Predictors of Depression during the Covid-19, Czech sample report v0.5
Sarka Tesarova, Ondrej Pekacek, Alessandro Porrovecchio, Jiri Remr
Last edited 31. 03. 2021
1. Study background
a. Goal of the study
This research project is based on the umbrella project “Pandemic Emergency in Social Perspective. Evidence from a large Web-survey research”, designed and organized by principal investigators Linda Lombi (Università Cattolica del Sacro Cuore, Milan) and Marco Terraneo (Università Bicocca-Milano).
The principal goal of the international cross-sectional study is to explore the predictors of depression within the European context of the Covid-19 pandemic, specifically during the lockdown and social distancing period of March-April 2020.
Our team has decided to primarily focus on the impact of modifiable behavioral/lifestyle factors, such as exercise, alcohol and tobacco consumption, but, also, the usage of social media as a source of information about the pandemic. Our intention is to create and validate a depression model that these literature-based predictors should predict. Furthermore, we intend to explore the indirect pathway between social media consumption and depression mediated by the level of anxiety.
Supplementary data for this project, such as the survey questionnaire, original dataset and are accessible in our Open Science Framework repository.
b. Sampling
Given the rapidly-developing nature of the Covid-19 pandemic, the principal research team (Lombi & Terraneo) chose a convenience sample, recruited through Facebook national groups using a snowballing technique. The goal was to collect at least 1000 responses per country.
The data collection has been conducted between March-April 2020 in the following eight countries: Italy, France, Germany, Spain, United Kingdom, Sweden, Poland, Czech Republic and were conducted by the members of the respective national teams (please see the research protocol in the OSF repository.
This relatively non-random sampling is likely to result in a non-representative sample for the national populations. This is one of the limitations of this research and is reflected in the “data collection and sampling” part of the research protocol outlined by Linda Lombi and Marco Terraneo.
This approach, therefore, does not aim to compare country-samples, but, rather, to compare segments of the national samples, with a particular focus on the vulnerable social groups, determined by socio-demographic, lifestyle professional and living condition aspects.
c. Analysis plan
In order to comply with the principles of Open Science, we intend to split our analysis to two parts.
- Within the first part, we test the literature-derived hypotheses on
the Czech sample (n=1484) of the international study and develop models.
We also explore the dataset (here referred to as
COV19_05_agroup.sav) inductively and consider the formulation of additional hypotheses for other predictors that might have been missed before the beginning of the study. To lower the chance of overfitting, we only consider the adding additional variables that have an empirical support based on our review of the existing literature. Towards the end of the first part of the project, we pre-register our hypotheses and other key research information (including this reproducible R code) at the OSF Registries. While some of the team members have briefly interacted with the international dataset, they have not been involved in the pre-registration and hypothesis forming process in order to reduce biases by separating the exploratory and confirmatory phases of the research. - In the second we will access the international dataset, which will include data from all of the countries that gathered at least 1000 responses. We will conduct confirmatory analyses, testing our models on this international sample, from which we will exclude the Czech sub-sample.
d. Core hypotheses
| Alternative Hypotheses | Variable | Literature |
|---|---|---|
| H1: Female gender is associated with higher levels of depression. | q01 | (Salk, Hyde, and Abramson 2017; Kowal et al. 2020; Wang et al. 2020; Luo et al. 2020; González-Sanguino et al. 2020) |
| H2: Higher age is associated with lower levels of depression. | q02 | (Kowal et al. 2020; Shevlin et al. 2020; Taylor et al. 2008; Losada-Baltar et al. 2020; González-Sanguino et al. 2020) |
| H3: People in a relationship experience lower levels of depression. | q03 | (Kowal et al. 2020; Jacob, Haro, and Koyanagi 2019) |
| H4: Parenthood is associated with significantly different levels of depression. | q04 | (Stanca 2012; Shevlin et al. 2020) |
| H5: Higher education is associated with lower levels of depression. | q11 | (Kowal et al. 2020; Gloster et al. 2020; Taylor et al. 2008) |
| H6: Use of social media is associated with higher levels of depression. | q18_02 | (Bendau et al. 2020; Dhir et al. 2018; Primack et al. 2017) |
| H7: Physical contact with friends and family is associated with lower levels of depression. | q35_01, q35_03 | (Gloster et al. 2020; Tull et al. 2020; Luo et al. 2020) |
| H8: Regular consumption of alcohol and tobacco is associated with higher levels of depression. | q38, q40 | (Stanton et al. 2020; Awaworyi Churchill and Farrell 2017) |
| H9: Regular workouts or physical activity are associated with lower levels of depression. | q42 | (Harvey et al. 2018; Schuch et al. 2016; Kvam et al. 2016; Krogh et al. 2017; Stubbs et al. 2018) |
| H10: Worse self-rated health quality is associated with higher levels of depression. | q47, q48, q47 | (Ambresin et al. 2014; Vindegaard and Benros 2020; Hossain et al. 2020) |
| H11: In addition to H6, we hypothesize an existence of a causal pathway leading from social media exposure to depression, which is mediated by anxiety and moderated by age. | q02, q18_02, anxiety_index | (Bendau et al. 2020; Rasmussen et al. 2020; Wheaton, Prikhidko, and Messner 2021; Vannucci, Flannery, and Ohannessian 2017; Mertens et al. 2020) |
2. Analysis of the Czech sample
a. Loading the dataset, required R packages and data wrangling
The code below can be run in R or in R IDE, such as R Studio. We used R Markdown within the R Studio to compose this report and used the open-source jamovi software (a R GUI) to conduct some of the exploratory analyses.
#The following packages might need to be installed onto your version of R prior to the running of the code below.
library(udpipe)
library(MASS)
library(medmod)
library(lavaan)
library(lavaanPlot)
library(wordcloud)
library(corrplot)
library(tidytext)
library(tidyverse)
library(haven)
library(jmv)
#Package lavaanPlot is currently not available on CRAN, install through devtools::install_github("alishinski/lavaanPlot")
#We load the original Czech dataset (in SPSS format) from a local directory.
data <- haven::read_sav(file = "COV19_05_agroup.sav")
#We also try to limit the decimals to three significant figures
options(digits = 3)#Firstly, because the source file is an SPSS file, we need to specify that we would like to see value labels (such as Male/Female) for selected variables, as opposed to just numeric values (such as 1/2). This is not essential for the analysis, but seeing the names of labels will enable better understanding of the results.
data <- data %>%
mutate(q01 = haven::as_factor(q01)) %>%
mutate(q03 = haven::as_factor(q03)) %>%
mutate(q04 = haven::as_factor(q04)) %>%
mutate(q11 = haven::as_factor(q11)) %>%
mutate(q38 = haven::as_factor(q38)) %>%
mutate(q40 = haven::as_factor(q40)) %>%
mutate(q42 = haven::as_factor(q42)) %>%
mutate(q47 = haven::as_factor(q47)) %>%
mutate(q48 = haven::as_factor(q48)) %>%
mutate(q49 = haven::as_factor(q49)) %>%
mutate(Q4_AGE_r = haven::as_factor(Q4_AGE_r, ordered = TRUE))
#For convenience, we translate the core variables labels from Czech to English.
levels(data$q01) <- list(female = "Žena",
male = "Muž")
levels(data$q03) <- list(single = "Svobodný/Svobodná",
relationship = "V partnerském vztahu",
married = "Ženatý/Vdaná",
divorced = "Rozvedený/Rozvedená",
widowed = "Vdovec/Vdova")
levels(data$q04) <- list(yes = "Ano", no = "Ne")
levels(data$q04) <- list(yes = "Ano", no = "Ne")
levels(data$q11) <- list(unfin_element = "Nedokončené základní",
element = "Základní vzdělání",
unfin_hs = "Vyučen/a nebo střední bez maturity",
hs = "Středoškolské s maturitou a vyšší odborné",
undergrad = "Vysokoškolské (bakalářský stupeň)",
postgrad = "Vysokoškolské (magisterské, doktorské nebo vyšší)")
levels(data$q38) <- list(yes = "Ano", no = "Ne")
levels(data$q40) <- list(yes = "Ano", no = "Ne")
levels(data$q42) <- list(yes = "Ano", no = "Ne")
levels(data$q47) <- list(excellent = "Velmi dobrý",
good = "Dobrý",
neutral = "Ani dobrý ani špatný",
bad = "Špatný",
very_bad = "Velmi špatný")
levels(data$q48) <- list(yes = "Ano", no = "Ne")
levels(data$q49) <- list(limits = "Velmi omezuje",
partially_limits = "Částečně omezuje",
no_limits = "Vůbec neomezuje")
#Secondly, we rename key variables to a more human-readable form.
data <- data %>%
rename(q01_gender = q01) %>%
rename(q02_age = q02) %>%
rename(q03_relationship_type = q03) %>%
rename(q04_children = q04) %>%
rename(q11_education = q11) %>%
rename(q18_02_soc_media_info = q18_02) %>%
rename(q35_01_contact_close_family = q35_01) %>%
rename(q35_03_contact_friends = q35_03) %>%
rename(q38_alcohol = q38) %>%
rename(q40_smoking = q40) %>%
rename(q42_sport = q42) %>%
rename(q47_self_reporting_health = q47) %>%
rename(q48_chronic_illness = q48) %>%
rename(q49_health_limitations = q49) %>%
#Thirdly, we rename variables related to anxiety, which we will use to construct an anxiety index with factor analysis (to use for path analysis afterwards).
rename(q30_anxiety_infection_covid = q30) %>%
rename(q31_anxiety_infection_friends = q31) %>%
rename(q33_01_anxiety_situation = q33_01) %>%
rename(q33_02_anxiety_low_control = q33_02) %>%
rename(q33_03_anxiety_survival_covid = q33_03) %>%
rename(q33_04_anxiety_change_employment = q33_04) %>%
rename(q33_05_anxiety_infecting_others = q33_05)
#Finally, we create dummy variable for social media as a source of information by replacing NA values. We also assign it appropriate factor levels
data$q18_02_soc_media_info[is.na(data$q18_02_soc_media_info)] <- 0
data <- data %>%
mutate(q18_02_soc_media_info = haven::as_factor(q18_02_soc_media_info, ordered = TRUE))
levels(data$q18_02_soc_media_info) <- list(yes = "Sociální média", no = "0")b. Sample descriptive statistics: Depression index (PHQ8)
The PHQ-8 dependent variable intend to determine the presence and severity of major depressive disorder. The PHQ-8 construction is standardized and based on the established methodology (Kroenke et al. 2009). The PHQ-8 questionnaire asks the number of days in the past 2 weeks the respondent had experienced a specific depressive symptom.
This variable was recoded by the international team from 8 survey items (see the OSF project page for the precise syntax) and is thus already present in the version of this dataset.
#To summarize the dependent continuous variable, we use the descriptives() function from the jmv package.
jmv::descriptives(
data = data,
vars = "PHQ8",
freq = TRUE,
box = TRUE,
median = FALSE,
range = TRUE,
se = TRUE,
quart = TRUE)##
## DESCRIPTIVES
##
## Descriptives
## ────────────────────────────
## PHQ8
## ────────────────────────────
## N 1484
## Missing 0
## Mean 4.71
## Std. error mean 0.120
## Range 24.0
## Minimum 0.00
## Maximum 24.0
## 25th percentile 1.00
## 50th percentile 3.00
## 75th percentile 7.00
## ────────────────────────────
c. Sample descriptive statistics: Demographic characteristics (Czech sample)
In the next step, we asses the demographic characteristics of the respondents in the Czech sample.
