BB3 Q of H_Analyses
Effect size difference scores
Convert to long data
#From S21 grad stats Ch 12 slide 32
#Body regulation
dat.long1 <- dat %>%
pivot_longer(
cols = c("BREG_T1","BREG_T2"), # the columns we want to pivot
names_to = ("Time"), # put the column names in a var called day
names_prefix = "BREG_", # trim “con” off the front of names
values_to = "BREG" # put the values in a variable called con
)
#Positive emotions
dat.long3 <- dat %>%
pivot_longer(
cols = c("pEM_T1","pEM_T2"), # the columns we want to pivot
names_to = ("Time"), # put the column names in a var called day
names_prefix = "pEM_", # trim “con” off the front of names
values_to = "pEM" # put the values in a variable called con
)
#Negative emotions
dat.long4 <- dat %>%
pivot_longer(
cols = c("nEM_T1","nEM_T2"), # the columns we want to pivot
names_to = ("Time"), # put the column names in a var called day
names_prefix = "nEM_", # trim “con” off the front of names
values_to = "nEM" # put the values in a variable called con
)
#Emotional regulation
dat.long5 <- dat %>%
pivot_longer(
cols = c("EREG_T1","EREG_T2"), # the columns we want to pivot
names_to = ("Time"), # put the column names in a var called day
names_prefix = "EREG_", # trim “con” off the front of names
values_to = "EREG" # put the values in a variable called con
)
b_longsub1 <- left_join(dat.long1, dat.long3)
b_longsub3<- left_join(b_longsub1, dat.long4)
BB3_long<- left_join(b_longsub3, dat.long5)
longnames <- as.data.frame(colnames(BB3_long))Power
With N = 28, with α = .05, G*power 3.1 calculates that there is a 1-β = .80 power to detect an effect size of Cohen’s dz = .43 (small/medium effect).
Effect size
From: https://imaging.mrc-cbu.cam.ac.uk/statswiki/FAQ/tdunpaired#:~:text=dz%3A%20Standardized%20difference%20scores,as%20the%20denominator%20of%20d.
dz: Standardized difference scores
A third way to compute a d-like effect size is to reduce each subject’s data to a single difference score—the mean difference between their responses in each condition—and then use the standard deviation of these difference scores as the denominator of d. Cohen actually discusses this statistic in his power analysis textbook (Cohen, 1988, p. 48), where he carefully distinguishes it from the classical Cohen’s d by calling it d_z. In R, we can compute this as:
(d_z <- md / sd(sub_means[,2] - sub_means[,1]))
dz = Mean of difference scores / sd of difference scores
small = .2, medium = .5, large ≥ .8
Descriptives
descripts.T1 <- dat[,c("RES.T1_av","BREG_T1", "pEM_T1","nEM_T1","EREG_T1")]
descripts.T2 <- dat[,c("RES.T1_av","BREG_T2", "pEM_T2","nEM_T2","EREG_T2")]T1
| Unique | Missing Pct. | Mean | SD | Min | Median | Max | Histogram | |
|---|---|---|---|---|---|---|---|---|
