ForCarol
Carol
2017-01-30
Correlation Stuff
Correlations - what are they?
Let’s test them for bmi, mvpa3.all, and mvpa4.all. This is not looking at significance, just the correlation estimates.
df %>%
dplyr::select(bmi, mvpa3.all, mvpa4.all) %>%
cor()| bmi | mvpa3.all | mvpa4.all | |
|---|---|---|---|
| bmi | 1.0000000 | -0.3931878 | -0.4541171 |
| mvpa3.all | -0.3931878 | 1.0000000 | 0.8854185 |
| mvpa4.all | -0.4541171 | 0.8854185 | 1.0000000 |
Next, we run a Spearman’s Correlation Test which is showing the significance of the association between paired samples. Essentially, is there correlation that is non-zero between these variables. This does not, however, test whether there is a significant association. Just correlation.
## First, bmi to mets3+
(m1 <- cor.test(df$bmi, df$mvpa3.all, method = 'kendall'))##
## Kendall's rank correlation tau
##
## data: df$bmi and df$mvpa3.all
## z = -2.652, p-value = 0.008001
## alternative hypothesis: true tau is not equal to 0
## sample estimates:
## tau
## -0.2591987## Next, bmi to mets4+
(m2 <- cor.test(df$bmi, df$mvpa4.all, method = "kendall"))##
## Kendall's rank correlation tau
##
## data: df$bmi and df$mvpa4.all
## z = -3.8679, p-value = 0.0001098
## alternative hypothesis: true tau is not equal to 0
## sample estimates:
## tau
## -0.3797835Explanatory Approach-ish
Let’s do some MANCOVA
Why do we want to do this? We want to check if certain variables significantly explain the variance associated with our two dependent variables. Let’s try this.
# Model where we asssess the significance of covariates on the two METS stuffs
mv1 <- manova(cbind(mvpa3.all, mvpa4.all) ~ age + k.pain + k.symptom + k.adl + k.qol + k.sport + intention + bmi, data = df)
summary(mv1)## Df Pillai approx F num Df den Df Pr(>F)
## age 1 0.208315 5.2626 2 40 0.0093551 **
## k.pain 1 0.100041 2.2232 2 40 0.1214666
## k.symptom 1 0.018232 0.3714 2 40 0.6921176
## k.adl 1 0.173803 4.2073 2 40 0.0219622 *
## k.qol 1 0.020334 0.4151 2 40 0.6630731
## k.sport 1 0.039011 0.8119 2 40 0.4511999
## intention 1 0.105263 2.3529 2 40 0.1081214
## bmi 1 0.302392 8.6694 2 40 0.0007451 ***
## Residuals 41
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1There seems to be indication that age, k.adl, and bmi may be significantly associated with the METS outcomes.
Let’s do the same thing, but with attitude, subjective, and perceived instead of intention.
# Model where we asssess the significance of covariates on the two METS stuffs
mv2 <- manova(cbind(mvpa3.all, mvpa4.all) ~ age + k.pain + k.symptom + k.adl + k.qol + k.sport + attitude + subjective + perceived + bmi, data = df)
summary(mv2)## Df Pillai approx F num Df den Df Pr(>F)
## age 1 0.206778 4.9529 2 38 0.012260 *
## k.pain 1 0.100150 2.1146 2 38 0.134659
## k.symptom 1 0.018054 0.3493 2 38 0.707400
## k.adl 1 0.172940 3.9729 2 38 0.027114 *
## k.qol 1 0.021044 0.4084 2 38 0.667577
## k.sport 1 0.038656 0.7640 2 38 0.472820
## attitude 1 0.080374 1.6606 2 38 0.203523
## subjective 1 0.043050 0.8548 2 38 0.433403
## perceived 1 0.030615 0.6001 2 38 0.553900
## bmi 1 0.313151 8.6626 2 38 0.000795 ***
## Residuals 39
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1Let’s do some multiple regression
mr1 <- lm(mvpa4.all ~ age + factor(sex) + k.pain + k.symptom + k.adl + k.qol + k.sport + attitude + subjective + perceived + bmi, data = df)
summary(mr1)##
## Call:
## lm(formula = mvpa4.all ~ age + factor(sex) + k.pain + k.symptom +
## k.adl + k.qol + k.sport + attitude + subjective + perceived +
## bmi, data = df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -48.232 -16.335 -4.115 12.744 84.619
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 259.25991 92.91070 2.790 0.008187 **