#To summarize the key demographic variables, we use the descriptives() function from the jmv package.
jmv::descriptives(
data = data,
vars = vars("q01_gender",
"Q4_AGE_r",
"q03_relationship_type",
"q04_children",
"q11_education"),
bar = TRUE,
freq = TRUE,
missing = FALSE,
mean = FALSE,
median = FALSE,
sd = FALSE,
min = FALSE,
max = FALSE)##
## DESCRIPTIVES
##
## Descriptives
## ─────────────────────────────────────────────────────────────────────────────────────────
## q01_gender Q4_AGE_r q03_relationship_type q04_children q11_education
## ─────────────────────────────────────────────────────────────────────────────────────────
## N 1484 1484 1484 1484 1484
## ─────────────────────────────────────────────────────────────────────────────────────────
##
##
## FREQUENCIES
##
## Frequencies of q01_gender
## ──────────────────────────────────────────────────
## Levels Counts % of Total Cumulative %
## ──────────────────────────────────────────────────
## female 1054 71.0 71.0
## male 430 29.0 100.0
## ──────────────────────────────────────────────────
##
##
## Frequencies of Q4_AGE_r
## ───────────────────────────────────────────────────────
## Levels Counts % of Total Cumulative %
## ───────────────────────────────────────────────────────
## 16-29 years 379 25.5 25.5
## 30-49 years 440 29.6 55.2
## 50-64 years 206 13.9 69.1
## 65+ 459 30.9 100.0
## ───────────────────────────────────────────────────────
##
##
## Frequencies of q03_relationship_type
## ────────────────────────────────────────────────────────
## Levels Counts % of Total Cumulative %
## ────────────────────────────────────────────────────────
## single 332 22.4 22.4
## relationship 283 19.1 41.4
## married 586 39.5 80.9
## divorced 155 10.4 91.4
## widowed 128 8.6 100.0
## ────────────────────────────────────────────────────────
##
##
## Frequencies of q04_children
## ──────────────────────────────────────────────────
## Levels Counts % of Total Cumulative %
## ──────────────────────────────────────────────────
## yes 937 63.1 63.1
## no 547 36.9 100.0
## ──────────────────────────────────────────────────
##
##
## Frequencies of q11_education
## ─────────────────────────────────────────────────────────
## Levels Counts % of Total Cumulative %
## ─────────────────────────────────────────────────────────
## unfin_element 5 0.3 0.3
## element 109 7.3 7.7
## unfin_hs 74 5.0 12.7
## hs 537 36.2 48.9
## undergrad 152 10.2 59.1
## postgrad 607 40.9 100.0
## ─────────────────────────────────────────────────────────
3. Building regression model to predict PHQ8
After descriptive statistics, we continue with building and fitting of the regression model based on our hypotheses.
The model has one independent continuous variable - PHQ8. The only other continuous variable in the model is q02_age, which is inputted as a covariate. The rest of the variables are either categorical (both nominal and ordinal) or binary. The linreg() function from the jmv package automatically handles them as dummy variables with reference levels and it is thus not necessary to create further dummy variables prior to this analysis.
a. Overview of correlations between individual predictors and outcome
As a first step in the regression model creation, we conduct a correlation analysis. Since we do not presume linearity between all of the variables, we use Spearman’s rank coefficient instead of Pearson’s r. The results below need to be interpreted with caution, since some of the variables are categorical (such as q03_relationship_type), without a defined order. For categorical variables, comparisons using Chi-Square test would be more appropriate, however, in this step, we are primarily looking at the relationship between the outcome (PHQ8) and the theorized predictors.
#While the dataset has been already imported, the values of factor variables were changed from numerics to text strings, therefore that dataset is unsuitable for correlation analysis. To solve this, we create a parallel dataset, again renaming the key variables to a more understandable form.
data_corr <- read_sav(file = "COV19_05_agroup.sav")
options(digits = 3)
data_corr <- data_corr %>%
rename(q01_gender = q01) %>%
rename(q02_age = q02) %>%
rename(q03_relationship_type = q03) %>%
rename(q04_children = q04) %>%
rename(q11_education = q11) %>%
rename(q18_02_soc_media_info = q18_02) %>%
rename(q35_01_contact_close_family = q35_01) %>%
rename(q35_03_contact_friends = q35_03) %>%
rename(q38_alcohol = q38) %>%
rename(q40_smoking = q40) %>%
rename(q42_sport = q42) %>%
rename(q47_self_reporting_health = q47) %>%
rename(q48_chronic_illness = q48) %>%
rename(q49_health_limitations = q49)
data_corr$q18_02_soc_media_info[is.na(data_corr$q18_02_soc_media_info)] <- 0
#Dataset simplification to only the variables specified with hypotheses
data_vars <- c("PHQ8",
"q02_age",
"q01_gender",
"q04_children",
"q42_sport",
"q40_smoking",
"q38_alcohol",
"q35_01_contact_close_family",
"q35_03_contact_friends",
"q18_02_soc_media_info",
"q03_relationship_type",
"q47_self_reporting_health",
"q49_health_limitations",
"q11_education",
"q48_chronic_illness")
data_corr <- data_corr[data_vars]
#SPEARMAN CORRELATION
corr_spearman <- cor(data_corr, method = "spearman", use = "complete.obs")
corrplot(corr_spearman, method = "color")
print(corr_spearman)## PHQ8 q02_age q01_gender q04_children q42_sport
## PHQ8 1.0000 -0.3125 -0.08050 0.2598 0.0272
## q02_age -0.3125 1.0000 -0.12945 -0.6552 0.1679
## q01_gender -0.0805 -0.1295 1.00000 0.1555 -0.0382
## q04_children 0.2598 -0.6552 0.15555 1.0000 -0.1219
## q42_sport 0.0272 0.1679 -0.03822 -0.1219 1.0000
## q40_smoking -0.0734 0.0779 -0.00954 -0.0545 -0.1106
## q38_alcohol -0.0390 0.1416 -0.15814 -0.1365 0.0938
## q35_01_contact_close_family 0.1016 -0.3148 0.11691 0.1797 -0.0775
## q35_03_contact_friends 0.0397 -0.0648 0.07751 0.0944 0.0303
## q18_02_soc_media_info 0.1886 -0.3475 0.00616 0.2114 -0.0165
## q03_relationship_type -0.2103 0.7162 -0.20108 -0.6772 0.1122
## q47_self_reporting_health 0.1845 0.2575 -0.00484 -0.1461 0.2626
## q49_health_limitations -0.1448 -0.2395 0.00610 0.1123 -0.1607
## q11_education -0.1286 0.2393 -0.03178 -0.2322 -0.1017
## q48_chronic_illness -0.0835 -0.2811 0.03011 0.1528 -0.1617
## q40_smoking q38_alcohol q35_01_contact_close_family
## PHQ8 -0.07337 -0.0390 0.1016
## q02_age 0.07788 0.1416 -0.3148
## q01_gender -0.00954 -0.1581 0.1169
## q04_children -0.05446 -0.1365 0.1797
## q42_sport -0.11059 0.0938 -0.0775
## q40_smoking 1.00000 0.0826 -0.0301
## q38_alcohol 0.08264 1.0000 -0.0661
## q35_01_contact_close_family -0.03008 -0.0661 1.0000
## q35_03_contact_friends -0.09691 0.0180 0.1312
## q18_02_soc_media_info -0.10582 -0.0246 0.0778
## q03_relationship_type 0.07538 0.1632 -0.2498
## q47_self_reporting_health -0.04259 0.1538 -0.0655
## q49_health_limitations -0.03849 -0.1438 0.0903
## q11_education 0.08480 -0.1065 -0.0804
## q48_chronic_illness -0.06276 -0.0881 0.0947
## q35_03_contact_friends q18_02_soc_media_info
## PHQ8 0.0397 0.18859
## q02_age -0.0648 -0.34752
## q01_gender 0.0775 0.00616
## q04_children 0.0944 0.21142
## q42_sport 0.0303 -0.01654
## q40_smoking -0.0969 -0.10582
## q38_alcohol 0.0180 -0.02457
## q35_01_contact_close_family 0.1312 0.07776
## q35_03_contact_friends 1.0000 0.07890
## q18_02_soc_media_info 0.0789 1.00000
## q03_relationship_type -0.0746 -0.23494
## q47_self_reporting_health -0.0116 -0.06681
## q49_health_limitations 0.0134 0.05345
## q11_education -0.0759 -0.12047
## q48_chronic_illness 0.0166 0.08028
## q03_relationship_type q47_self_reporting_health
## PHQ8 -0.2103 0.18453
## q02_age 0.7162 0.25751
## q01_gender -0.2011 -0.00484
## q04_children -0.6772 -0.14615
## q42_sport 0.1122 0.26260
## q40_smoking 0.0754 -0.04259
## q38_alcohol 0.1632 0.15378
## q35_01_contact_close_family -0.2498 -0.06555
## q35_03_contact_friends -0.0746 -0.01156
## q18_02_soc_media_info -0.2349 -0.06681
## q03_relationship_type 1.0000 0.19053
## q47_self_reporting_health 0.1905 1.00000
## q49_health_limitations -0.2078 -0.46591
## q11_education 0.1888 -0.09659
## q48_chronic_illness -0.2113 -0.43942
## q49_health_limitations q11_education
## PHQ8 -0.1448 -0.12859
## q02_age -0.2395 0.23927
## q01_gender 0.0061 -0.03178
## q04_children 0.1123 -0.23216
## q42_sport -0.1607 -0.10165
## q40_smoking -0.0385 0.08480
## q38_alcohol -0.1438 -0.10647
## q35_01_contact_close_family 0.0903 -0.08044
## q35_03_contact_friends 0.0134 -0.07589
## q18_02_soc_media_info 0.0535 -0.12047
## q03_relationship_type -0.2078 0.18877
## q47_self_reporting_health -0.4659 -0.09659
## q49_health_limitations 1.0000 0.04214
## q11_education 0.0421 1.00000
## q48_chronic_illness 0.4639 -0.00365
## q48_chronic_illness
## PHQ8 -0.08345
## q02_age -0.28108
## q01_gender 0.03011
## q04_children 0.15278
## q42_sport -0.16171
## q40_smoking -0.06276
## q38_alcohol -0.08812
## q35_01_contact_close_family 0.09473
## q35_03_contact_friends 0.01659
## q18_02_soc_media_info 0.08028
## q03_relationship_type -0.21127
## q47_self_reporting_health -0.43942
## q49_health_limitations 0.46390
## q11_education -0.00365
## q48_chronic_illness 1.00000b. Theory derived, inductively built regression model
In the first set of models, we avoid potentially biased modifications, such as pairwise comparisons, which could lead to overfitting. Instead, we build five successive models in total (“blocks” in the syntax).