| RES.T1_av | 11 | 0 | 3.5 | 0.5 | 2.7 | 3.3 | 5.0 | |
| BREG_T1 | 12 | 0 | 3.6 | 0.6 | 2.2 | 3.6 | 5.0 | |
| pEM_T1 | 12 | 0 | 2.9 | 0.7 | 2.0 | 2.8 | 4.7 | |
| nEM_T1 | 13 | 0 | 1.6 | 0.4 | 1.0 | 1.5 | 2.4 | |
| EREG_T1 | 8 | 0 | 3.4 | 0.9 | 1.0 | 3.5 | 5.0 |
T2
| Unique | Missing Pct. | Mean | SD | Min | Median | Max | Histogram | |
|---|---|---|---|---|---|---|---|---|
| RES.T1_av | 11 | 0 | 3.5 | 0.5 | 2.7 | 3.3 | 5.0 | |
| BREG_T2 | 14 | 0 | 3.5 | 0.8 | 1.4 | 3.5 | 4.8 | |
| pEM_T2 | 12 | 0 | 3.5 | 0.7 | 2.5 | 3.5 | 5.0 | |
| nEM_T2 | 6 | 0 | 1.2 | 0.2 | 1.0 | 1.1 | 1.8 | |
| EREG_T2 | 7 | 4 | 4.1 | 0.8 | 1.5 | 4.0 | 5.0 |
Pairwise Correlations
T1
## RES.T1_av BREG_T1 pEM_T1 nEM_T1 EREG_T1
## RES.T1_av 1.00 0.38 0.18 -0.37 0.15
## BREG_T1 0.38 1.00 0.60 -0.21 0.28
## pEM_T1 0.18 0.60 1.00 -0.53 0.02
## nEM_T1 -0.37 -0.21 -0.53 1.00 0.14
## EREG_T1 0.15 0.28 0.02 0.14 1.00
##
## n= 28
##
##
## P
## RES.T1_av BREG_T1 pEM_T1 nEM_T1 EREG_T1
## RES.T1_av 0.0469 0.3512 0.0553 0.4558
## BREG_T1 0.0469 0.0007 0.2823 0.1562
## pEM_T1 0.3512 0.0007 0.0034 0.8999
## nEM_T1 0.0553 0.2823 0.0034 0.4806
## EREG_T1 0.4558 0.1562 0.8999 0.4806
T2
## RES.T1_av BREG_T2 pEM_T2 nEM_T2 EREG_T2
## RES.T1_av 1.00 0.33 0.02 -0.17 0.11
## BREG_T2 0.33 1.00 0.59 -0.32 -0.05
## pEM_T2 0.02 0.59 1.00 -0.35 -0.13
## nEM_T2 -0.17 -0.32 -0.35 1.00 -0.08
## EREG_T2 0.11 -0.05 -0.13 -0.08 1.00
##
## n
## RES.T1_av BREG_T2 pEM_T2 nEM_T2 EREG_T2
## RES.T1_av 28 28 28 28 27
## BREG_T2 28 28 28 28 27
## pEM_T2 28 28 28 28 27
## nEM_T2 28 28 28 28 27
## EREG_T2 27 27 27 27 27
##
## P
## RES.T1_av BREG_T2 pEM_T2 nEM_T2 EREG_T2
## RES.T1_av 0.0890 0.9084 0.3990 0.5993
## BREG_T2 0.0890 0.0010 0.0955 0.8095
## pEM_T2 0.9084 0.0010 0.0679 0.5071
## nEM_T2 0.3990 0.0955 0.0679 0.6964
## EREG_T2 0.5993 0.8095 0.5071 0.6964
Body Regulation (n.s.)
Items
T1 prompt: How often do the following statements apply to you? (1=never, 5 = always)
T2 & 3 prompt: How much do each of the following statements apply to you right now?
BREG1_T1: I notice how my body changes when I feel happy / joyful.
BREG1_T2: I am noticing how my body changes when I feel happy / joyful.
BREG2_T1: When I feel overwhelmed, I can find a calm place inside.
BREG2_T2: I notice that my body feels different while viewing dance.
BREG3_T1: When I bring awareness to my body, I feel a sense of calm.
BREG3_T2: When I bring awareness to my body, I feel a sense of calm.
BREG4_T1: I can use my breath to reduce tension.
BREG4_T2: I can use my breath to reduce tension.
BREG5_T1: When I am caught up in thoughts, I can calm my mind by focusing on my body/breathing.
BREG5_T2: When I am caught up in thoughts, I can calm my mind by focusing on my body/breathing.
Paired t model (n.s.)