## age -2.12250 0.62067 -3.420 0.001511 **
## factor(sex)Male 20.62589 11.92761 1.729 0.091881 .
## k.pain -0.67506 0.56390 -1.197 0.238670
## k.symptom 0.35018 0.43693 0.801 0.427850
## k.adl 1.01517 0.69531 1.460 0.152502
## k.qol -0.56243 0.46728 -1.204 0.236179
## k.sport -0.06168 0.21392 -0.288 0.774657
## attitude -1.69589 7.32357 -0.232 0.818117
## subjective 11.58565 6.06509 1.910 0.063673 .
## perceived -9.18957 5.55544 -1.654 0.106334
## bmi -4.69111 1.10892 -4.230 0.000142 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 30.54 on 38 degrees of freedom
## Multiple R-squared: 0.5177, Adjusted R-squared: 0.378
## F-statistic: 3.708 on 11 and 38 DF, p-value: 0.001227mr2 <- lm(mvpa4.all ~ age + factor(sex) + k.pain + k.symptom + k.adl + k.qol + k.sport + intention + bmi, data = df)
summary(mr2)##
## Call:
## lm(formula = mvpa4.all ~ age + factor(sex) + k.pain + k.symptom +
## k.adl + k.qol + k.sport + intention + bmi, data = df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -38.049 -16.487 -7.462 4.004 85.875
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 182.70435 74.33761 2.458 0.018410 *
## age -1.86275 0.56713 -3.285 0.002129 **
## factor(sex)Male 13.12889 12.19554 1.077 0.288139
## k.pain -0.52116 0.57291 -0.910 0.368444
## k.symptom 0.10971 0.41199 0.266 0.791375
## k.adl 0.55355 0.70041 0.790 0.434003
## k.qol -0.48625 0.46651 -1.042 0.303527
## k.sport 0.05837 0.19795 0.295 0.769616
## intention 11.84863 8.17433 1.449 0.154996
## bmi -3.70186 0.91849 -4.030 0.000243 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 30.8 on 40 degrees of freedom
## Multiple R-squared: 0.4834, Adjusted R-squared: 0.3671
## F-statistic: 4.159 on 9 and 40 DF, p-value: 0.0007702mr1 <- lm(mvpa4.all ~ age + sex + k.pain + k.symptom + k.adl + k.qol + k.sport + attitude + subjective + perceived + bmi, data = df)
summary(mr1)##
## Call:
## lm(formula = mvpa4.all ~ age + sex + k.pain + k.symptom + k.adl +
## k.qol + k.sport + attitude + subjective + perceived + bmi,
## data = df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -48.232 -16.335 -4.115 12.744 84.619
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 259.25991 92.91070 2.790 0.008187 **
