We required at least 20 data points to be included in data set
reduced_Fulladdiction–has all data including demographics
reduced_addiction_withMissing–has variables we are focusing on with misisng addiction_imputed–has variables we are focusing on with no misisng due to copymean imputation
library(readit)
library(INIT)
library(tidyverse)## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.1 ──## ✔ ggplot2 3.4.4 ✔ purrr 1.0.2
## ✔ tibble 3.2.1 ✔ dplyr 1.1.4
## ✔ tidyr 1.3.0 ✔ stringr 1.5.1
## ✔ readr 2.1.4 ✔ forcats 1.0.0## Warning: package 'tibble' was built under R version 4.3.1## Warning: package 'tidyr' was built under R version 4.2.2## Warning: package 'readr' was built under R version 4.2.2## Warning: package 'forcats' was built under R version 4.2.2## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()library(psych)##
## Attaching package: 'psych'## The following objects are masked from 'package:ggplot2':
##
## %+%, alphalibrary(longitudinalData)## Loading required package: clv## Loading required package: cluster## Loading required package: class## Warning: package 'class' was built under R version 4.3.0## Loading required package: rgl## This build of rgl does not include OpenGL functions. Use
## rglwidget() to display results, e.g. via options(rgl.printRglwidget = TRUE).## Loading required package: misc3d## Warning in fun(libname, pkgname): couldn't connect to display ":0"#library(igraph)
library(qgraph)
library(gimme)## Registered S3 method overwritten by 'quantmod':
## method from
## as.zoo.data.frame zoolibrary(psych)
library(forecast)
library(broom)## Warning: package 'broom' was built under R version 4.3.1library(metafor)## Loading required package: Matrix## Warning: package 'Matrix' was built under R version 4.3.1##
## Attaching package: 'Matrix'## The following objects are masked from 'package:tidyr':
##
## expand, pack, unpack## Loading required package: metadat## Loading required package: numDeriv##
## Loading the 'metafor' package (version 4.2-0). For an
## introduction to the package please type: help(metafor)library(brms)## Loading required package: Rcpp## Warning: package 'Rcpp' was built under R version 4.3.1## Loading 'brms' package (version 2.17.0). Useful instructions
## can be found by typing help('brms'). A more detailed introduction
## to the package is available through vignette('brms_overview').##
## Attaching package: 'brms'## The following object is masked from 'package:forecast':
##
## ma## The following object is masked from 'package:psych':
##
## cs## The following object is masked from 'package:stats':
##
## arlibrary(lavaan)## This is lavaan 0.6-16
## lavaan is FREE software! Please report any bugs.##
## Attaching package: 'lavaan'## The following object is masked from 'package:psych':
##
## cor2covlibrary(lme4)##
## Attaching package: 'lme4'## The following object is masked from 'package:brms':
##
## ngrpslibrary(bayesmeta)## Loading required package: forestplot## Loading required package: grid## Loading required package: checkmate## Loading required package: abind## Loading required package: mvtnorm##
## Attaching package: 'bayesmeta'## The following object is masked from 'package:stats':
##
## convolvelibrary(DT)## Warning: package 'DT' was built under R version 4.3.0###data files ####
###note non imputed include all data, whereas imputed exclude those who did not do at least 10 time points
# GIMME_Input <- "~/student led research/student led research 2024/data/andrew individual/gimme data"
# GIMME_Output<-"~/student led research/student led research 2024/output/gimme"
#
calculate_missing_percentages <- function(data, grouping_var) {
# Ensure the grouping variable is treated as a symbol for dynamic referencing
grouping_var_sym <- rlang::ensym(grouping_var)
# First dataframe: Calculate the percentage of missing values
missing_data_df <- data %>%
dplyr::group_by(!!grouping_var_sym) %>%
dplyr::mutate(TotalObservations = n()) %>%
dplyr::summarise(TotalObservations = first(TotalObservations), # Keep the total observations count
across(everything(), ~sum(is.na(.)) / length(.) * 100, .names = "perc_missing_{.col}")) %>%
ungroup()
# Second dataframe: Summarize the first dataframe
summary_df <- missing_data_df %>%
dplyr::select(-!!grouping_var_sym) %>% # Exclude the grouping variable from the summary
dplyr::summarise(across(.cols = everything(),
.fns = list(mean = ~mean(.x, na.rm = TRUE),
median = ~median(.x, na.rm = TRUE)),
.names = "{.col}_{.fn}"))
# Return a list containing both dataframes
return(list(MissingData = missing_data_df, Summary = summary_df))
}
###restructure meta analtic data, stacking it long
transform_to_long_format <- function(data_frame) {
long_data <- data_frame %>%
pivot_longer(-ID, names_to = "variable", values_to = "value") %>%
mutate(
IV = sub("_arimaCont.*", "", variable),
DV = sub(".*\\.(.*)", "\\1", variable), # Dynamically extract DV
Metric = ifelse(grepl("coef", variable), "Score", "SE"),
variable = NULL # Remove the original variable column
) %>%
pivot_wider(names_from = Metric, values_from = value) %>%
select(ID, IV, DV, Score, SE)
return(long_data)
}
addiction_withMissing <- readit("~/student led research/student led research 2024/data/cristobaladdiction/addiction_wtime.csv");## File guessed to be CSV ("~/student led research/student led research 2024/data/cristobaladdiction/addiction_wtime.csv")## Rows: 5027 Columns: 14
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## dbl (14): JustAFewMoreMinutes, FailedToDecreaseInternetTime, LostSleepOverIn...