First model (from the set of 5) uses only the demographic characteristics as predictors. Second model add the effect of the social media consumption. Third model adds lifestyle variables, such as alcohol, smoking, sport and social contacts. The fourth model further adds the variables related to self-rated health quality. Finally, the fifth model controls for the age by adding it as a covariate. The performance of each model could be seen in the output below.
jmv::linReg(
data = data,
dep = "PHQ8",
covs = "q02_age",
factors = vars("q01_gender",
"q04_children",
"q42_sport",
"q40_smoking",
"q38_alcohol",
"q35_01_contact_close_family",
"q35_03_contact_friends",
"q18_02_soc_media_info",
"q03_relationship_type",
"q47_self_reporting_health",
"q49_health_limitations",
"q11_education",
"q48_chronic_illness"),
blocks = list(
list(
"q01_gender",
"q03_relationship_type",
"q04_children",
"q11_education"),
list(
"q18_02_soc_media_info"),
list(
"q40_smoking",
"q42_sport",
"q38_alcohol",
"q35_01_contact_close_family",
"q35_03_contact_friends"),
list(
"q47_self_reporting_health",
"q48_chronic_illness",
"q49_health_limitations"),
list(
"q02_age")),
refLevels = list(
list(
var="q01_gender",
ref="female"),
list(
var="q04_children",
ref="no"),
list(
var="q42_sport",
ref="no"),
list(
var="q40_smoking",
ref="yes"),
list(
var="q38_alcohol",
ref="yes"),
list(
var="q35_01_contact_close_family",
ref="1"),
list(
var="q35_03_contact_friends",
ref="1"),
list(
var="q18_02_soc_media_info",
ref="yes"),
list(
var="q03_relationship_type",
ref="single"),
list(
var="q47_self_reporting_health",
ref="very_bad"),
list(
var="q49_health_limitations",
ref="limits"),
list(
var="q11_education",
ref="unfin_element"),
list(
var="q48_chronic_illness",
ref="yes")),
r2Adj = TRUE,
aic = TRUE,
bic = TRUE,
rmse = TRUE,
modelTest = TRUE,
anova = TRUE,
ci = TRUE,
stdEst = TRUE,
ciStdEst = TRUE,
durbin = TRUE,
collin = TRUE)##
## LINEAR REGRESSION
##
## Model Fit Measures
## ─────────────────────────────────────────────────────────────────────────────────────────────────────
## Model R R² Adjusted R² AIC BIC RMSE F df1 df2 p
## ─────────────────────────────────────────────────────────────────────────────────────────────────────
## 1 0.297 0.0880 0.0811 8518 8587 4.42 12.73 11 1450 < .001
## 2 0.320 0.1022 0.0948 8497 8571 4.38 13.75 12 1449 < .001
## 3 0.336 0.1132 0.1015 8493 8604 4.35 9.69 19 1442 < .001
## 4 0.446 0.1989 0.1844 8358 8507 4.14 13.71 26 1435 < .001
## 5 0.484 0.2346 0.2202 8294 8447 4.05 16.28 27 1434 < .001
## ─────────────────────────────────────────────────────────────────────────────────────────────────────
##
##
## Model Comparisons
## ───────────────────────────────────────────────────────────────────
## Model Model ΔR² F df1 df2 p
## ───────────────────────────────────────────────────────────────────
## 1 - 2 0.0142 22.89 1 1449 < .001
## 2 - 3 0.0110 2.55 7 1442 0.013
## 3 - 4 0.0857 21.93 7 1435 < .001
## 4 - 5 0.0357 66.88 1 1434 < .001
## ───────────────────────────────────────────────────────────────────
##
##
## MODEL SPECIFIC RESULTS
##
## MODEL 1
##
## Omnibus ANOVA Test
## ─────────────────────────────────────────────────────────────────────────────────────
## Sum of Squares df Mean Square F p
## ─────────────────────────────────────────────────────────────────────────────────────
## q01_gender 332 1 331.8 16.87 < .001
## q03_relationship_type 167 4 41.8 2.13 0.075
## q04_children 362 1 362.0 18.41 < .001
## q11_education 246 5 49.2 2.50 0.029
## Residuals 28515 1450 19.7
## ─────────────────────────────────────────────────────────────────────────────────────
## Note. Type 3 sum of squares
##
##
## Model Coefficients - PHQ8
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Predictor Estimate SE Lower Upper t p Stand. Estimate Lower Upper
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Intercept ᵃ 6.008 2.008 2.07 9.947 2.993 0.003
## q01_gender:
## male – female -1.096 0.267 -1.62 -0.573 -4.108 < .001 -0.2370 -0.350 -0.1238
## q03_relationship_type:
## relationship – single -0.612 0.384 -1.37 0.142 -1.593 0.111 -0.1323 -0.295 0.0306
## married – single -1.141 0.443 -2.01 -0.272 -2.576 0.010 -0.2467 -0.435 -0.0588
## divorced – single -0.781 0.542 -1.84 0.282 -1.442 0.150 -0.1688 -0.399 0.0609
## widowed – single -1.423 0.558 -2.52 -0.328 -2.550 0.011 -0.3076 -0.544 -0.0710
## q04_children:
## yes – no -1.572 0.366 -2.29 -0.853 -4.290 < .001 -0.3399 -0.495 -0.1845
## q11_education:
## element – unfin_element 2.058 2.037 -1.94 6.053 1.011 0.312 0.4449 -0.419 1.3085
## unfin_hs – unfin_element 1.155 2.058 -2.88 5.191 0.561 0.575 0.2496 -0.623 1.1221
## hs – unfin_element 0.871 1.996 -3.05 4.787 0.436 0.663 0.1882 -0.658 1.0347
## undergrad – unfin_element 0.952 2.020 -3.01 4.915 0.471 0.637 0.2058 -0.651 1.0624
## postgrad – unfin_element 0.387 1.994 -3.52 4.297 0.194 0.846 0.0836 -0.762 0.9289
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## ᵃ Represents reference level
##
##
## ASSUMPTION CHECKS
##
## Durbin–Watson Test for Autocorrelation
## ────────────────────────────────────────────
## Autocorrelation DW Statistic p
## ────────────────────────────────────────────
## 0.0456 1.91 0.062
## ────────────────────────────────────────────
##
##
## Collinearity Statistics
## ──────────────────────────────────────────────
## VIF Tolerance
## ──────────────────────────────────────────────
## q01_gender 1.05 0.956
## q03_relationship_type 1.12 0.891
## q04_children 1.52 0.656
## q11_education 1.03 0.971
## ──────────────────────────────────────────────
##
##
## MODEL 2
##
## Omnibus ANOVA Test
## ─────────────────────────────────────────────────────────────────────────────────────
## Sum of Squares df Mean Square F p
## ─────────────────────────────────────────────────────────────────────────────────────
## q01_gender 308 1 308.2 15.91 < .001
## q03_relationship_type 115 4 28.7 1.48 0.205
## q04_children 321 1 321.1 16.57 < .001
## q11_education 207 5 41.4 2.14 0.059
## q18_02_soc_media_info 444 1 443.5 22.89 < .001
## Residuals 28071 1449 19.4
## ─────────────────────────────────────────────────────────────────────────────────────
## Note. Type 3 sum of squares
##
##
## Model Coefficients - PHQ8
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Predictor Estimate SE Lower Upper t p Stand. Estimate Lower Upper
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Intercept ᵃ 6.614 1.997 2.70 10.5310 3.312 < .001
## q01_gender:
## male – female -1.057 0.265 -1.58 -0.5372 -3.988 < .001 -0.228 -0.341 -0.11611
## q03_relationship_type:
## relationship – single -0.680 0.382 -1.43 0.0684 -1.782 0.075 -0.147 -0.309 0.01479
## married – single -0.975 0.441 -1.84 -0.1095 -2.210 0.027 -0.211 -0.398 -0.02368
## divorced – single -0.670 0.538 -1.73 0.3861 -1.244 0.214 -0.145 -0.373 0.08347
## widowed – single -1.071 0.559 -2.17 0.0252 -1.916 0.056 -0.231 -0.468 0.00546
## q04_children:
## yes – no -1.483 0.364 -2.20 -0.7683 -4.071 < .001 -0.321 -0.475 -0.16608
## q11_education:
## element – unfin_element 2.164 2.021 -1.80 6.1293 1.070 0.285 0.468 -0.389 1.32490
## unfin_hs – unfin_element 1.236 2.042 -2.77 5.2427 0.605 0.545 0.267 -0.599 1.13327
## hs – unfin_element 0.998 1.982 -2.89 4.8851 0.504 0.615 0.216 -0.625 1.05596
## undergrad – unfin_element 1.062 2.005 -2.87 4.9952 0.529 0.597 0.229 -0.621 1.07976
## postgrad – unfin_element 0.599 1.979 -3.28 4.4814 0.303 0.762 0.129 -0.710 0.96870
## q18_02_soc_media_info:
## no – yes -1.282 0.268 -1.81 -0.7563 -4.785 < .001 -0.277 -0.391 -0.16349
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## ᵃ Represents reference level
##
##
## ASSUMPTION CHECKS
##
## Durbin–Watson Test for Autocorrelation
## ────────────────────────────────────────────
## Autocorrelation DW Statistic p
## ────────────────────────────────────────────
## 0.0382 1.92 0.120
## ────────────────────────────────────────────
##
##
## Collinearity Statistics
## ──────────────────────────────────────────────
## VIF Tolerance
## ──────────────────────────────────────────────
## q01_gender 1.05 0.955
## q03_relationship_type 1.13 0.888
## q04_children 1.53 0.655
## q11_education 1.03 0.970
## q18_02_soc_media_info 1.04 0.958
## ──────────────────────────────────────────────
##
##
## MODEL 3
##
## Omnibus ANOVA Test
## ────────────────────────────────────────────────────────────────────────────────────────────
## Sum of Squares df Mean Square F p
## ────────────────────────────────────────────────────────────────────────────────────────────
## q01_gender 347.64 1 347.64 18.079 < .001
## q03_relationship_type 101.47 4 25.37 1.319 0.261
## q04_children 321.13 1 321.13 16.700 < .001
## q11_education 162.04 5 32.41 1.685 0.135
## q18_02_soc_media_info 400.55 1 400.55 20.831 < .001
## q40_smoking 47.46 1 47.46 2.468 0.116
## q42_sport 151.15 1 151.15 7.861 0.005
## q38_alcohol 24.07 1 24.07 1.252 0.263
## q35_01_contact_close_family 87.11 2 43.55 2.265 0.104
## q35_03_contact_friends 9.08 2 4.54 0.236 0.790
## Residuals 27727.79 1442 19.23
## ────────────────────────────────────────────────────────────────────────────────────────────
## Note. Type 3 sum of squares
##
##
## Model Coefficients - PHQ8
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Predictor Estimate SE Lower Upper t p Stand. Estimate Lower Upper
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Intercept ᵃ 7.1830 2.028 3.2048 11.1612 3.5419 < .001
## q01_gender:
## male – female -1.1365 0.267 -1.6607 -0.6122 -4.2520 < .001 -0.24566 -0.35899 -0.13232
## q03_relationship_type:
## relationship – single -0.6549 0.382 -1.4041 0.0942 -1.7149 0.087 -0.14157 -0.30350 0.02037
## married – single -0.8872 0.442 -1.7543 -0.0201 -2.0072 0.045 -0.19178 -0.37921 -0.00435
## divorced – single -0.5314 0.541 -1.5924 0.5296 -0.9825 0.326 -0.11487 -0.34421 0.11447
## widowed – single -0.9802 0.566 -2.0897 0.1294 -1.7328 0.083 -0.21187 -0.45172 0.02798
## q04_children:
## yes – no -1.4940 0.366 -2.2112 -0.7769 -4.0866 < .001 -0.32295 -0.47796 -0.16793
## q11_education:
## element – unfin_element 2.2846 2.023 -1.6836 6.2528 1.1294 0.259 0.49384 -0.36393 1.35161
## unfin_hs – unfin_element 1.2582 2.039 -2.7418 5.2582 0.6170 0.537 0.27197 -0.59267 1.13661
## hs – unfin_element 1.1801 1.983 -2.7090 5.0692 0.5952 0.552 0.25509 -0.58558 1.09575
## undergrad – unfin_element 1.3243 2.008 -2.6153 5.2639 0.6594 0.510 0.28626 -0.56533 1.13785
## postgrad – unfin_element 0.8562 1.982 -3.0308 4.7433 0.4321 0.666 0.18509 -0.65514 1.02531
## q18_02_soc_media_info:
## no – yes -1.2261 0.269 -1.7531 -0.6991 -4.5641 < .001 -0.26504 -0.37895 -0.15113
## q40_smoking:
## no – yes -0.5417 0.345 -1.2180 0.1346 -1.5711 0.116 -0.11709 -0.26329 0.02910
## q42_sport:
## yes – no -0.6723 0.240 -1.1426 -0.2019 -2.8037 0.005 -0.14532 -0.24699 -0.04365
## q38_alcohol:
## no – yes -0.2760 0.247 -0.7600 0.2079 -1.1188 0.263 -0.05966 -0.16427 0.04495
## q35_01_contact_close_family:
## 2 – 1 -0.0122 0.257 -0.5169 0.4924 -0.0475 0.962 -0.00264 -0.11173 0.10644
## 3 – 1 0.7723 0.385 0.0164 1.5282 2.0042 0.045 0.16694 0.00355 0.33033
## q35_03_contact_friends:
## 2 – 1 0.4077 0.613 -0.7942 1.6096 0.6654 0.506 0.08813 -0.17168 0.34794
## 3 – 1 0.2754 1.409 -2.4877 3.0386 0.1955 0.845 0.05953 -0.53775 0.65681
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## ᵃ Represents reference level