## age -2.12250 0.62067 -3.420 0.001511 **
## sexMale 20.62589 11.92761 1.729 0.091881 .
## k.pain -0.67506 0.56390 -1.197 0.238670
## k.symptom 0.35018 0.43693 0.801 0.427850
## k.adl 1.01517 0.69531 1.460 0.152502
## k.qol -0.56243 0.46728 -1.204 0.236179
## k.sport -0.06168 0.21392 -0.288 0.774657
## attitude -1.69589 7.32357 -0.232 0.818117
## subjective 11.58565 6.06509 1.910 0.063673 .
## perceived -9.18957 5.55544 -1.654 0.106334
## bmi -4.69111 1.10892 -4.230 0.000142 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 30.54 on 38 degrees of freedom
## Multiple R-squared: 0.5177, Adjusted R-squared: 0.378
## F-statistic: 3.708 on 11 and 38 DF, p-value: 0.001227Predictive Approach
Let’s try a stepwise regression approach, optimizing for AIC
METS 3+
df3 <- df %>%
dplyr::select(age,
sex,
comorb,
k.pain,
k.symptom,
k.adl,
k.sport,
k.qol,
attitude,
subjective,
perceived,
bmi,
mvpa3.all)
t3 <- lm(mvpa3.all ~ ., data = df3)
step <- stepAIC(t3, direction="both")## Start: AIC=395.85
## mvpa3.all ~ age + sex + comorb + k.pain + k.symptom + k.adl +
## k.sport + k.qol + attitude + subjective + perceived + bmi
##
## Df Sum of Sq RSS AIC
## - k.sport 1 131 81682 393.93
## - comorb 1 265 81815 394.01
## - k.symptom 1 690 82240 394.27
## - attitude 1 1197 82747 394.58
## - subjective 1 2074 83624 395.10
## - perceived 1 2434 83984 395.32
## <none> 81550 395.85
## - k.pain 1 5012 86562 396.83
## - k.adl 1 6641 88192 397.76
## - sex 1 7132 88682 398.04
## - k.qol 1 7221 88772 398.09
## - age 1 30859 112409 409.89
## - bmi 1 34546 116096 411.51
##
## Step: AIC=393.93
## mvpa3.all ~ age + sex + comorb + k.pain + k.symptom + k.adl +
## k.qol + attitude + subjective + perceived + bmi
##
## Df Sum of Sq RSS AIC
## - comorb 1 250 81931 392.08
## - k.symptom 1 884 82566 392.47
## - attitude 1 1120 82801 392.61
## - subjective 1 2503 84185 393.44
## <none> 81682 393.93
## - perceived 1 3592 85273 394.08
## - k.pain 1 6171 87852 395.57
## + k.sport 1 131 81550 395.85
## - sex 1 7034 88716 396.06
## - k.qol 1 7120 88802 396.11
## - k.adl 1 7735 89417 396.45
## - age 1 31811 113492 408.37
## - bmi 1 36654 118336 410.46
##
## Step: AIC=392.08
## mvpa3.all ~ age + sex + k.pain + k.symptom + k.adl + k.qol +
## attitude + subjective + perceived + bmi
##
## Df Sum of Sq RSS AIC
## - k.symptom 1 885 82816 390.62
## - attitude 1 971 82902 390.67
## - subjective 1 2307 84238 391.47
## <none> 81931 392.08
## - perceived 1 3856 85787 392.38
## - k.pain 1 6401 88332 393.84
## + comorb 1 250 81682 393.93
## + k.sport 1 116 81815 394.01
## - sex 1 6868 88799 394.11
## - k.qol 1 7079 89010 394.22
## - k.adl 1 7917 89848 394.69
## - age 1 33274 115205 407.12
## - bmi 1 39939 121870 409.93
##
## Step: AIC=390.62
## mvpa3.all ~ age + sex + k.pain + k.adl + k.qol + attitude + subjective +
## perceived + bmi
##
## Df Sum of Sq RSS AIC
