Skip to content

Latest commit

 

History

History
683 lines (534 loc) · 20.1 KB

Main_Paxton_Dale-all_analyses.md

File metadata and controls

683 lines (534 loc) · 20.1 KB
Raw

Data Analysis for Main, Paxton, & Dale (accepted, Emotion)

Interest/Validation

setwd('/Main_Paxton_Dale-Analyses')

# preliminaries 
rm(list=ls())
getwd()
## [1] "/Main_Paxton_Dale-Analyses"

source('globals_functions.R')
## Loading required package: Matrix
## Loading required package: tseriesChaos
## Loading required package: deSolve
## 
## Attaching package: 'deSolve'
## 
## The following object is masked from 'package:graphics':
## 
##     matplot
## 
## Loading required package: fields
## Loading required package: spam
## Loading required package: grid
## Spam version 1.3-0 (2015-10-24) is loaded.
## Type 'help( Spam)' or 'demo( spam)' for a short introduction 
## and overview of this package.
## Help for individual functions is also obtained by adding the
## suffix '.spam' to the function name, e.g. 'help( chol.spam)'.
## 
## Attaching package: 'spam'
## 
## The following objects are masked from 'package:base':
## 
##     backsolve, forwardsolve
## 
## Loading required package: maps
## 
##  # ATTENTION: maps v3.0 has an updated 'world' map.        #
##  # Many country borders and names have changed since 1990. #
##  # Type '?world' or 'news(package="maps")'. See README_v3. #
## 
## 
## Loading required package: plot3D
## Loading required package: pracma
## 
## Attaching package: 'pracma'
## 
## The following object is masked from 'package:deSolve':
## 
##     rk4
## 
## The following objects are masked from 'package:Matrix':
## 
##     expm, lu, tril, triu

# specify parents' and adolescents' affect
a_target_emotion = c(int,val)
p_target_emotion = c(int,val)

# run RQA
source('get_rqa_measures.R')
## [1] 2
## [1] 3
## [1] 4
## [1] 5
## [1] 6
## [1] 7
## [1] 8
## [1] 9
## [1] 10
## [1] 11
## [1] 12
## [1] 13
## [1] 14
## [1] 15
## [1] 16
## [1] 17
## [1] 18
## [1] 19
## [1] 20
## [1] 21
## [1] 22
## [1] 23
## [1] 24
## [1] 25
## [1] 26
## [1] 27
## [1] 28
## [1] 29
## [1] 30
## [1] 31
## [1] 32
## [1] 33
## [1] 34
## [1] 35
## [1] 36
## [1] 37
## [1] 38
## [1] 39
## [1] 40
## [1] 41
## [1] 42
## [1] 43
## [1] 44
## [1] 45
## [1] 46
## [1] 47
## [1] 48
## [1] 49
## [1] 50

Let's look at the location of maximum lags. If one or the other is leading, we will find maximum lag location to be more on the left (parent leading) or right (adolescent leading). In general we do not find this imbalance: A one-sample t-test with maximum observed lag as the dependent variable is not significant less or greater than 0. In fact, it is centered around 0 (see plots); interestingly, in the interest/validation case, the maximum lags do show a bimodality, suggesting the turn taking pattern holds even in the maximum lag value.

# only include those for which a maximum RR is observed
descriptives = descriptives[descriptives$maxrec>0,]

# print it out
pander(t.test(descriptives[descriptives$shuff=='observed',]$maxlag),style="rmarkdown")
Test statistic df P value Alternative hypothesis
-0.5962 33 0.5551 two.sided

Table: One Sample t-test: descriptives[descriptives$shuff == "observed", ]$maxlag

# plot interaction with satisfaction -- use PDF if we want to save as a file in a subdirectory
# pdf(file='/Main_Paxton_Dale-Analyses/plots/interestvalidation_satisfaction.pdf',width=8,height=5.25)
source('plot_drps_satisfaction.R')