##
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.addiction_imputed <- readit("~/student led research/student led research 2024/data/cristobaladdiction/Addiction_imputed.csv")## File guessed to be CSV ("~/student led research/student led research 2024/data/cristobaladdiction/Addiction_imputed.csv")
## Rows: 4105 Columns: 14── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## dbl (14): ID, Time, JustAFewMoreMinutes, FailedToDecreaseInternetTime, LostS...
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.Fulladdiction <- readit("~/student led research/student led research 2024/data/cristobaladdiction/data.xlsx");## File guessed to be xls/xlsx (Excel) ("~/student led research/student led research 2024/data/cristobaladdiction/data.xlsx")# getting time variable
Fulladdiction <-Fulladdiction %>%
arrange(Name, Date, MeasurementPoint) %>% # Ensure the data is ordered
group_by(Name) %>%
mutate(time = row_number()) %>% # Create 'time' as a sequential number
ungroup()
Fulladdiction <- Fulladdiction %>%
dplyr::rename(
JustAFewMoreMinutes = IAT_1,
FailedToDecreaseInternetTime = IAT_2,
LostSleepOverInternet = IAT_3,
FeltBadOffline = IAT_4,
PreferredInternetOverPeople = IAT_5,
Anxious =GAD_1,
ID = Name,
Time=time,
Tired =Tired_Awake,
Unhappy =Unhappy_Happy,
Agitated=Agitated_Calm,
NoEnergy=WoEnergy_FullEnergy,
Bad=Bad_Good,
Tense=Tense_Relaxed
)
length(unique(Fulladdiction$ID)) ## 122## [1] 122length(unique(addiction_withMissing$ID)) #`122## [1] 122length(unique(addiction_imputed$ID)) #84## [1] 84# Filter to include only IDs that are found in addiction_imputed and therefore have at least 20 values
reduced_addiction_withMissing <- addiction_withMissing %>%
filter(ID %in% addiction_imputed$ID)
reduced_Fulladdiction <- Fulladdiction %>%
filter(ID %in% addiction_imputed$ID)
length(unique(reduced_Fulladdiction$ID)) ## 84## [1] 84length(unique(reduced_addiction_withMissing$ID)) #84## [1] 84length(unique(addiction_imputed$ID)) #84## [1] 84## Mean age: 23.54762## SD age: 2.957908##
## Femenino Masculino Otro
## 44 36 4This shows you the simple correlations and significant levels
Within are correlations within person Between are correlations between people
This will produce arimax estimates for every internet addiction iv and every well being dv
If score is 2 times the standard error, then that result is significant for that person
## Rows: 2016 Columns: 5
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (2): IV, DV
## dbl (3): ID, Score, SE
##
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.Now we ask for the pooled effect of the internet addiction items linking to the well-being times
Score= strength of relationship betrween addiction and well-being
intercept is the pooled effect-ie. average link between internet addiction and well-being (should be negative)
Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESSIntercept -0.06 0.02 -0.11 -0.02 1.00 2374 4981
so on average, there is a negative link between addiction and well being…(estimate more than 2 x standard error)
sd(Intercept):
Estimate: 0.18 Est.Error: 0.02
This represents the estimated standard deviation of the intercept across the different levels of ID The estimate of 0.18 with a standard error of 0.02 suggests that there’s variability in the baseline scores across ID
so intercept +_2*SD is the range of links between well-being and addiction. -.06-.36=-.42…strongest link -.06 + .36= .30 …addition associated with higher well being for some
sd(IVFeltBadOffline):
Estimate: 0.15
This is the estimated standard deviation for the effect of the IV variable FeltBadOffline on Score, accounting for random effects across ID. It indicates how much the influence of feeling bad when offline varies across different individuals (ID).