## - attitude 1 717 83534 389.05
## - subjective 1 1981 84797 389.80
## - perceived 1 3234 86051 390.53
## <none> 82816 390.62
## - k.pain 1 5529 88346 391.85
## + k.symptom 1 885 81931 392.08
## - k.qol 1 6196 89012 392.23
## + k.sport 1 302 82515 392.44
## + comorb 1 251 82566 392.47
## - sex 1 6756 89572 392.54
## - k.adl 1 7513 90330 392.96
## - age 1 32446 115262 405.15
## - bmi 1 40309 123125 408.45
##
## Step: AIC=389.05
## mvpa3.all ~ age + sex + k.pain + k.adl + k.qol + subjective +
## perceived + bmi
##
## Df Sum of Sq RSS AIC
## - subjective 1 2112 85646 388.30
## <none> 83534 389.05
## - perceived 1 4632 88165 389.75
## - k.pain 1 5512 89045 390.24
## - k.qol 1 5865 89399 390.44
## + attitude 1 717 82816 390.62
## + k.symptom 1 631 82902 390.67
## - sex 1 6717 90251 390.92
## + k.sport 1 174 83359 390.94
## + comorb 1 116 83417 390.98
## - k.adl 1 7361 90895 391.27
## - age 1 33512 117045 403.91
## - bmi 1 47463 130997 409.55
##
## Step: AIC=388.3
## mvpa3.all ~ age + sex + k.pain + k.adl + k.qol + perceived +
## bmi
##
## Df Sum of Sq RSS AIC
## - perceived 1 3028 88674 388.03
## <none> 85646 388.30
## + subjective 1 2112 83534 389.05
## - sex 1 5535 91180 389.43
## - k.qol 1 5878 91524 389.62
## + attitude 1 848 84797 389.80
## - k.pain 1 6381 92027 389.89
## + k.sport 1 528 85117 389.99
## + k.symptom 1 344 85302 390.10
## + comorb 1 6 85640 390.29
## - k.adl 1 7340 92985 390.41
## - age 1 31407 117053 401.92
## - bmi 1 45455 131101 407.58
##
## Step: AIC=388.03
## mvpa3.all ~ age + sex + k.pain + k.adl + k.qol + bmi
##
## Df Sum of Sq RSS AIC
## <none> 88674 388.03
## + perceived 1 3028 85646 388.30
## - sex 1 4560 93233 388.54
## + attitude 1 1973 86701 388.91
## - k.qol 1 5649 94323 389.12
## + k.sport 1 1479 87195 389.19
## - k.adl 1 6505 95179 389.57
## + subjective 1 508 88165 389.75
## + comorb 1 96 88577 389.98
## + k.symptom 1 20 88653 390.02
## - k.pain 1 8495 97169 390.61
## - age 1 31455 120128 401.21
## - bmi 1 43398 132072 405.95step$anova # display results| Step | Df | Deviance | Resid. Df | Resid. Dev | AIC |
|---|---|---|---|---|---|
| NA | NA | 37 | 81550.21 | 395.8476 | |
| - k.sport | 1 | 131.4061 | 38 | 81681.61 | 393.9281 |
| - comorb | 1 | 249.7261 | 39 | 81931.34 | 392.0807 |
| - k.symptom | 1 | 884.9722 | 40 | 82816.31 | 390.6179 |
| - attitude | 1 | 717.1921 | 41 | 83533.50 | 389.0490 |
| - subjective | 1 | 2112.1999 | 42 | 85645.70 | 388.2976 |
| - perceived | 1 | 3027.8388 | 43 | 88673.54 | 388.0347 |
sm3 <- lm(mvpa3.all ~ age + sex + k.pain + k.adl + k.qol + bmi, data = df3)
summary(sm3)##
## Call:
## lm(formula = mvpa3.all ~ age + sex + k.pain + k.adl + k.qol +
## bmi, data = df3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -70.750 -29.333 3.017 23.061 143.592
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 457.5734 88.6834 5.160 5.99e-06 ***