##
##
## ASSUMPTION CHECKS
##
## Durbin–Watson Test for Autocorrelation
## ────────────────────────────────────────────
## Autocorrelation DW Statistic p
## ────────────────────────────────────────────
## 0.0332 1.93 0.164
## ────────────────────────────────────────────
##
##
## Collinearity Statistics
## ────────────────────────────────────────────────────
## VIF Tolerance
## ────────────────────────────────────────────────────
## q01_gender 1.06 0.944
## q03_relationship_type 1.13 0.882
## q04_children 1.54 0.650
## q11_education 1.04 0.960
## q18_02_soc_media_info 1.05 0.952
## q40_smoking 1.03 0.972
## q42_sport 1.04 0.965
## q38_alcohol 1.05 0.956
## q35_01_contact_close_family 1.04 0.964
## q35_03_contact_friends 1.02 0.982
## ────────────────────────────────────────────────────
##
##
## MODEL 4
##
## Omnibus ANOVA Test
## ─────────────────────────────────────────────────────────────────────────────────────────────
## Sum of Squares df Mean Square F p
## ─────────────────────────────────────────────────────────────────────────────────────────────
## q01_gender 450.278 1 450.278 25.7967 < .001
## q03_relationship_type 237.479 4 59.370 3.4013 0.009
## q04_children 333.208 1 333.208 19.0897 < .001
## q11_education 147.134 5 29.427 1.6859 0.135
## q18_02_soc_media_info 433.458 1 433.458 24.8331 < .001
## q40_smoking 36.725 1 36.725 2.1040 0.147
## q42_sport 1.605 1 1.605 0.0919 0.762
## q38_alcohol 137.581 1 137.581 7.8821 0.005
## q35_01_contact_close_family 113.124 2 56.562 3.2405 0.039
## q35_03_contact_friends 21.632 2 10.816 0.6196 0.538
## q47_self_reporting_health 1067.666 4 266.916 15.2918 < .001
## q48_chronic_illness 0.680 1 0.680 0.0390 0.844
## q49_health_limitations 248.520 2 124.260 7.1189 < .001
## Residuals 25047.725 1435 17.455
## ─────────────────────────────────────────────────────────────────────────────────────────────
## Note. Type 3 sum of squares
##
##
## Model Coefficients - PHQ8
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Predictor Estimate SE Lower Upper t p Stand. Estimate Lower Upper
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Intercept ᵃ 11.0446 2.608 5.929 16.16033 4.235 < .001
## q01_gender:
## male – female -1.2957 0.255 -1.796 -0.79527 -5.079 < .001 -0.28008 -0.3882 -0.17191
## q03_relationship_type:
## relationship – single -0.7688 0.365 -1.485 -0.05276 -2.106 0.035 -0.16619 -0.3210 -0.01140
## married – single -1.1604 0.423 -1.990 -0.33095 -2.744 0.006 -0.25083 -0.4301 -0.07154
## divorced – single -0.9934 0.518 -2.010 0.02346 -1.916 0.056 -0.21474 -0.4346 0.00507
## widowed – single -1.9693 0.548 -3.044 -0.89416 -3.593 < .001 -0.42568 -0.6581 -0.19328
## q04_children:
## yes – no -1.5232 0.349 -2.207 -0.83934 -4.369 < .001 -0.32926 -0.4771 -0.18143
## q11_education:
## element – unfin_element 3.1013 1.932 -0.689 6.89110 1.605 0.109 0.67038 -0.1488 1.48958
## unfin_hs – unfin_element 1.5954 1.947 -2.224 5.41449 0.819 0.413 0.34485 -0.4807 1.17040
## hs – unfin_element 2.2107 1.896 -1.509 5.93028 1.166 0.244 0.47786 -0.3262 1.28189
## undergrad – unfin_element 2.4573 1.921 -1.310 6.22475 1.279 0.201 0.53116 -0.2832 1.34554
## postgrad – unfin_element 1.9542 1.895 -1.764 5.67239 1.031 0.303 0.42242 -0.3813 1.22614
## q18_02_soc_media_info:
## no – yes -1.2779 0.256 -1.781 -0.77489 -4.983 < .001 -0.27624 -0.3850 -0.16750
## q40_smoking:
## no – yes -0.4795 0.331 -1.128 0.16896 -1.451 0.147 -0.10365 -0.2438 0.03652
## q42_sport:
## yes – no -0.0711 0.234 -0.531 0.38871 -0.303 0.762 -0.01536 -0.1147 0.08402
## q38_alcohol:
## no – yes -0.6664 0.237 -1.132 -0.20079 -2.808 0.005 -0.14406 -0.2447 -0.04340
## q35_01_contact_close_family:
## 2 – 1 0.0349 0.245 -0.446 0.51629 0.142 0.887 0.00755 -0.0965 0.11160
## 3 – 1 0.9026 0.367 0.182 1.62349 2.456 0.014 0.19512 0.0393 0.35093
## q35_03_contact_friends:
## 2 – 1 0.6399 0.584 -0.507 1.78637 1.095 0.274 0.13831 -0.1095 0.38614
## 3 – 1 0.3221 1.344 -2.315 2.95908 0.240 0.811 0.06963 -0.5004 0.63963
## q47_self_reporting_health:
## excellent – very_bad -4.1412 1.794 -7.661 -0.62130 -2.308 0.021 -0.89515 -1.6560 -0.13430
## good – very_bad -3.4874 1.777 -6.973 -0.00214 -1.963 0.050 -0.75384 -1.5072 -4.64e-4
## neutral – very_bad -1.9461 1.774 -5.426 1.53345 -1.097 0.273 -0.42066 -1.1728 0.33147
## bad – very_bad 0.4448 1.790 -3.067 3.95698 0.248 0.804 0.09615 -0.6630 0.85534
## q48_chronic_illness:
## no – yes -0.0525 0.266 -0.574 0.46914 -0.197 0.844 -0.01134 -0.1241 0.10141
## q49_health_limitations:
## partially_limits – limits -1.1321 0.644 -2.396 0.13132 -1.758 0.079 -0.24472 -0.5178 0.02839
## no_limits – limits -1.9976 0.660 -3.292 -0.70351 -3.028 0.003 -0.43180 -0.7115 -0.15207
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## ᵃ Represents reference level
##
##
## ASSUMPTION CHECKS
##
## Durbin–Watson Test for Autocorrelation
## ────────────────────────────────────────────
## Autocorrelation DW Statistic p
## ────────────────────────────────────────────
## 0.0463 1.90 0.048
## ────────────────────────────────────────────
##
##
## Collinearity Statistics
## ────────────────────────────────────────────────────
## VIF Tolerance
## ────────────────────────────────────────────────────
## q01_gender 1.06 0.942
## q03_relationship_type 1.14 0.876
## q04_children 1.54 0.650
## q11_education 1.05 0.954
## q18_02_soc_media_info 1.05 0.950
## q40_smoking 1.04 0.966
## q42_sport 1.06 0.941
## q38_alcohol 1.06 0.946
## q35_01_contact_close_family 1.04 0.963
## q35_03_contact_friends 1.02 0.980
## q47_self_reporting_health 1.09 0.918
## q48_chronic_illness 1.20 0.834
## q49_health_limitations 1.19 0.844
## ────────────────────────────────────────────────────
##
##
## MODEL 5
##
## Omnibus ANOVA Test
## ────────────────────────────────────────────────────────────────────────────────────────────
## Sum of Squares df Mean Square F p
## ────────────────────────────────────────────────────────────────────────────────────────────
## q01_gender 380.8 1 380.83 22.820 < .001
## q03_relationship_type 148.8 4 37.20 2.229 0.064
## q04_children 53.6 1 53.62 3.213 0.073
## q11_education 43.0 5 8.60 0.515 0.765
## q18_02_soc_media_info 145.1 1 145.10 8.695 0.003
## q40_smoking 30.5 1 30.47 1.826 0.177
## q42_sport 10.6 1 10.63 0.637 0.425
## q38_alcohol 114.1 1 114.14 6.839 0.009
## q35_01_contact_close_family 40.0 2 20.01 1.199 0.302
## q35_03_contact_friends 35.1 2 17.57 1.053 0.349
## q47_self_reporting_health 1187.5 4 296.87 17.789 < .001
## q48_chronic_illness 17.9 1 17.88 1.071 0.301
## q49_health_limitations 307.3 2 153.63 9.206 < .001
## q02_age 1116.1 1 1116.10 66.878 < .001
## Residuals 23931.6 1434 16.69
## ────────────────────────────────────────────────────────────────────────────────────────────
## Note. Type 3 sum of squares
##
##
## Model Coefficients - PHQ8
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Predictor Estimate SE Lower Upper t p Stand. Estimate Lower Upper
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Intercept ᵃ 14.4098 2.58304 9.3429 19.4768 5.5786 < .001
## q01_gender:
## male – female -1.1931 0.24976 -1.6830 -0.7032 -4.7770 < .001 -0.25790 -0.3638 -0.1520
## q03_relationship_type:
## relationship – single -0.8067 0.35698 -1.5069 -0.1064 -2.2597 0.024 -0.17437 -0.3257 -0.0230
## married – single -0.5262 0.42066 -1.3513 0.2990 -1.2508 0.211 -0.11373 -0.2921 0.0646
## divorced – single 0.1274 0.52510 -0.9027 1.1574 0.2426 0.808 0.02754 -0.1951 0.2502
## widowed – single 0.0381 0.58946 -1.1182 1.1944 0.0647 0.948 0.00824 -0.2417 0.2582
## q04_children:
## yes – no -0.6409 0.35755 -1.3423 0.0605 -1.7925 0.073 -0.13854 -0.2902 0.0131
## q11_education:
## element – unfin_element 1.9852 1.89403 -1.7302 5.7005 1.0481 0.295 0.42911 -0.3740 1.2322
## unfin_hs – unfin_element 1.1438 1.90452 -2.5921 4.8798 0.6006 0.548 0.24725 -0.5603 1.0548
## hs – unfin_element 1.7382 1.85502 -1.9006 5.3770 0.9370 0.349 0.37573 -0.4108 1.1623
## undergrad – unfin_element 1.6541 1.88054 -2.0348 5.3430 0.8796 0.379 0.35755 -0.4398 1.1549
## postgrad – unfin_element 1.6105 1.85389 -2.0261 5.2472 0.8687 0.385 0.34814 -0.4380 1.1342
## q18_02_soc_media_info:
## no – yes -0.7624 0.25856 -1.2696 -0.2552 -2.9486 0.003 -0.16480 -0.2744 -0.0552
## q40_smoking:
## no – yes -0.4368 0.32329 -1.0710 0.1973 -1.3513 0.177 -0.09443 -0.2315 0.0427
## q42_sport:
## yes – no -0.1832 0.22959 -0.6336 0.2672 -0.7980 0.425 -0.03961 -0.1370 0.0577
## q38_alcohol:
## no – yes -0.6073 0.23222 -1.0628 -0.1518 -2.6152 0.009 -0.13127 -0.2297 -0.0328
## q35_01_contact_close_family:
## 2 – 1 -0.1732 0.24130 -0.6465 0.3002 -0.7177 0.473 -0.03744 -0.1398 0.0649
## 3 – 1 0.3870 0.36481 -0.3286 1.1027 1.0609 0.289 0.08366 -0.0710 0.2383
## q35_03_contact_friends:
## 2 – 1 0.8162 0.57191 -0.3056 1.9381 1.4272 0.154 0.17643 -0.0661 0.4189
## 3 – 1 0.4104 1.31449 -2.1681 2.9890 0.3122 0.755 0.08872 -0.4687 0.6461
## q47_self_reporting_health:
## excellent – very_bad -4.8721 1.75682 -8.3184 -1.4259 -2.7733 0.006 -1.05316 -1.7981 -0.3082
## good – very_bad -3.9829 1.73835 -7.3929 -0.5729 -2.2912 0.022 -0.86094 -1.5980 -0.1238
## neutral – very_bad -2.3948 1.73529 -5.7988 1.0091 -1.3801 0.168 -0.51767 -1.2535 0.2181
## bad – very_bad -0.0856 1.75192 -3.5222 3.3510 -0.0489 0.961 -0.01850 -0.7614 0.7244
## q48_chronic_illness:
## no – yes -0.2706 0.26138 -0.7833 0.2422 -1.0351 0.301 -0.05848 -0.1693 0.0523
## q49_health_limitations:
## partially_limits – limits -1.1691 0.62980 -2.4046 0.0663 -1.8564 0.064 -0.25272 -0.5198 0.0143
## no_limits – limits -2.1602 0.64537 -3.4262 -0.8943 -3.3473 < .001 -0.46695 -0.7406 -0.1933
## q02_age -0.0726 0.00888 -0.0900 -0.0552 -8.1779 < .001 -0.32301 -0.4005 -0.2455
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## ᵃ Represents reference level
##
##
## ASSUMPTION CHECKS
##
## Durbin–Watson Test for Autocorrelation
## ────────────────────────────────────────────
## Autocorrelation DW Statistic p
## ────────────────────────────────────────────
## 0.0379 1.92 0.118
## ────────────────────────────────────────────
##
##
## Collinearity Statistics
## ────────────────────────────────────────────────────
## VIF Tolerance
## ────────────────────────────────────────────────────
## q01_gender 1.06 0.941
## q03_relationship_type 1.18 0.849
## q04_children 1.61 0.620
## q11_education 1.05 0.950
## q18_02_soc_media_info 1.09 0.921
## q40_smoking 1.04 0.966
## q42_sport 1.06 0.939
## q38_alcohol 1.06 0.946
## q35_01_contact_close_family 1.05 0.955
## q35_03_contact_friends 1.02 0.980
## q47_self_reporting_health 1.09 0.916
## q48_chronic_illness 1.21 0.829
## q49_health_limitations 1.19 0.843
## q02_age 1.71 0.585
## ────────────────────────────────────────────────────c. Stepwise regression model, step 1: algorithmic selection
As an alternative approach to the theory-derived, inductively build set of models, we choose to use the stepwise regression - combining forward with stepwise selection of the predictors. By using both of the Akaike information criterion (AIC) and Bayesian information criterion (BIC) to select the best-performing model, the algorithm from the MASS package arrives at a more simple model - 9 predictor variables using AIC and 7 predictor variables using BIC (as opposed to 14 with the previous manual approach). However, while this model performs well with this particular sample, there is a significant chance of underperformance on the international sample, since stepwise regression is prone to overfitting.