##
## Paired t-test
##
## data: dat$BREG_T2 and dat$BREG_T1
## t = -1.179, df = 27, p-value = 0.2487
## alternative hypothesis: true mean difference is not equal to 0
## 95 percent confidence interval:
## -0.3376527 0.0912241
## sample estimates:
## mean difference
## -0.1232143
Conclusion
There is not a significant positive effect of show (with time as proxy) on body regulation (t = 0.52, p = .249).
!Positive Emotions (SIGNIFICANT)!
Items
T1-2 prompt: How much are you feeling each of the following emotions right now? (1 = none at all; 5 = A great deal)
pEM1: amused, fun-loving, silly.
pEM2: content, serene, peaceful.
pEM3: glad, happy, joyful.
pEM4: grateful, appreciative, thankful.
pEM5: hopeful, optimistic, encouraged.
pEM6: love, closeness, trust.
Paired t model
##
## Paired t-test
##
## data: dat$pEM_T2 and dat$pEM_T1
## t = 4.1893, df = 27, p-value = 0.0002677
## alternative hypothesis: true mean difference is not equal to 0
## 95 percent confidence interval:
## 0.2945938 0.8601681
## sample estimates:
## mean difference
## 0.577381
Conclusion
There is a medium/large significant positive effect of show (with time as proxy) on positive emotions, t(27) = 4.19, p < .001, dz = .79. Positive emotions increase throughout the show.
!Negative Emotions (SIGNIFICANT)!
Items
T1-2 prompt: How much are you feeling each of the following emotions right now? (1 = none at all; 5 = A great deal)
nEM1: angry, irritated, annoyed.
nEM2: sad, downhearted, unhappy,
nEM3: scared, fearful, afraid.
nEM4: disgust, distaste, revulsion.
nEM5: repentant, guilty, blameworthy.
nEM6: ashamed, humiliated, disgraced.
nEM7: contemptuous, scornful, disdainful.
nEM8: anxious, nervous, pressured.
Paired t model
##
## Paired t-test
##
## data: dat$nEM_T2 and dat$nEM_T1
## t = -5.0987, df = 27, p-value = 2.337e-05
## alternative hypothesis: true mean difference is not equal to 0
## 95 percent confidence interval:
## -0.5536362 -0.2359046
## sample estimates:
## mean difference
## -0.3947704
Conclusion
There is a large significant negative effect of show (with time as proxy) on negative emotions (t(27) = -5.10, p < .001, dz = 0.96).
Graph
## T1 T2
## 1.555485 1.160714
## T1 T2
## 0.406693 0.209323
p <- ggplot(BB3_long, aes(Time, nEM, fill = Time)) +
geom_violinhalf(position = position_nudge(x = 0.1, y = 0)) +
geom_boxplot(width=0.4, color="black", alpha=0.75) +
geom_jitter(color="black", size=0.4, alpha=0.5)+
coord_flip()+
theme(legend.position = "none",
panel.grid.minor = element_blank(),
panel.border = element_rect(color = "black", fill = NA, size = 1),
axis.text.x = element_text(size = 14,face = "bold"))
p = p + scale_fill_paletteer_d("fishualize::Halichoeres_garnoti") + ylab("") + xlab("")
p! Emotion Regulation
Items
T1
How much do you agree with each statement? (1 = strongly disagree; 5 = strongly agree)
EREG1: I am good at managing my emotions.
EREG2: I am always successful at regulating my emotions when the need arises.
T2 & 3
EREG1: I think I did pretty well at managing my emotions
EREG2: I was successful at regulating my emotions
EREG3: I tried to manage my emotions
Paired t model (significant)
##
## Paired t-test
##
## data: dat$EREG_T2 and dat$EREG_T1
## t = 3.9756, df = 26, p-value = 0.0004981
## alternative hypothesis: true mean difference is not equal to 0
## 95 percent confidence interval:
## 0.3488114 1.0956331
## sample estimates:
## mean difference
## 0.7222222
Conclusion
There is a medium/large significant positive effect of show (with time as proxy) on emotional regulation (t (27) = 3.98, p < .001, dz = 0.77).