cor(Intercept,IVFeltBadOffline):
Estimate: -0.35
This shows the estimated correlation between the intercept and the effect of FeltBadOffline. A negative value (-0.35) means that, across ID, higher baseline scores tend to be associated with a smaller effect of FeltBadOffline on the outcome (or vice versa).
cor(IVFeltBadOffline,IVPreferredInternetOverPeople):
Estimate: 0.50
This represents the estimated correlation between the effects of FeltBadOffline and PreferredInternetOverPeople on Score, again across ID. A positive correlation (0.50) suggests that individuals for whom FeltBadOffline has a stronger effect on Score also tend to have a stronger effect of PreferredInternetOverPeople.
## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: Score | se(SE) ~ 1 + IV + DV + (1 + IV + DV | ID)
## Data: combined_results (Number of observations: 2016)
## Draws: 4 chains, each with iter = 10000; warmup = 5000; thin = 1;
## total post-warmup draws = 20000
##
## Group-Level Effects:
## ~ID (Number of levels: 84)
## Estimate Est.Error
## sd(Intercept) 0.18 0.02
## sd(IVFeltBadOffline) 0.15 0.01
## sd(IVJustAFewMoreMinutes) 0.08 0.01
## sd(IVPreferredInternetOverPeople) 0.19 0.02
## sd(DVBad) 0.11 0.01
## sd(DVNoEnergy) 0.12 0.01
## sd(DVTense) 0.08 0.01
## sd(DVTired) 0.13 0.01
## sd(DVUnhappy) 0.11 0.01
## cor(Intercept,IVFeltBadOffline) -0.35 0.10
## cor(Intercept,IVJustAFewMoreMinutes) -0.08 0.14
## cor(IVFeltBadOffline,IVJustAFewMoreMinutes) 0.20 0.14
## cor(Intercept,IVPreferredInternetOverPeople) -0.46 0.09
## cor(IVFeltBadOffline,IVPreferredInternetOverPeople) 0.50 0.09
## cor(IVJustAFewMoreMinutes,IVPreferredInternetOverPeople) 0.07 0.14
## cor(Intercept,DVBad) -0.12 0.13
## cor(IVFeltBadOffline,DVBad) -0.30 0.12
## cor(IVJustAFewMoreMinutes,DVBad) 0.18 0.16
## cor(IVPreferredInternetOverPeople,DVBad) -0.10 0.13
## cor(Intercept,DVNoEnergy) -0.15 0.12
## cor(IVFeltBadOffline,DVNoEnergy) 0.07 0.13
## cor(IVJustAFewMoreMinutes,DVNoEnergy) 0.19 0.15
## cor(IVPreferredInternetOverPeople,DVNoEnergy) 0.03 0.13
## cor(DVBad,DVNoEnergy) 0.63 0.10
## cor(Intercept,DVTense) -0.34 0.13
## cor(IVFeltBadOffline,DVTense) -0.11 0.15
## cor(IVJustAFewMoreMinutes,DVTense) 0.22 0.18
## cor(IVPreferredInternetOverPeople,DVTense) -0.01 0.15
## cor(DVBad,DVTense) 0.41 0.15
## cor(DVNoEnergy,DVTense) 0.49 0.14
## cor(Intercept,DVTired) -0.19 0.12
## cor(IVFeltBadOffline,DVTired) -0.08 0.13
## cor(IVJustAFewMoreMinutes,DVTired) -0.09 0.15
## cor(IVPreferredInternetOverPeople,DVTired) -0.19 0.13
## cor(DVBad,DVTired) 0.36 0.13
## cor(DVNoEnergy,DVTired) 0.74 0.08
## cor(DVTense,DVTired) 0.29 0.16
## cor(Intercept,DVUnhappy) -0.16 0.13
## cor(IVFeltBadOffline,DVUnhappy) -0.22 0.13
## cor(IVJustAFewMoreMinutes,DVUnhappy) 0.17 0.16
## cor(IVPreferredInternetOverPeople,DVUnhappy) -0.09 0.13
## cor(DVBad,DVUnhappy) 0.72 0.09