#We are using the MASS package, which contains stepAIC() function for stepwise regression model selection
#Dataset simplification to only the variables specified with hypotheses
data_linreg <- data[data_vars]
# Fit the full linear model using lm() function from base R
full.model_MASS <- lm(PHQ8 ~.,
data = data_linreg,
na.action = na.omit)
# Stepwise regression model using MASS package, ranks on AIC
step.model_AIC <- stepAIC(full.model_MASS,
direction = "both",
trace = FALSE)
# Stepwise regression model using MASS package, ranks on BIC
step.model_BIC <- stepAIC(full.model_MASS,
direction = "both",
trace = FALSE,
k = log(nrow(data_linreg)))
# Summary of model based on AIC
base::summary(step.model_AIC)##
## Call:
## lm(formula = PHQ8 ~ q02_age + q01_gender + q04_children + q40_smoking +
## q38_alcohol + q18_02_soc_media_info + q03_relationship_type +
## q47_self_reporting_health + q49_health_limitations, data = data_linreg,
## na.action = na.omit)
##
## Residuals:
## Min 1Q Median 3Q Max
## -11.408 -2.676 -0.792 1.814 19.958
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 9.94214 0.88010 11.30 < 2e-16 ***
## q02_age -0.07382 0.00839 -8.80 < 2e-16 ***
## q01_gendermale -1.14657 0.24372 -4.70 2.8e-06 ***
## q04_childrenno 0.69569 0.35378 1.97 0.04944 *
## q40_smokingno -0.48830 0.31777 -1.54 0.12459
## q38_alcoholno -0.57567 0.22900 -2.51 0.01205 *
## q18_02_soc_media_info.L -0.54033 0.18150 -2.98 0.00296 **
## q03_relationship_typerelationship -0.86034 0.35004 -2.46 0.01410 *
## q03_relationship_typemarried -0.60582 0.41427 -1.46 0.14386
## q03_relationship_typedivorced 0.06784 0.52148 0.13 0.89651
## q03_relationship_typewidowed 0.04565 0.58606 0.08 0.93792
## q47_self_reporting_healthgood 0.95560 0.28048 3.41 0.00067 ***
## q47_self_reporting_healthneutral 2.60654 0.36489 7.14 1.4e-12 ***
## q47_self_reporting_healthbad 4.93669 0.64552 7.65 3.7e-14 ***
## q47_self_reporting_healthvery_bad 5.03553 1.74612 2.88 0.00399 **
## q49_health_limitationspartially_limits -1.06210 0.61980 -1.71 0.08681 .
## q49_health_limitationsno_limits -2.16885 0.63396 -3.42 0.00064 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 4.08 on 1445 degrees of freedom
## (22 observations deleted due to missingness)
## Multiple R-squared: 0.23, Adjusted R-squared: 0.221
## F-statistic: 27 on 16 and 1445 DF, p-value: <2e-16# Summary of model based on BIC
base::summary(step.model_BIC)##
## Call:
## lm(formula = PHQ8 ~ q02_age + q01_gender + q04_children + q18_02_soc_media_info +
## q47_self_reporting_health + q49_health_limitations, data = data_linreg,
## na.action = na.omit)
##
## Residuals:
## Min 1Q Median 3Q Max
## -12.295 -2.602 -0.827 1.889 19.504
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 8.53669 0.78167 10.92 < 2e-16 ***
## q02_age -0.06945 0.00733 -9.47 < 2e-16 ***
## q01_gendermale -1.07301 0.23911 -4.49 7.8e-06 ***
## q04_childrenno 1.00656 0.29004 3.47 0.00053 ***
## q18_02_soc_media_info.L -0.55883 0.18096 -3.09 0.00205 **
## q47_self_reporting_healthgood 0.95620 0.28129 3.40 0.00069 ***
## q47_self_reporting_healthneutral 2.60287 0.36421 7.15 1.4e-12 ***
## q47_self_reporting_healthbad 4.88445 0.64414 7.58 6.0e-14 ***
## q47_self_reporting_healthvery_bad 4.76829 1.74961 2.73 0.00650 **
## q49_health_limitationspartially_limits -1.05367 0.61879 -1.70 0.08882 .
## q49_health_limitationsno_limits -2.12026 0.63212 -3.35 0.00082 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 4.1 on 1451 degrees of freedom
## (22 observations deleted due to missingness)
## Multiple R-squared: 0.22, Adjusted R-squared: 0.215
## F-statistic: 40.9 on 10 and 1451 DF, p-value: <2e-16d. Stepwise regression model, step 2: final AIC-driven model
In order to allow direct comparison with the manually-selected model, we input the chosen model (based on the AIC criterion) from the previous step into the linreg() function of the jmv package.
#To construct this regression model, we use the linReg() function from the jmv package.
jmv::linReg(
data = data,
dep = "PHQ8",
covs = "q02_age",
factors = vars("q01_gender",
"q40_smoking",
"q04_children",
"q38_alcohol",
"q18_02_soc_media_info",
"q03_relationship_type",
"q47_self_reporting_health",
"q49_health_limitations"),
blocks = list(
list(
"q02_age",
"q01_gender",
"q04_children",
"q47_self_reporting_health",
"q49_health_limitations",
"q18_02_soc_media_info"),
list(
"q38_alcohol",
"q40_smoking",
"q03_relationship_type")),
refLevels = list(
list(
var="q01_gender",
ref="male"),
list(
var="q04_children",
ref="yes"),
list(
var="q38_alcohol",
ref="yes"),
list(
var="q40_smoking",
ref="yes"),
list(
var="q18_02_soc_media_info",
ref="yes"),
list(
var="q03_relationship_type",
ref="single"),
list(
var="q47_self_reporting_health",
ref="very_bad"),
list(
var="q49_health_limitations",
ref="limits")),
r2Adj = TRUE,
aic = TRUE,
bic = TRUE,
rmse = TRUE,
modelTest = TRUE,
anova = TRUE,
ci = TRUE,
stdEst = TRUE,
ciStdEst = TRUE,
durbin = TRUE,
collin = TRUE)##
## LINEAR REGRESSION
##
## Model Fit Measures
## ───────────────────────────────────────────────────────────────────────────────────────────────────
## Model R R² Adjusted R² AIC BIC RMSE F df1 df2 p
## ───────────────────────────────────────────────────────────────────────────────────────────────────
## 1 0.469 0.220 0.215 8288 8351 4.08 40.9 10 1451 < .001
## 2 0.480 0.230 0.221 8281 8376 4.06 27.0 16 1445 < .001
## ───────────────────────────────────────────────────────────────────────────────────────────────────
##
##
## Model Comparisons
## ─────────────────────────────────────────────────────────────────
## Model Model ΔR² F df1 df2 p
## ─────────────────────────────────────────────────────────────────
## 1 - 2 0.0100 3.14 6 1445 0.005
## ─────────────────────────────────────────────────────────────────
##
##
## MODEL SPECIFIC RESULTS
##
## MODEL 1
##
## Omnibus ANOVA Test
## ─────────────────────────────────────────────────────────────────────────────────────────
## Sum of Squares df Mean Square F p
## ─────────────────────────────────────────────────────────────────────────────────────────
## q02_age 1508 1 1507.7 89.70 < .001
## q01_gender 338 1 338.5 20.14 < .001
## q04_children 202 1 202.4 12.04 < .001
## q47_self_reporting_health 1424 4 355.9 21.17 < .001
## q49_health_limitations 361 2 180.3 10.73 < .001
## q18_02_soc_media_info 160 1 160.3 9.54 0.002
## Residuals 24390 1451 16.8
## ─────────────────────────────────────────────────────────────────────────────────────────
## Note. Type 3 sum of squares
##
##
## Model Coefficients - PHQ8
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Predictor Estimate SE Lower Upper t p Stand. Estimate Lower Upper
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Intercept ᵃ 12.6271 1.77613 9.1431 16.1112 7.1094 < .001
## q02_age -0.0694 0.00733 -0.0838 -0.0551 -9.4708 < .001 -0.3089 -0.3729 -0.2449
## q01_gender:
## female – male 1.0730 0.23911 0.6040 1.5420 4.4875 < .001 0.2319 0.1306 0.3333
## q04_children:
## no – yes 1.0066 0.29004 0.4376 1.5755 3.4704 < .001 0.2176 0.0946 0.3406
## q47_self_reporting_health:
## excellent – very_bad -4.7683 1.74961 -8.2003 -1.3363 -2.7253 0.007 -1.0307 -1.7726 -0.2888
## good – very_bad -3.8121 1.73534 -7.2161 -0.4080 -2.1967 0.028 -0.8240 -1.5598 -0.0882
## neutral – very_bad -2.1654 1.73549 -5.5698 1.2389 -1.2477 0.212 -0.4681 -1.2040 0.2678
## bad – very_bad 0.1162 1.75383 -3.3241 3.5565 0.0662 0.947 0.0251 -0.7185 0.7688
## q49_health_limitations:
## partially_limits – limits -1.0537 0.61879 -2.2675 0.1601 -1.7028 0.089 -0.2278 -0.4901 0.0346
## no_limits – limits -2.1203 0.63212 -3.3602 -0.8803 -3.3542 < .001 -0.4583 -0.7263 -0.1903
## q18_02_soc_media_info:
## no – yes -0.7903 0.25591 -1.2923 -0.2883 -3.0882 0.002 -0.1708 -0.2793 -0.0623
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## ᵃ Represents reference level
