
## t value Pr(>|t|)
## (Intercept) 21.140 <2e-16 ***
## panaspos_c_mean_4wk.c 0.812 0.418
## panasneg_c_mean_4wk.c -0.510 0.611
## panaspos_c_shannon_4wk.c 1.487 0.140
## panasneg_c_shannon_4wk.c -0.510 0.611
## I(panaspos_c_mean_4wk.c * panaspos_c_shannon_4wk.c) 0.755 0.452
## I(panasneg_c_mean_4wk.c * panasneg_c_shannon_4wk.c) -0.914 0.363
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 17.57 on 110 degrees of freedom
## Multiple R-squared: 0.3401, Adjusted R-squared: 0.3041
## F-statistic: 9.447 on 6 and 110 DF, p-value: 2.34e-08mod6f_adjrsq <- summary(mod6f)$adj.r.squared
mod6f_beta1 <- summary(mod6f)$coefficients[2]
mod6f_beta2 <- summary(mod6f)$coefficients[3]
mod6f_beta3 <- summary(mod6f)$coefficients[4]
mod6f_beta4 <- summary(mod6f)$coefficients[5]Model fit comparisons:
anova(mod6d,mod6e); fit1 <- anova(mod6d,mod6e)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c
## Model 2: physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c +
## panaspos_c_shannon_4wk.c + panasneg_c_shannon_4wk.c
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 114 35925
## 2 112 34380 2 1545.4 2.5173 0.08523 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1anova(mod6e,mod6f); fit2 <- anova(mod6e,mod6f)[6][[2,1]]## Analysis of Variance Table
##
## Model 1: physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c +
## panaspos_c_shannon_4wk.c + panasneg_c_shannon_4wk.c
## Model 2: physhealth ~ panaspos_c_mean_4wk.c + panasneg_c_mean_4wk.c +
## panaspos_c_shannon_4wk.c + panasneg_c_shannon_4wk.c + I(panaspos_c_mean_4wk.c *
## panaspos_c_shannon_4wk.c) + I(panasneg_c_mean_4wk.c * panasneg_c_shannon_4wk.c)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 112 34380
## 2 110 33955 2 424.99 0.6884 0.5045Summary and conclusions: This additional set of analyses is given to more closely match what was reported in Benson, Ram, Almeida, Zautra, & Ong, 2017 (with the exception here that the Benson et al., 2017 paper did not report any regression results using the single occassion emotion item data). These results echo what was reported in the model 4 series (d, e, f) and model 5 series (d, e, f). The model fit comparisons suggests that mod6d (just mean positive emotion and mean negative emotion) provides the best/ most parsimonious fit to these data.