## cor(DVNoEnergy,DVUnhappy) 0.39 0.13
## cor(DVTense,DVUnhappy) 0.32 0.16
## cor(DVTired,DVUnhappy) 0.28 0.14
## l-95% CI u-95% CI Rhat
## sd(Intercept) 0.15 0.22 1.00
## sd(IVFeltBadOffline) 0.13 0.19 1.00
## sd(IVJustAFewMoreMinutes) 0.06 0.11 1.00
## sd(IVPreferredInternetOverPeople) 0.16 0.22 1.00
## sd(DVBad) 0.09 0.14 1.00
## sd(DVNoEnergy) 0.10 0.15 1.00
## sd(DVTense) 0.06 0.11 1.00
## sd(DVTired) 0.10 0.16 1.00
## sd(DVUnhappy) 0.09 0.14 1.00
## cor(Intercept,IVFeltBadOffline) -0.54 -0.14 1.00
## cor(Intercept,IVJustAFewMoreMinutes) -0.35 0.20 1.00
## cor(IVFeltBadOffline,IVJustAFewMoreMinutes) -0.09 0.45 1.00
## cor(Intercept,IVPreferredInternetOverPeople) -0.63 -0.26 1.00
## cor(IVFeltBadOffline,IVPreferredInternetOverPeople) 0.30 0.67 1.00
## cor(IVJustAFewMoreMinutes,IVPreferredInternetOverPeople) -0.21 0.34 1.00
## cor(Intercept,DVBad) -0.37 0.13 1.00
## cor(IVFeltBadOffline,DVBad) -0.53 -0.05 1.00
## cor(IVJustAFewMoreMinutes,DVBad) -0.14 0.47 1.00
## cor(IVPreferredInternetOverPeople,DVBad) -0.35 0.14 1.00
## cor(Intercept,DVNoEnergy) -0.39 0.09 1.00
## cor(IVFeltBadOffline,DVNoEnergy) -0.18 0.32 1.00
## cor(IVJustAFewMoreMinutes,DVNoEnergy) -0.12 0.47 1.00
## cor(IVPreferredInternetOverPeople,DVNoEnergy) -0.21 0.28 1.00
## cor(DVBad,DVNoEnergy) 0.41 0.80 1.00
## cor(Intercept,DVTense) -0.58 -0.05 1.00
## cor(IVFeltBadOffline,DVTense) -0.40 0.20 1.00
## cor(IVJustAFewMoreMinutes,DVTense) -0.14 0.56 1.00
## cor(IVPreferredInternetOverPeople,DVTense) -0.31 0.28 1.00
## cor(DVBad,DVTense) 0.08 0.67 1.00
## cor(DVNoEnergy,DVTense) 0.20 0.72 1.00
## cor(Intercept,DVTired) -0.42 0.06 1.00
## cor(IVFeltBadOffline,DVTired) -0.33 0.17 1.00
## cor(IVJustAFewMoreMinutes,DVTired) -0.39 0.21 1.00
## cor(IVPreferredInternetOverPeople,DVTired) -0.43 0.06 1.00
## cor(DVBad,DVTired) 0.08 0.59 1.00
## cor(DVNoEnergy,DVTired) 0.56 0.87 1.00
## cor(DVTense,DVTired) -0.04 0.57 1.00
## cor(Intercept,DVUnhappy) -0.40 0.10 1.00
## cor(IVFeltBadOffline,DVUnhappy) -0.47 0.04 1.00
## cor(IVJustAFewMoreMinutes,DVUnhappy) -0.14 0.48 1.00
## cor(IVPreferredInternetOverPeople,DVUnhappy) -0.34 0.17 1.00
## cor(DVBad,DVUnhappy) 0.51 0.87 1.00
## cor(DVNoEnergy,DVUnhappy) 0.12 0.62 1.00
## cor(DVTense,DVUnhappy) -0.03 0.60 1.00
## cor(DVTired,DVUnhappy) -0.01 0.53 1.00
## Bulk_ESS Tail_ESS
## sd(Intercept) 4584 8200
## sd(IVFeltBadOffline) 3661 8178
## sd(IVJustAFewMoreMinutes) 5800 9532
## sd(IVPreferredInternetOverPeople) 3950 8073
## sd(DVBad) 5598 10221
## sd(DVNoEnergy) 4955 8651
## sd(DVTense) 4241 7089
## sd(DVTired) 5615 9778
## sd(DVUnhappy) 5865 10302
## cor(Intercept,IVFeltBadOffline) 3424 7372
## cor(Intercept,IVJustAFewMoreMinutes) 7584 11918
## cor(IVFeltBadOffline,IVJustAFewMoreMinutes) 6544 10918
## cor(Intercept,IVPreferredInternetOverPeople) 4068 7744
## cor(IVFeltBadOffline,IVPreferredInternetOverPeople) 