##
##
## ASSUMPTION CHECKS
##
## Durbin–Watson Test for Autocorrelation
## ────────────────────────────────────────────
## Autocorrelation DW Statistic p
## ────────────────────────────────────────────
## 0.0394 1.92 0.100
## ────────────────────────────────────────────
##
##
## Collinearity Statistics
## ──────────────────────────────────────────────────
## VIF Tolerance
## ──────────────────────────────────────────────────
## q02_age 1.41 0.711
## q01_gender 1.01 0.986
## q04_children 1.30 0.767
## q47_self_reporting_health 1.07 0.934
## q49_health_limitations 1.14 0.875
## q18_02_soc_media_info 1.07 0.934
## ──────────────────────────────────────────────────
##
##
## MODEL 2
##
## Omnibus ANOVA Test
## ─────────────────────────────────────────────────────────────────────────────────────────
## Sum of Squares df Mean Square F p
## ─────────────────────────────────────────────────────────────────────────────────────────
## q02_age 1290.6 1 1290.6 77.46 < .001
## q01_gender 368.8 1 368.8 22.13 < .001
## q04_children 64.4 1 64.4 3.87 0.049
## q47_self_reporting_health 1422.5 4 355.6 21.34 < .001
## q49_health_limitations 378.5 2 189.3 11.36 < .001
## q18_02_soc_media_info 147.7 1 147.7 8.86 0.003
## q38_alcohol 105.3 1 105.3 6.32 0.012
## q40_smoking 39.3 1 39.3 2.36 0.125
## q03_relationship_type 175.8 4 44.0 2.64 0.033
## Residuals 24075.8 1445 16.7
## ─────────────────────────────────────────────────────────────────────────────────────────
## Note. Type 3 sum of squares
##
##
## Model Coefficients - PHQ8
## ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Predictor Estimate SE Lower Upper t p Stand. Estimate Lower Upper
## ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Intercept ᵃ 14.2132 1.81886 10.64527 17.7811 7.8143 < .001
## q02_age -0.0738 0.00839 -0.09027 -0.0574 -8.8011 < .001 -0.32837 -0.402 -0.2552
## q01_gender:
## female – male 1.1466 0.24372 0.66849 1.6247 4.7045 < .001 0.24784 0.145 0.3512
## q04_children:
## no – yes 0.6957 0.35378 0.00171 1.3897 1.9664 0.049 0.15038 3.69e-4 0.3004
## q47_self_reporting_health:
## excellent – very_bad -5.0355 1.74612 -8.46072 -1.6103 -2.8838 0.004 -1.08848 -1.829 -0.3481
## good – very_bad -4.0799 1.73172 -7.47688 -0.6830 -2.3560 0.019 -0.88192 -1.616 -0.1476
## neutral – very_bad -2.4290 1.73158 -5.82566 0.9677 -1.4028 0.161 -0.52505 -1.259 0.2092
## bad – very_bad -0.0988 1.74868 -3.52906 3.3314 -0.0565 0.955 -0.02137 -0.763 0.7201
## q49_health_limitations:
## partially_limits – limits -1.0621 0.61980 -2.27790 0.1537 -1.7136 0.087 -0.22958 -0.492 0.0332
## no_limits – limits -2.1689 0.63396 -3.41244 -0.9253 -3.4211 < .001 -0.46882 -0.738 -0.2000
## q18_02_soc_media_info:
## no – yes -0.7641 0.25668 -1.26765 -0.2606 -2.9770 0.003 -0.16518 -0.274 -0.0563
## q38_alcohol:
## no – yes -0.5757 0.22900 -1.02488 -0.1265 -2.5139 0.012 -0.12444 -0.222 -0.0273
## q40_smoking:
## no – yes -0.4883 0.31777 -1.11163 0.1350 -1.5367 0.125 -0.10555 -0.240 0.0292
## q03_relationship_type:
## relationship – single -0.8603 0.35004 -1.54698 -0.1737 -2.4578 0.014 -0.18597 -0.334 -0.0375
## married – single -0.6058 0.41427 -1.41846 0.2068 -1.4624 0.144 -0.13095 -0.307 0.0447
## divorced – single 0.0678 0.52148 -0.95510 1.0908 0.1301 0.897 0.01467 -0.206 0.2358
## widowed – single 0.0457 0.58606 -1.10396 1.1953 0.0779 0.938 0.00987 -0.239 0.2584
## ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## ᵃ Represents reference level
##
##
## ASSUMPTION CHECKS
##
## Durbin–Watson Test for Autocorrelation
## ────────────────────────────────────────────
## Autocorrelation DW Statistic p
## ────────────────────────────────────────────
## 0.0385 1.92 0.128
## ────────────────────────────────────────────
##
##
## Collinearity Statistics
## ──────────────────────────────────────────────────
## VIF Tolerance
## ──────────────────────────────────────────────────
## q02_age 1.62 0.619
## q01_gender 1.04 0.963
## q04_children 1.60 0.626
## q47_self_reporting_health 1.07 0.931
## q49_health_limitations 1.15 0.870
## q18_02_soc_media_info 1.08 0.927
## q38_alcohol 1.04 0.958
## q40_smoking 1.02 0.982
## q03_relationship_type 1.17 0.858
## ──────────────────────────────────────────────────4. Anxiety factor as a mediator for depression
a. Creation of the Anxiety index, step 1: overview of survey items
Aside from the regression model, we intend to explore the mediating role of anxiety between the consumption of social media and depression through a mediation/moderation analysis (in section 5).
In this section, we aim to construct an anxiety index from several survey items. As a first step, we select the survey items, which should be the manifestation of the latent factor of anxiety.
These survey items are:
| Survey question (1-10 scale) | Original variable | Renamed variable name |
|---|---|---|
| How scared are you of the risk of getting sick? | q30 | q30_anxiety_infection_covid |
| How scared are you of the risk that someone in your family or network of friends will get COVID-19? | q31 | q31_anxiety_infection_friends |
| I feel very anxious about the health emergency. | q33_01 | q33_01_anxiety_situation |
| I think I have little control over whether I get the infection. | q33_02 | q33_02_anxiety_low_control |
| I am scared that I will not be able to survive if I get sick due to COVID-19 or I got sick and I was scared that I would not survive. | q33_03 | q33_03_anxiety_survival_covid |
| I thought about quitting my job / dropping out of school due to COVID-19. | q33_04 | q33_04_anxiety_change_employment |
| I am afraid of transmitting the coronavirus to others. | q33_05 | q33_05_anxiety_infecting_others |
b. Creation of the Anxiety index, step 2: anxiety items descriptives
After the selection, we analyze these survey items with a set of descriptive statistics. To follow the established principles pertaining to the factor analyses, we also split the sample into two randomly chosen halves. The first half of the data set will be used for the Exploratory Factor Analysis, while the second half will be used by the Reliability and Confirmatory Factor Analyses.
jmv::descriptives(
data = data,
vars = vars("q30_anxiety_infection_covid",
"q31_anxiety_infection_friends",
"q33_01_anxiety_situation",
"q33_02_anxiety_low_control",
"q33_03_anxiety_survival_covid",
"q33_04_anxiety_change_employment",
"q33_05_anxiety_infecting_others"),
hist = TRUE,
min = FALSE,
max = FALSE)##
## DESCRIPTIVES
##
## Descriptives
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## q30_anxiety_infection_covid q31_anxiety_infection_friends q33_01_anxiety_situation q33_02_anxiety_low_control q33_03_anxiety_survival_covid q33_04_anxiety_change_employment q33_05_anxiety_infecting_others
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## N 1482 1482 1484 1484 1484 1484 1484
## Missing 2 2 0 0 0 0 0
## Mean 4.30 5.78 5.70 4.36 3.24 1.82 5.76
## Median 4.00 6.00 5.00 4.00 2.00 1.00 6.00
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
# We also split the sample into two halfs. The "training" half, on which we conduct the EFA analysis and "test" part, on which we test our construct through CFA.
set.seed(2021)
train_set <- data %>% slice_sample(n = 742)
test_set <- setdiff(data,train_set)c. Creation of the Anxiety index, step 3: Exploratory Factor Analysis of anxiety items
In the next step, we conduct an Exploratory Factor Analysis on these variables.
In line with best practices, we conduct the assumption checks (KMO and Bartlett’s Sphericity tests), set a cutoff for eigenvalue of >1 and hide factor loading below 0.4.
The result is therefore a one-factor construct, which includes all of the variables, except for the “q33_04_anxiety_change_employment”, which does not seem to be a good fit for the manifestation of anxiety within this group of variables.
#To conduct the EFA, we use the efa() function from the jmv package on the "train" data set (as opposed to the "test" dataset used for CFA)
jmv::efa(
data = train_set,
vars = vars("q30_anxiety_infection_covid",
"q31_anxiety_infection_friends",
"q33_01_anxiety_situation",
"q33_02_anxiety_low_control",
"q33_03_anxiety_survival_covid",
"q33_04_anxiety_change_employment",
"q33_05_anxiety_infecting_others"),
nFactorMethod = "eigen",
nFactors = 1,
minEigen = 1,
rotation = "promax",
hideLoadings = 0.4,
screePlot = TRUE,
factorSummary = TRUE,
kmo = TRUE,
bartlett = TRUE)##
## EXPLORATORY FACTOR ANALYSIS
##
## Factor Loadings
## ───────────────────────────────────────────────────────────
## 1 Uniqueness
## ───────────────────────────────────────────────────────────
## q30_anxiety_infection_covid 0.861 0.258
## q31_anxiety_infection_friends 0.813 0.340
## q33_01_anxiety_situation 0.577 0.667
## q33_02_anxiety_low_control 0.443 0.804
## q33_03_anxiety_survival_covid 0.454 0.794
## q33_04_anxiety_change_employment 0.953
## q33_05_anxiety_infecting_others 0.533 0.716
## ───────────────────────────────────────────────────────────
## Note. 'Minimum residual' extraction method was used
## in combination with a 'promax' rotation
##
##
## FACTOR STATISTICS
##
## Summary
## ──────────────────────────────────────────────────────────
## Factor SS Loadings % of Variance Cumulative %
## ──────────────────────────────────────────────────────────
## 1 2.47 35.3 35.3
## ──────────────────────────────────────────────────────────
##
##
## ASSUMPTION CHECKS
##
## Bartlett's Test of Sphericity
## ─────────────────────────────
## χ² df p
## ─────────────────────────────
## 1376 21 < .001
## ─────────────────────────────
##
##
## KMO Measure of Sampling Adequacy
## ─────────────────────────────────────────────
## MSA
## ─────────────────────────────────────────────
## Overall 0.779
## q30_anxiety_infection_covid 0.743
## q31_anxiety_infection_friends 0.719
## q33_01_anxiety_situation 0.883
## q33_02_anxiety_low_control 0.849
## q33_03_anxiety_survival_covid 0.823
## q33_04_anxiety_change_employment 0.730
## q33_05_anxiety_infecting_others 0.812
## ─────────────────────────────────────────────
d. Creation of the Anxiety index, step 3: Reliability Analysis of the index items
Secondly, we conduct a reliability analysis of the anxiety factor. We use a cutoff value of 0.7 for both McDonald’s omega and Cronbach’s Alpha. The scale passes this cutoff and the statistics would not be improved if any of the items were dropped.