# dev.off()

# print model results using standardized and unstandardized coefficients
source('lmer_stats.R')
pander(coefs.simple,style="rmarkdown") # simple (just lag terms): standardized
  Estimate Std..Error t.value p
(Intercept) -2.094e-12 0.1349 -1.552e-11 1
LinearLag -0.02228 0.02984 -0.7466 0.4553
QuadLag 0.07685 0.03048 2.521 0.01169 
pander(coefs.simple.raw,style="rmarkdown") # simple (just lag terms): unstandardized
  Estimate Std..Error t.value p
(Intercept) 0.01368 0.003341 4.095 4.214e-05
LinearLag -0.004309 0.005772 -0.7466 0.4553
QuadLag 0.01486 0.005895 2.521 0.01169 
pander(coefs.satisfaction,style="rmarkdown") # lag terms and satisfaction: standardized
  Estimate Std..Error t.value p
(Intercept) -1.065e-12 0.1246 -8.547e-12 1
SatisfactionF 0.3819 0.1226 3.116 0.001833
LinearLag 0.04341 0.1313 0.3305 0.741
QuadLag -0.08813 0.06891 -1.279 0.2009
SatisXLinear -0.06733 0.131 -0.5138 0.6074
SatisXQuad 0.1696 0.08094 2.095 0.03615 
pander(coefs.satisfaction.raw,style="rmarkdown") # lag terms and satisfaction: unstandardized
  Estimate Std..Error t.value
(Intercept) -0.02837 0.01385 -2.049
SatisfactionF 0.01062 0.003407 3.116
LinearLag 0.008396 0.0254 0.3305
QuadLag -0.01704 0.01333 -1.279
SatisfactionF:LinearLag -0.003207 0.006242 -0.5138
SatisfactionF:QuadLag 0.008078 0.003855 2.095

Table: Table continues below

  p
(Intercept) 0.04044
SatisfactionF 0.001833
LinearLag 0.741
QuadLag 0.2009
SatisfactionF:LinearLag 0.6074
SatisfactionF:QuadLag 0.03615 
pander(coefs.age,style="rmarkdown") # lag terms and age: standardized
  Estimate Std..Error t.value p
(Intercept) -1.297e-12 0.124 -1.046e-11 1
AgeF 0.389 0.124 3.138 0.001701
LinearLag 0.2818 0.2089 1.349 0.1773
QuadLag -0.1678 0.2161 -0.7765 0.4374
AgeXLinear -0.3069 0.2161 -1.42 0.1556
AgeXQuad 0.2469 0.2236 1.104 0.2695 
pander(coefs.age.raw,style="rmarkdown") # lag terms and age: unstandardized
  Estimate Std..Error t.value p
(Intercept) -0.05825 0.02313 -2.519 0.01178
AgeF 0.004842 0.001543 3.138 0.001701
LinearLag 0.05451 0.04041 1.349 0.1773
QuadLag -0.03246 0.04181 -0.7765 0.4374
AgeF:LinearLag -0.003959 0.002789 -1.42 0.1556
AgeF:QuadLag 0.003186 0.002885 1.104 0.2695

Negative Emotion

# preliminaries
rm(list=ls())
source('globals_functions.R')

# set the target emotion for these analyses and then run RQA
a_target_emotion = neg
p_target_emotion = neg
source('get_rqa_measures.R')
## [1] 2
## [1] 3
## [1] 4
## [1] 5
## [1] 6
## [1] 7
## [1] 8
## [1] 9
## [1] 10
## [1] 11
## [1] 12
## [1] 13
## [1] 14
## [1] 15
## [1] 16
## [1] 17
## [1] 18
## [1] 19
## [1] 20
## [1] 21
## [1] 22
## [1] 23
## [1] 24
## [1] 25
## [1] 26
## [1] 27
## [1] 28
## [1] 29
## [1] 30
## [1] 31
## [1] 32
## [1] 33
## [1] 34
## [1] 35
## [1] 36
## [1] 37
## [1] 38
## [1] 39
## [1] 40
## [1] 41
## [1] 42
## [1] 43
## [1] 44
## [1] 45
## [1] 46
## [1] 47
## [1] 48
## [1] 49
## [1] 50

# only include those for which a maximum RR is observed
descriptives = descriptives[descriptives$maxrec>0,]

# print it out
pander(t.test(descriptives[descriptives$shuff=='observed',]$maxlag),style="rmarkdown")
Test statistic df P value Alternative hypothesis
-0.4161 39 0.6796 two.sided

Table: One Sample t-test: descriptives[descriptives$shuff == "observed", ]$maxlag

# plot interaction with satisfaction -- use PDF if we want to save as a file in a subdirectory
# pdf(file='/Main_Paxton_Dale-Analyses/plots/negative_satisfaction.pdf',width=8,height=5.25)
source('plot_drps_satisfaction.R')

# dev.off()

# plot interaction with age -- use PDF if we want to save as a file in a subdirectory
# pdf(file='/Main_Paxton_Dale-Analyses/plots/negative_age.pdf',width=8,height=5.25)
source('plot_drps_age.R')