## age -3.1773 0.8135 -3.906 0.000327 ***
## sexMale 25.1008 16.8800 1.487 0.144307
## k.pain -1.4874 0.7328 -2.030 0.048607 *
## k.adl 1.7515 0.9862 1.776 0.082790 .
## k.qol -1.0564 0.6383 -1.655 0.105187
## bmi -5.8919 1.2843 -4.587 3.86e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 45.41 on 43 degrees of freedom
## Multiple R-squared: 0.5205, Adjusted R-squared: 0.4536
## F-statistic: 7.78 on 6 and 43 DF, p-value: 1.064e-05lm.1.ef <- sjp.lm(sm3, type = "dots", y.offset = .4,
geom.colors = color, prnt.plot = FALSE,
axis.labels = c("a, b, c, d, e, f"))
lm.1.ef$plot +
theme(strip.text.x = element_text(size = 10),
plot.title = element_text(family = "OpenSans-CondensedBold", size = 12),
plot.subtitle=element_text(family="OpenSans-CondensedLightItalic"),
panel.grid.major.y=element_line(color="#2b2b2b", linetype="dotted", size=0.15),
plot.margin=unit(rep(0.5, 4), "cm")) +
ggtitle('Linear Model Estimates: METs 3+',
subtitle = "mvpa3+ ~ age + sex + KOOS.pain + KOOS.adl + KOOS.qol + BMI")
METS 4+
df4 <- df %>%
dplyr::select(age,
sex,
comorb,
k.pain,
k.symptom,
k.adl,
k.sport,
k.qol,
attitude,
subjective,
perceived,
bmi,
mvpa4.all)
t4 <- lm(mvpa4.all ~ ., data = df4)
step <- stepAIC(t4, direction="both")## Start: AIC=352.43
## mvpa4.all ~ age + sex + comorb + k.pain + k.symptom + k.adl +
## k.sport + k.qol + attitude + subjective + perceived + bmi
##
## Df Sum of Sq RSS AIC
## - k.sport 1 53.4 34275 350.51
## - attitude 1 175.2 34397 350.68
## - k.symptom 1 580.3 34802 351.27
## - k.pain 1 1116.4 35338 352.03
## - comorb 1 1209.1 35431 352.17
## <none> 34222 352.43
## - k.qol 1 1413.4 35635 352.45
## - k.adl 1 1781.0 36003 352.97
## - perceived 1 2114.9 36337 353.43
## - sex 1 3685.0 37907 355.54
## - subjective 1 3970.8 38192 355.92
## - age 1 12110.0 46332 365.58
## - bmi 1 17713.2 51935 371.29
##
## Step: AIC=350.51
## mvpa4.all ~ age + sex + comorb + k.pain + k.symptom + k.adl +
## k.qol + attitude + subjective + perceived + bmi
##
## Df Sum of Sq RSS AIC
## - attitude 1 201.9 34477 348.80
## - k.symptom 1 530.1 34805 349.27
## - k.pain 1 1075.5 35351 350.05
## - comorb 1 1233.2 35508 350.27
## <none> 34275 350.51
## - k.qol 1 1587.3 35862 350.77
## - k.adl 1 1746.6 36022 350.99
## - perceived 1 2246.0 36521 351.68
## + k.sport 1 53.4 34222 352.43
## - subjective 1 3995.3 38270 354.02
## - sex 1 4034.3 38309 354.07
## - age 1 12092.8 46368 363.62
## - bmi 1 17987.0 52262 369.60
##
## Step: AIC=348.8
## mvpa4.all ~ age + sex + comorb + k.pain + k.symptom + k.adl +
## k.qol + subjective + perceived + bmi
##
## Df Sum of Sq RSS AIC
## - k.symptom 1 443.7 34921 347.44
## - k.pain 1 1039.0 35516 348.29
## - comorb 1 1097.3 35574 348.37
## <none> 34477 348.80
## - k.qol 1 1461.9 35939 348.88
## - k.adl 1 1704.7 36182 349.21
## + attitude 1 201.9 34275 350.51
## + k.sport 1 80.2 34397 350.68
## - perceived 1 2823.3 37300 350.74
## - sex 1 3936.3 38413 352.21
## - subjective 1 3992.9 38470 352.28
## - age 1 13030.0 47507 362.83
## - bmi 1 22884.1 57361 372.25
##
## Step: AIC=347.44
## mvpa4.all ~ age + sex + comorb + k.pain + k.adl + k.qol + subjective +
## perceived + bmi
##
## Df Sum of Sq RSS AIC
## - k.pain 1 720.7 35641 346.46
## - k.qol 1 1103.1 36024 347.00
## - comorb 1 1139.8 36060 347.05
## <none> 34921 347.44
## - k.adl 1 1572.9 36494 347.64
## - perceived 1 2422.4 