#To conduct the reliability analysis, we use the reliability() function from the jmv package on the "test" data set (as opposed to the "train" dataset used for EFA)
jmv::reliability(
data = test_set,
vars = vars("q30_anxiety_infection_covid",
"q31_anxiety_infection_friends",
"q33_01_anxiety_situation",
"q33_02_anxiety_low_control",
"q33_03_anxiety_survival_covid",
"q33_05_anxiety_infecting_others"),
omegaScale = TRUE,
alphaItems = TRUE,
omegaItems = TRUE)##
## RELIABILITY ANALYSIS
##
## Scale Reliability Statistics
## ─────────────────────────────────────────
## Cronbach's α McDonald's ω
## ─────────────────────────────────────────
## scale 0.784 0.797
## ─────────────────────────────────────────
##
##
## Item Reliability Statistics
## ───────────────────────────────────────────────────────────────────
## Cronbach's α McDonald's ω
## ───────────────────────────────────────────────────────────────────
## q30_anxiety_infection_covid 0.719 0.726
## q31_anxiety_infection_friends 0.724 0.738
## q33_01_anxiety_situation 0.752 0.772
## q33_02_anxiety_low_control 0.761 0.782
## q33_03_anxiety_survival_covid 0.775 0.790
## q33_05_anxiety_infecting_others 0.773 0.784
## ───────────────────────────────────────────────────────────────────e. Creation of the Anxiety index, step 4: Confirmatory Factor Analysis of the index items
According to the commonly used cut-offs for estimating CFA fit, we report that the chi-square is 159 (p < 0.001) and the Standardized Root Mean Square Residual is 0.0521 (cut-off SRMR <0.08), which indicates a good fit. However, Root Mean Square Error of Approximation (90% CI) is 0.130-0.171 (cut-off < 0.08) and the Comparative Fit Index is 0.887 (cut-off ≥.90), which does not indicate a good-fit.
#To conduct the CFA, we use the cfa() function from the jmv package on the "test" data set (as opposed to the "train" dataset used for EFA)
jmv::cfa(
data = test_set,
factors = list(
list(
label="Anxiety",
vars=c(
"q30_anxiety_infection_covid",
"q31_anxiety_infection_friends",
"q33_01_anxiety_situation",
"q33_02_anxiety_low_control",
"q33_03_anxiety_survival_covid",
"q33_05_anxiety_infecting_others"))),
resCov = list(),
ci = TRUE,
stdEst = TRUE,
factCovEst = FALSE,
fitMeasures = c("cfi", "tli", "rmsea", "srmr"),
corRes = TRUE)##
## CONFIRMATORY FACTOR ANALYSIS
##
## Factor Loadings
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Factor Indicator Estimate SE Lower Upper Z p Stand. Estimate
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Anxiety q30_anxiety_infection_covid 1.99 0.0787 1.84 2.15 25.3 < .001 0.834
## q31_anxiety_infection_friends 2.08 0.0869 1.91 2.25 23.9 < .001 0.800
## q33_01_anxiety_situation 1.57 0.1015 1.37 1.77 15.4 < .001 0.566
## q33_02_anxiety_low_control 1.27 0.0962 1.08 1.45 13.1 < .001 0.495
## q33_03_anxiety_survival_covid 1.31 0.1037 1.11 1.51 12.6 < .001 0.480
## q33_05_anxiety_infecting_others 1.70 0.1198 1.46 1.93 14.2 < .001 0.534
## ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
##
##
## MODEL FIT
##
## Test for Exact Fit
## ───────────────────────
## χ² df p
## ───────────────────────
## 159 9 < .001
## ───────────────────────
##
##
## Fit Measures
## ───────────────────────────────────────────────────────
## CFI TLI SRMR RMSEA Lower Upper
## ───────────────────────────────────────────────────────
## 0.887 0.812 0.0521 0.150 0.130 0.171
## ───────────────────────────────────────────────────────
##
##
## POST-HOC MODEL PERFORMANCE
##
## Residuals for Observed Correlation Matrix
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## q30_anxiety_infection_covid q31_anxiety_infection_friends q33_01_anxiety_situation q33_02_anxiety_low_control q33_03_anxiety_survival_covid q33_05_anxiety_infecting_others
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## q30_anxiety_infection_covid 0.021 -0.028 -0.005 0.057 -0.075
## q31_anxiety_infection_friends -0.016 -0.084 -0.099 0.098
## q33_01_anxiety_situation 0.128 0.035 0.009
## q33_02_anxiety_low_control 0.120 0.026
## q33_03_anxiety_survival_covid -0.060
## q33_05_anxiety_infecting_others
## ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────f. Creation of Anxiety Index , step 5: creation and descriptives
After Reliability analysis and CFA, we combine the multiple variables into one named anxiety_index. We also render visualization and descriptive statistics for the new anxiety_index variable.
#Creating the anxiety index, consisting of the average the values of the multiple variables selected through factor analysis to represent the underlying construct of anxiety.
anxiety_index <- apply(cbind(data$q30_anxiety_infection_covid,
data$q31_anxiety_infection_friends,
data$q33_01_anxiety_situation,
data$q33_02_anxiety_low_control,
data$q33_03_anxiety_survival_covid,
data$q33_05_anxiety_infecting_others), 1, mean)
#Adding the vector as an column to the existing dataset.
data <- cbind(data, anxiety_index)
#To summarize the anxiety_index variable, we use the descriptives() function from the jmv package.
jmv::descriptives(
data = data,
missing = TRUE,
vars = "anxiety_index",
sd = TRUE,
median = FALSE,
quart = TRUE,
range = TRUE,
box = TRUE)##
## DESCRIPTIVES
##
## Descriptives
## ───────────────────────────────────────
## anxiety_index
## ───────────────────────────────────────
## N 1482
## Missing 2
## Mean 4.86
## Standard deviation 1.86
## Range 9.00
## Minimum 1.00
## Maximum 10.0
## 25th percentile 3.50
## 50th percentile 4.83
## 75th percentile 6.17
## ───────────────────────────────────────
5. Simplified age-moderated & anxiety-mediated model
a. Simple mediation model (without moderator)
To explore our hypothesized pathway (see H11) between social media exposure and depression, partially mediated by anxiety, we conduct a mediation analysis.
#To create the mediation model, we use the med() function from the medmod package.Before running the model, we need to transform the social media string dummy (yes/no) back to its numeric form.
levels(data$q18_02_soc_media_info) <- list("1" = "yes", "2" = "no")
data$q18_02_soc_media_info <- as.numeric(as.character(data$q18_02_soc_media_info))
# Centering continuous variables (without scaling)
data <- data %>%
mutate(PHQ8.c = scale(PHQ8, scale = FALSE)) %>%
mutate(q02_age.c = scale(q02_age, scale = FALSE)) %>%
filter(!is.na(anxiety_index)) %>%
mutate(anxiety_index.c = scale(anxiety_index, scale = FALSE)) %>%
mutate(q18_02_soc_media_info.c = scale(q18_02_soc_media_info,
scale = FALSE))
#Seed for reproducibility of the results (due to bootstrap)
set.seed(2021)
medmod::med(
data = data,
dep = "PHQ8.c",
med = "anxiety_index.c",
pred = "q18_02_soc_media_info.c",
estMethod = "bootstrap",
bootstrap = 1000,
ci = TRUE,
pm = TRUE,
paths = TRUE,
label = TRUE)##
## MEDIATION
##
## Mediation Estimates
## ──────────────────────────────────────────────────────────────────────────────────────────────────────
## Effect Label Estimate SE Lower Upper Z p % Mediation
## ──────────────────────────────────────────────────────────────────────────────────────────────────────
## Indirect a × b -0.0904 0.0748 -0.249 0.0532 -1.21 0.227 4.74
## Direct c -1.8179 0.2641 -2.313 -1.2869 -6.88 < .001 95.26
## Total c + a × b -1.9083 0.2737 -2.414 -1.3789 -6.97 < .001 100.00
## ──────────────────────────────────────────────────────────────────────────────────────────────────────
##
##
## Path Estimates
## ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## Label Estimate SE Lower Upper Z p
## ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
## q18_02_soc_media_info.c → anxiety_index.c a -0.136 0.1107 -0.358 0.0770 -1.23 0.220
## anxiety_index.c → PHQ8.c b 0.667 0.0645 0.546 0.8016 10.34 < .001
## q18_02_soc_media_info.c → PHQ8.c c -1.818 0.2641 -2.313 -1.2869 -6.88 < .001
## ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────b. Complex age-moderated mediation model
To account for the significant effect of age, which is presumed to influence both the social media exposure and the depression pathway, we build a complex model using the lavaan package, where age acts as an additional moderator.
#Mediation analysis (path analysis framework, SEM) using lavaan package.
#For reproducibility of results
set.seed(2021)
#First, we specify the pathways
spec_mod <- "
#Regressions
anxiety_index.c ~ a1*q18_02_soc_media_info.c + a2*q02_age.c + a3*q18_02_soc_media_info.c:q02_age.c
PHQ8.c ~ cdash1*q18_02_soc_media_info.c + cdash2*q02_age.c + cdash3*q18_02_soc_media_info.c:q02_age.c + b1*anxiety_index.c
#Mean of centered age moderator (for use in simple slopes)
q02_age.c ~ q02_age.c.mean*1
#Variance of centered age moderator (for use in simple slopes)
q02_age.c ~~ q02_age.c.var*q02_age.c
#Indirect effects conditional on moderator (a1 + a3*ModValue)*b1
indirect.SDbelow := (a1 + a3*(q02_age.c.mean-sqrt(q02_age.c.var)))*b1
indirect.SDabove := (a1 + a3*(q02_age.c.mean+sqrt(q02_age.c.var)))*b1
#Direct effects conditional on moderator (cdash1 + cdash3*ModValue)
direct.SDbelow := cdash1 + cdash3*(q02_age.c.mean-sqrt(q02_age.c.var))
direct.SDabove := cdash1 + cdash3*(q02_age.c.mean+sqrt(q02_age.c.var))
#Total effects conditional on moderator
total.SDbelow := direct.SDbelow + indirect.SDbelow
total.SDabove := direct.SDabove + indirect.SDabove
#Proportion mediated conditional on moderator
prop.mediated.SDbelow := indirect.SDbelow / total.SDbelow
prop.mediated.SDabove := indirect.SDabove / total.SDabove
#Index of moderated mediation
index.mod.med := a3*b1
"
#Secondly, we fit/estimate the model
fit_mod <- lavaan::sem(model = spec_mod,
data = data,
se = "bootstrap",
bootstrap = 1000)## Warning in lavaan::lavaan(model = spec_mod, data = data, se = "bootstrap", :
## lavaan WARNING: syntax contains parameters involving exogenous covariates;
## switching to fixed.x = FALSE#Finally, we summarize the model
lavaan::summary(fit_mod,
fit.measures = FALSE,
rsquare = TRUE,
standardize = TRUE)## lavaan 0.6-6 ended normally after 50 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of free parameters 18
##
## Number of observations 1482
##
## Model Test User Model:
##
## Test statistic 293.363
## Degrees of freedom 2
## P-value (Chi-square) 0.000
##
## Parameter Estimates:
##
## Standard errors Bootstrap
## Number of requested bootstrap draws 1000
## Number of successful bootstrap draws 1000
##
## Regressions:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## anxiety_index.c ~
## q18_02_ (a1) -0.031 0.148 -0.209 0.834 -0.031 -0.007
## q02_g.c (a2) -0.002 0.003 -0.668 0.504 -0.002 -0.020
## q18_02_ (a3) 0.010 0.007 1.472 0.141 0.010 0.046
## PHQ8.c ~
## q18_02_ (cds1) -0.517 0.284 -1.820 0.069 -0.517 -0.050
## q02_g.c (cds2) -0.062 0.006 -10.246 0.000 -0.062 -0.276
## q18_02_ (cds3) 0.041 0.015 2.668 0.008 0.041 0.074
## anxty_. (b1) 0.655 0.064 10.204 0.000 0.655 0.265
##
## Covariances:
## Estimate Std.Err z-value P(>|z|) Std.lv
## q18_02_soc_media_info.c ~~
## q18_02___.:02_ -1.447 0.125 -11.609 0.000 -1.447
## Std.all
##
## -0.391
##
## Intercepts:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## q02_g.c (q02_) 0.022 0.535 0.040 0.968 0.022 0.001
## .anxty_. -0.034 0.053 -0.642 0.521 -0.034 -0.018
## .PHQ8.c -0.136 0.120 -1.136 0.256 -0.136 -0.030
## q18_02_ 0.000 0.012 0.000 1.000 0.000 0.000
## q18_02_ 3.259 0.208 15.647 0.000 3.259 0.394
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## q02_g.c (q02_) 423.795 9.022 46.974 0.000 423.795 1.000
## .anxty_. 3.450 0.118 29.277 0.000 3.450 0.997
## .PHQ8.c 17.749 0.838 21.183 0.000 17.749 0.837
## q18_02_ 0.200 0.005 37.732 0.000 0.200 1.000
## q18_02_ 68.457 2.447 27.976 0.000 68.457 1.000
##
## R-Square:
## Estimate
## anxiety_indx.c 0.003
## PHQ8.c 0.163
##
## Defined Parameters:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## indirect.SDblw -0.161 0.085 -1.884 0.060 -0.161 -0.014
## indirect.SDabv 0.121 0.175 0.690 0.490 0.121 0.010
## direct.SDbelow -1.363 0.354 -3.846 0.000 -1.363 -0.124
## direct.SDabove 0.331 0.488 0.678 0.498 0.331 0.024
## total.SDbelow -1.524 0.363 -4.199 0.000 -1.524 -0.138
## total.SDabove 0.451 0.521 0.866 0.387 0.451 0.034
## prp.mdtd.SDblw 0.106 0.062 1.692 0.091 0.106 0.103
## prop.mdtd.SDbv 0.268 4.561 0.059 0.953 0.268 0.304
## index.mod.med 0.007 0.005 1.460 0.144 0.007 0.012#We can also vizalize the path diagram with lavaanPlot package
lavaanPlot(model = fit_mod)6. Visualizing the qualitative responses using Word Clouds
Part of the survey (q50) was dedicated to the comments of the respondents on their situation. To visualize this textual data, we use two pairs of two word clouds.