# dev.off()

# print model results using standardized and unstandardized ("raw") coefficients
source('lmer_stats.R')
pander(coefs.simple,style="rmarkdown") # simple (just lag terms): standardized
  Estimate Std..Error t.value p
(Intercept) -5.873e-12 0.1432 -4.1e-11 1
LinearLag -0.01346 0.01003 -1.342 0.1797
QuadLag -0.02796 0.008803 -3.176 0.001494 
pander(coefs.simple.raw,style="rmarkdown") # simple (just lag terms): unstandardized
  Estimate Std..Error t.value p
(Intercept) 0.0958 0.01809 5.297 1.18e-07
LinearLag -0.01328 0.009894 -1.342 0.1797
QuadLag -0.02757 0.00868 -3.176 0.001494 
pander(coefs.satisfaction,style="rmarkdown") # lag terms and satisfaction: standardized
  Estimate Std..Error t.value p
(Intercept) -9.869e-12 0.1324 -7.457e-11 1
SatisfactionF -0.4053 0.1324 -3.063 0.002195
LinearLag 0.00563 0.04612 0.1221 0.9028
QuadLag -0.04619 0.02463 -1.875 0.06074
SatisXLinear -0.01957 0.04612 -0.4243 0.6713
SatisXQuad 0.01752 0.0287 0.6104 0.5416 
pander(coefs.satisfaction.raw,style="rmarkdown") # lag terms and satisfaction: unstandardized
  Estimate Std..Error t.value
(Intercept) 0.3234 0.07617 4.246
SatisfactionF -0.05745 0.01876 -3.063
LinearLag 0.005551 0.04548 0.1221
QuadLag -0.04554 0.02428 -1.875
SatisfactionF:LinearLag -0.004752 0.0112 -0.4243
SatisfactionF:QuadLag 0.004254 0.006969 0.6104

Table: Table continues below

  p
(Intercept) 2.179e-05
SatisfactionF 0.002195
LinearLag 0.9028
QuadLag 0.06074
SatisfactionF:LinearLag 0.6713
SatisfactionF:QuadLag 0.5416 
pander(coefs.age,style="rmarkdown") # lag terms and age: standardized
  Estimate Std..Error t.value p
(Intercept) -5.049e-12 0.1438 -3.511e-11 1
AgeF -0.1152 0.1438 -0.8007 0.4233
LinearLag -0.1639 0.07314 -2.241 0.02502
QuadLag -0.06263 0.06878 -0.9107 0.3625
AgeXLinear 0.1518 0.07314 2.075 0.03795
AgeXQuad 0.03492 0.07113 0.4909 0.6235 
pander(coefs.age.raw,style="rmarkdown") # lag terms and age: unstandardized
  Estimate Std..Error t.value p
(Intercept) 0.2043 0.1368 1.494 0.1351
AgeF -0.007306 0.009124 -0.8007 0.4233
LinearLag -0.1616 0.07211 -2.241 0.02502
QuadLag -0.06175 0.06781 -0.9106 0.3625
AgeF:LinearLag 0.009985 0.004811 2.075 0.03795
AgeF:QuadLag 0.002297 0.004679 0.4909 0.6235

Positive Emotion

# preliminaries
rm(list=ls())
source('globals_functions.R')

# set the target emotion for these analyses and then run RQA
a_target_emotion = pos
p_target_emotion = pos
source('get_rqa_measures.R')
## [1] 2
## [1] 3
## [1] 4
## [1] 5
## [1] 6
## [1] 7
## [1] 8
## [1] 9
## [1] 10
## [1] 11
## [1] 12
## [1] 13
## [1] 14
## [1] 15
## [1] 16
## [1] 17
## [1] 18
## [1] 19
## [1] 20
## [1] 21
## [1] 22
## [1] 23
## [1] 24
## [1] 25
## [1] 26
## [1] 27
## [1] 28
## [1] 29
## [1] 30
## [1] 31
## [1] 32
## [1] 33
## [1] 34
## [1] 35
## [1] 36
## [1] 37
## [1] 38
## [1] 39
## [1] 40
## [1] 41
## [1] 42
## [1] 43
## [1] 44
## [1] 45
## [1] 46
## [1] 47
## [1] 48
## [1] 49
## [1] 50

# only include those for which a maximum RR is observed
descriptives = descriptives[descriptives$maxrec>0,]

# print it out
pander(t.test(descriptives[descriptives$shuff=='observed',]$maxlag),style="rmarkdown")
Test statistic df P value Alternative hypothesis
-1.369 38 0.1791 two.sided