3048 6314
## cor(IVJustAFewMoreMinutes,IVPreferredInternetOverPeople) 2727 6516
## cor(Intercept,DVBad) 5643 10154
## cor(IVFeltBadOffline,DVBad) 5601 9726
## cor(IVJustAFewMoreMinutes,DVBad) 4409 8277
## cor(IVPreferredInternetOverPeople,DVBad) 6115 9619
## cor(Intercept,DVNoEnergy) 5402 9287
## cor(IVFeltBadOffline,DVNoEnergy) 5316 9163
## cor(IVJustAFewMoreMinutes,DVNoEnergy) 3836 6990
## cor(IVPreferredInternetOverPeople,DVNoEnergy) 5897 11157
## cor(DVBad,DVNoEnergy) 4876 8908
## cor(Intercept,DVTense) 8601 11997
## cor(IVFeltBadOffline,DVTense) 7609 12102
## cor(IVJustAFewMoreMinutes,DVTense) 5243 8308
## cor(IVPreferredInternetOverPeople,DVTense) 8460 13694
## cor(DVBad,DVTense) 5897 10899
## cor(DVNoEnergy,DVTense) 5864 11044
## cor(Intercept,DVTired) 6361 10658
## cor(IVFeltBadOffline,DVTired) 5637 10366
## cor(IVJustAFewMoreMinutes,DVTired) 4352 8245
## cor(IVPreferredInternetOverPeople,DVTired) 6945 10978
## cor(DVBad,DVTired) 4067 8372
## cor(DVNoEnergy,DVTired) 7498 11813
## cor(DVTense,DVTired) 5897 11262
## cor(Intercept,DVUnhappy) 6761 12223
## cor(IVFeltBadOffline,DVUnhappy) 6750 10833
## cor(IVJustAFewMoreMinutes,DVUnhappy) 5022 9141
## cor(IVPreferredInternetOverPeople,DVUnhappy) 7953 11933
## cor(DVBad,DVUnhappy) 5046 9568
## cor(DVNoEnergy,DVUnhappy) 5983 12182
## cor(DVTense,DVUnhappy) 5813 10271
## cor(DVTired,DVUnhappy) 5980 10225
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept -0.06 0.02 -0.11 -0.02 1.00
## IVFeltBadOffline 0.01 0.02 -0.03 0.05 1.00
## IVJustAFewMoreMinutes 0.03 0.01 0.00 0.05 1.00
## IVPreferredInternetOverPeople 0.04 0.02 -0.00 0.08 1.00
## DVBad 0.00 0.02 -0.03 0.03 1.00
## DVNoEnergy -0.00 0.02 -0.04 0.03 1.00
## DVTense 0.00 0.01 -0.02 0.03 1.00
## DVTired -0.02 0.02 -0.05 0.02 1.00
## DVUnhappy 0.00 0.02 -0.03 0.03 1.00
## Bulk_ESS Tail_ESS
## Intercept 2374 4981
## IVFeltBadOffline 3212 6326
## IVJustAFewMoreMinutes 8416 12320
## IVPreferredInternetOverPeople 3429 6076
## DVBad 5119 8847
## DVNoEnergy 5198 9542
## DVTense 6763 10555
## DVTired 5988 10341
## DVUnhappy 6546 10504
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.00 0.00 0.00 0.00 NA NA NA
##
## Draws were sampled using sampling(NUTS). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).level of heterogeneity of every combination of variable
Rough guidelines for interpreting I2 in the meta-analytic literature are that values less than 25% reflect low inconsistency, 25% to 50% reflect moderate inconsistency, 50 to 75% reflect high inconsistency, and over 75% show very high inconsistency (Higgins et al., 2003). While there are no absolute cutoffs, in the Cochrane library of meta-analyses, for example, the median I2 is 21% (Ioannidis et al., 2007). If I2 exceeds even 50%, it is common to search for subgroups or to avoid reporting pooled effects (Lo et al., 2019).