37343 348.79
## + k.symptom 1 443.7 34477 348.80
## + attitude 1 115.6 34805 349.27
## + k.sport 1 11.5 34909 349.42
## - subjective 1 3716.0 38637 350.50
## - sex 1 3904.1 38825 350.74
## - age 1 12796.6 47717 361.05
## - bmi 1 22753.4 57674 370.53
##
## Step: AIC=346.46
## mvpa4.all ~ age + sex + comorb + k.adl + k.qol + subjective +
## perceived + bmi
##
## Df Sum of Sq RSS AIC
## - k.adl 1 855.0 36496 345.65
## - k.qol 1 1277.1 36918 346.22
## - comorb 1 1298.6 36940 346.25
## <none> 35641 346.46
## + k.pain 1 720.7 34921 347.44
## + k.symptom 1 125.4 35516 348.29
## + attitude 1 121.6 35520 348.29
## + k.sport 1 12.3 35629 348.44
## - perceived 1 3158.9 38800 348.71
## - subjective 1 4197.9 39839 350.03
## - sex 1 4330.1 39971 350.19
## - age 1 14245.4 49887 361.27
## - bmi 1 27998.5 63640 373.45
##
## Step: AIC=345.65
## mvpa4.all ~ age + sex + comorb + k.qol + subjective + perceived +
## bmi
##
## Df Sum of Sq RSS AIC
## - k.qol 1 429.4 36926 344.23
## - comorb 1 1326.2 37823 345.43
## <none> 36496 345.65
## + k.adl 1 855.0 35641 346.46
## - perceived 1 2456.3 38953 346.90
## + k.symptom 1 253.8 36243 347.30
## + attitude 1 96.9 36399 347.51
## + k.sport 1 45.3 36451 347.59
## + k.pain 1 2.8 36494 347.64
## - subjective 1 3882.4 40379 348.70
## - sex 1 4179.6 40676 349.07
## - age 1 18660.9 55157 364.30
## - bmi 1 28080.5 64577 372.18
##
## Step: AIC=344.23
## mvpa4.all ~ age + sex + comorb + subjective + perceived + bmi
##
## Df Sum of Sq RSS AIC
## - comorb 1 1317.1 38243 343.98
## <none> 36926 344.23
## + k.qol 1 429.4 36496 345.65
## + k.pain 1 224.6 36701 345.93
## + attitude 1 60.7 36865 346.15
## + k.sport 1 12.4 36913 346.22
## + k.adl 1 7.3 36918 346.22
## + k.symptom 1 0.1 36926 346.23
## - perceived 1 3885.0 40811 347.23
## - subjective 1 4315.2 41241 347.76
## - sex 1 4351.1 41277 347.80
## - age 1 18825.0 55751 362.83
## - bmi 1 27944.6 64870 370.41
##
## Step: AIC=343.98
## mvpa4.all ~ age + sex + subjective + perceived + bmi
##
## Df Sum of Sq RSS AIC
## <none> 38243 343.98
## + comorb 1 1317.1 36926 344.23
## + k.qol 1 420.2 37823 345.43
## + k.pain 1 265.1 37978 345.64
## + k.adl 1 10.1 38233 345.97
## + k.sport 1 9.9 38233 345.97
## + attitude 1 1.5 38241 345.98
## + k.symptom 1 0.4 38242 345.98
## - sex 1 3373.8 41617 346.21
## - subjective 1 3630.0 41873 346.52
## - perceived 1 4316.6 42559 347.33
## - age 1 17508.2 55751 360.83
## - bmi 1 27212.0 65455 368.85step$anova # display results| Step | Df | Deviance | Resid. Df | Resid. Dev | AIC |
|---|---|---|---|---|---|
| NA | NA | 37 | 34221.61 | 352.4295 | |
| - k.sport | 1 | 53.39329 | 38 | 34275.00 | 350.5074 |
| - attitude | 1 | 201.92885 | 39 | 34476.93 | 348.8011 |
| - k.symptom | 1 | 443.74241 | 40 | 34920.68 | 347.4406 |
| - k.pain | 1 | 720.68802 | 41 | 35641.36 | 346.4620 |
| - k.adl | 1 | 855.04468 | 42 | 36496.41 | 345.6473 |
| - k.qol | 1 | 429.36101 | 43 | 36925.77 | 344.2321 |
| - comorb | 1 | 1317.06350 | 44 | 38242.83 | 343.9844 |
sm4 <- lm(mvpa4.all ~ age + sex + subjective + perceived + bmi, data = df)
summary(sm4)##
## Call:
## lm(formula = mvpa4.all ~ age + sex + subjective + perceived +
## bmi, data = df)
##
## Residuals:
## Min 1Q Median 3Q Max
## -51.086 -16.338 -5.065 9.688 89.835
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 272.9046 52.4266 5.205 4.88e-06 ***