Table: One Sample t-test: descriptives[descriptives$shuff == "observed", ]$maxlag

# plot interaction with satisfaction -- use PDF if we want to save as a file in a subdirectory
# pdf(file='/Main_Paxton_Dale-Analyses/plots/positive_satisfaction.pdf',width=8,height=5.25)
source('plot_drps_satisfaction.R')

# dev.off()

# plot interaction with age -- use PDF if we want to save as a file in a subdirectory
# pdf(file='/Main_Paxton_Dale-Analyses/plots/positive_age.pdf',width=8,height=5.25)
source('plot_drps_age.R')

# dev.off()

# print model results using standardized and unstandardized ("raw") coefficients
source('lmer_stats.R')
pander(coefs.simple,style="rmarkdown") # simple (just lag terms): standardized
  Estimate Std..Error t.value p
(Intercept) 2.632e-13 0.1225 2.148e-12 1
LinearLag 0.007195 0.01379 0.5217 0.6018
QuadLag -0.196 0.04619 -4.243 2.201e-05 
pander(coefs.simple.raw,style="rmarkdown") # simple (just lag terms): unstandardized
  Estimate Std..Error t.value p
(Intercept) 0.009871 0.00273 3.616 0.0002991
LinearLag 0.001251 0.002399 0.5217 0.6018
QuadLag -0.03409 0.008035 -4.243 2.201e-05 
pander(coefs.satisfaction,style="rmarkdown") # lag terms and satisfaction: standardized
  Estimate Std..Error t.value p
(Intercept) 2.803e-13 0.1213 2.31e-12 1
SatisfactionF 0.1739 0.1093 1.591 0.1117
LinearLag -0.09852 0.06148 -1.602 0.1091
QuadLag -0.07803 0.1235 -0.6318 0.5275
SatisXLinear 0.1084 0.06148 1.762 0.07801
SatisXQuad -0.1225 0.1454 -0.8422 0.3997 
pander(coefs.satisfaction.raw,style="rmarkdown") # lag terms and satisfaction: unstandardized
  Estimate Std..Error t.value
(Intercept) -0.007355 0.01116 -0.659
SatisfactionF 0.004348 0.002733 1.591
LinearLag -0.01714 0.01069 -1.602
QuadLag -0.01357 0.02148 -0.6318
SatisfactionF:LinearLag 0.004641 0.002634 1.762
SatisfactionF:QuadLag -0.005247 0.00623 -0.8422

Table: Table continues below

  p
(Intercept) 0.5099
SatisfactionF 0.1117
LinearLag 0.1091
QuadLag 0.5275
SatisfactionF:LinearLag 0.07801
SatisfactionF:QuadLag 0.3997 
pander(coefs.age,style="rmarkdown") # lag terms and age: standardized
  Estimate Std..Error t.value p
(Intercept) 3.416e-13 0.123 2.778e-12 1
AgeF -0.1107 0.1223 -0.9053 0.3653
LinearLag -0.07085 0.1057 -0.67 0.5028
QuadLag -0.3491 0.3627 -0.9626 0.3357
AgeXLinear 0.07869 0.1086 0.7246 0.4687
AgeXQuad 0.155 0.3752 0.4133 0.6794 
pander(coefs.age.raw,style="rmarkdown") # lag terms and age: unstandardized
  Estimate Std..Error t.value p
(Intercept) 0.02828 0.02052 1.378 0.1681
AgeF -0.001239 0.001369 -0.9053 0.3653
LinearLag -0.01232 0.01839 -0.67 0.5028
QuadLag -0.06073 0.06309 -0.9626 0.3357
AgeF:LinearLag 0.0009131 0.00126 0.7246 0.4687
AgeF:QuadLag 0.001799 0.004353 0.4133 0.6794

Correlation between Dyad Satisfaction and Adolescent Age

Given systematic similarities between the age and satisfaction models, we checked whether age and satisfaction correlate with one another. As presented below, we do not find a reliable correlation between age and satisfaction variables.

# grab only one slice from each dyad
corr_check = drps_raw[drps_raw$RawLag==0,]
pander(cor.test(corr_check$Satisfaction,corr_check$Age))
Test statistic df P value Alternative hypothesis
1.077 47 0.2868 two.sided

Table: Pearson's product-moment correlation: corr_check$Satisfaction and corr_check$Age