## age -2.2864 0.5094 -4.488 5.12e-05 ***
## sexMale 22.0673 11.2005 1.970 0.0551 .
## subjective 11.2010 5.4809 2.044 0.0470 *
## perceived -8.6078 3.8625 -2.229 0.0310 *
## bmi -4.4121 0.7885 -5.595 1.32e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 29.48 on 44 degrees of freedom
## Multiple R-squared: 0.4794, Adjusted R-squared: 0.4202
## F-statistic: 8.103 on 5 and 44 DF, p-value: 1.741e-05lm.2.ef <- sjp.lm(sm4, type = "dots", y.offset = .4,
geom.colors = color, prnt.plot = FALSE,
axis.labels = c("a, b, c, d, e, f"))
lm.2.ef$plot +
theme(strip.text.x = element_text(size = 10),
plot.title = element_text(family = "OpenSans-CondensedBold", size = 12),
plot.subtitle=element_text(family="OpenSans-CondensedLightItalic"),
panel.grid.major.y=element_line(color="#2b2b2b", linetype="dotted", size=0.15),
plot.margin=unit(rep(0.5, 4), "cm")) +
ggtitle('Linear Model Estimates: METs 4+',
subtitle = "mvpa3+ ~ sex + SubjectiveNorm + age + BMI + PBC ")
tab <- sjt.lm(sm3,sm4, group.pred = FALSE, no.output=TRUE, CSS = css.table,
show.aic = TRUE,
show.ci = TRUE,
show.dev = TRUE,
show.re.var = TRUE,
depvar.labels = c("METs 3+", "METs 4+"),
show.header=TRUE
)$knitr| Predictors | Dependent Variables | METs 3+ | METs 4+ | |||||
| B | CI | p | B | CI | p | |||
| (Intercept) | 457.57 | 278.73 – 636.42 | <.001 | 272.90 | 167.25 – 378.56 | <.001 | ||
| age | -3.18 | -4.82 – -1.54 | <.001 | -2.29 | -3.31 – -1.26 | <.001 | ||
| sex (Male) | 25.10 | -8.94 – 59.14 | .144 | 22.07 | -0.51 – 44.64 | .055 | ||
| k.pain | -1.49 | -2.97 – -0.01 | .049 | |||||
| k.adl | 1.75 | -0.24 – 3.74 | .083 | |||||
| k.qol | -1.06 | -2.34 – 0.23 | .105 | |||||
| bmi | -5.89 | -8.48 – -3.30 | <.001 | -4.41 | -6.00 – -2.82 | <.001 | ||
| subjective | 11.20 | 0.15 – 22.25 | .047 | |||||
| perceived | -8.61 | -16.39 – -0.82 | .031 | |||||
| Observations | 50 | 50 | ||||||
| R2 / adj. R2 | .521 / .454 | .479 / .420 | ||||||
| AIC | 531.929 | 487.878 | ||||||
| Deviance | 88673.542 | 38242.833 | ||||||