First Word Cloud pair visualizes the most common tokens and lemma (size and color represents frequency of the word). Second Word Cloud pair uses sentiment analysis technique to create two distinct word clouds (using only tokens, not lemma), one visualizes only words with positive emotional sentiment, while the second only words with negative sentiment.
a. Word Clouds of all textual responses, step 1: pre-processing and tokens Word Cloud
#Remove stop words - first, we load the public stop word list
stop_words_cz <- read_csv("https://raw.githubusercontent.com/stopwords-iso/stopwords-cs/master/stopwords-cs.txt", col_names = "word")
## Should the above link become obsolete, alternative source can be reached using "stopwords" library: stop_words_cz <- as_tibble_col(stopwords::stopwords("cs", source = "stopwords-iso"), column_name = "word")
#To be able to lemmatize tokens, we need to download Czech model
udmodel <- udpipe_download_model(language = "czech-pdt", overwrite = FALSE)
# Reshape the data frame into one column called "word"
tidy_dat <- gather(dplyr::as_tibble(data$q50), key, word) %>%
dplyr::select(word)
# Tokenize - one word per row of a dataframe/tibble
tokens <- tidy_dat %>%
unnest_tokens(word, word) %>%
dplyr::count(word, sort = TRUE) %>%
ungroup()
#Removing stop words by using anti_join() applied on the stop words list
tokens_clean <- tokens %>%
anti_join(stop_words_cz)
#Next, we remove numbers (optional step)
nums <- tokens_clean %>%
dplyr::filter(str_detect(word, "^[0-9]")) %>%
dplyr::select(word) %>% unique()
tokens_clean <- tokens_clean %>%
anti_join(nums, by = "word")
#We can also remove unique stop words that are still present (optional step)
uni_sw <- data.frame(word = c("např"))
tokens_clean <- tokens_clean %>%
anti_join(uni_sw, by = "word")
#Define a color palette for the Word Cloud
palette <- brewer.pal(8,"Dark2")
#Using seed for reproducibility of the Word Cloud look
set.seed(2021)
#We generate a Word Cloud using tokens
tokens_clean %>% with(wordcloud(word,
n,
random.order = FALSE,
scale = c(7,.5),
min.freq = 1,
max.words = 50,
colors = palette))
b. Word Clouds of all textual responses, step 2: pre-processing and lemma Word Cloud
#Lemmatization of tokens, using udpipe package
udpipe_tokens_lemma <- udpipe(x = tokens_clean$word,
object = udmodel)
tidy_dat_lemma <- gather(dplyr::as_tibble(udpipe_tokens_lemma$lemma),
key,
word) %>%
dplyr::select(word) %>%
dplyr::count(word, sort = TRUE)
tokens_clean_lemma <- tidy_dat_lemma %>%
anti_join(stop_words_cz)
#Using seed for reproducibility of the word cloud look
set.seed(2021)
#We generate a Word Cloud using lemma
tokens_clean_lemma %>% with(wordcloud(word,
n,
random.order = FALSE,
scale = c(6,.5),
min.freq = 1,
max.words = 50,
colors = palette))
c. Sentiment analysis of qualitative responses, step 1: pre-processing and Word Cloud with positive valency
#First, we load Czech Subjectivity Lexicon from ÚFAL MFF, which assesses sentiment for every word as positive or negative
sentiment_cz <- read_delim("https://lindat.mff.cuni.cz/repository/xmlui/bitstream/handle/11858/00-097C-0000-0022-FF60-B/sublex_1_0.csv?sequence=1&isAllowed=y",
"\t",
escape_double = FALSE,
col_names = FALSE,
trim_ws = TRUE) %>%
rename("word" = "X3", "sentiment" = "X4")
#Next, we create tidy tibble with tokens created in the previous section and we use inner_join function to separately save only the tokens with positive and negative valency
tokens_sentiment_positive <- tokens_clean_lemma %>%
inner_join(sentiment_cz %>%
filter(sentiment == "POS")) %>%
count(word, sort = TRUE)
tokens_sentiment_negative <- tokens_clean_lemma %>%
inner_join(sentiment_cz %>%
filter(sentiment == "NEG")) %>%
count(word, sort = TRUE)
#We generate a Word Cloud using only tokens with positive valency
#Using seed for reproducibility of the Word Cloud look
set.seed(5471)
tokens_sentiment_positive %>% with(wordcloud(word,
n,
random.order = FALSE,
scale = c(5,.5),
max.words = 50,
min.freq = 1,
colors = palette,
random.color = TRUE))
d. Sentiment analysis of qualitative responses, step 2: Word Cloud with negative valency
#Finally, a Word Cloud using only tokens with negative valency
#Using seed for reproducibility of the Word Cloud look
set.seed(5471)
tokens_sentiment_negative %>% with(wordcloud(word,
n,
random.order = FALSE,
scale = c(3,.5),
max.words = 50,
min.freq = 1,
colors = palette,
random.color = TRUE))
7. Bibliography
Ambresin, G., P. Chondros, C. Dowrick, H. Herrman, and J. M. Gunn. 2014. “Self-Rated Health and Long-Term Prognosis of Depression.” The Annals of Family Medicine 12 (1): 57–65. https://doi.org/10.1370/afm.1562.
Awaworyi Churchill, S., and L. Farrell. 2017. “Alcohol and Depression: Evidence from the 2014 Health Survey for England.” Drug and Alcohol Dependence 180 (November): 86–92. https://doi.org/10.1016/j.drugalcdep.2017.08.006.
Bendau, Antonia, Moritz Bruno Petzold, Lena Pyrkosch, Lea Mascarell Maricic, Felix Betzler, Janina Rogoll, Julia Große, Andreas Ströhle, and Jens Plag. 2020. “Associations between COVID-19 related media consumption and symptoms of anxiety, depression and COVID-19 related fear in the general population in Germany.” European Archives of Psychiatry and Clinical Neuroscience 1 (July): 3. https://doi.org/10.1007/s00406-020-01171-6.
Dhir, Amandeep, Yossiri Yossatorn, Puneet Kaur, and Sufen Chen. 2018. “Online social media fatigue and psychological wellbeing—A study of compulsive use, fear of missing out, fatigue, anxiety and depression.” International Journal of Information Management 40 (June): 141–52. https://doi.org/10.1016/j.ijinfomgt.2018.01.012.
Gloster, Andrew T., Demetris Lamnisos, Jelena Lubenko, Giovambattista Presti, Valeria Squatrito, Marios Constantinou, Christiana Nicolaou, et al. 2020. “Impact of COVID-19 pandemic on mental health: An international study.” Edited by Joel Msafiri Francis. PLOS ONE 15 (12): e0244809. https://doi.org/10.1371/journal.pone.0244809.
González-Sanguino, Clara, Berta Ausín, Miguel Ángel Castellanos, Jesús Saiz, Aída López-Gómez, Carolina Ugidos, and Manuel Muñoz. 2020. “Mental Health Consequences During the Initial Stage of the 2020 Coronavirus Pandemic (Covid-19) in Spain.” Brain, Behavior, and Immunity 87 (July): 172–76. https://doi.org/10.1016/j.bbi.2020.05.040.
Harvey, Samuel B., Simon Øverland, Stephani L. Hatch, Simon Wessely, Arnstein Mykletun, and Matthew Hotopf. 2018. “Exercise and the Prevention of Depression: Results of the Hunt Cohort Study.” American Journal of Psychiatry 175 (1): 28–36. https://doi.org/10.1176/appi.ajp.2017.16111223.
Hossain, Md. Tanvir, Benojir Ahammed, Sanjoy Kumar Chanda, Nusrat Jahan, Mahfuza Zaman Ela, and Md. Nazrul Islam. 2020. “Social and Electronic Media Exposure and Generalized Anxiety Disorder Among People During Covid-19 Outbreak in Bangladesh: A Preliminary Observation.” Edited by Amir H. Pakpour. PLOS ONE 15 (9): e0238974. https://doi.org/10.1371/journal.pone.0238974.
Jacob, Louis, Josep Maria Haro, and Ai Koyanagi. 2019. “Relationship Between Living Alone and Common Mental Disorders in the 1993, 2000 and 2007 National Psychiatric Morbidity Surveys.” Edited by Soraya Seedat. PLOS ONE 14 (5): e0215182. https://doi.org/10.1371/journal.pone.0215182.
Kowal, Marta, Tao Coll‐Martín, Gözde Ikizer, Jesper Rasmussen, Kristina Eichel, Anna Studzińska, Karolina Koszałkowska, et al. 2020. “Who is the Most Stressed During the COVID‐19 Pandemic? Data From 26 Countries and Areas.” Applied Psychology: Health and Well-Being 12 (4): 946–66. https://doi.org/10.1111/aphw.12234.
Kroenke, Kurt, Tara W. Strine, Robert L. Spitzer, Janet B. W. Williams, Joyce T. Berry, and Ali H. Mokdad. 2009. “The PHQ-8 as a measure of current depression in the general population.” Journal of Affective Disorders 114 (1-3): 163–73. https://doi.org/10.1016/j.jad.2008.06.026.
Krogh, Jesper, Carsten Hjorthøj, Helene Speyer, Christian Gluud, and Merete Nordentoft. 2017. “Exercise for patients with major depression: A systematic review with meta-analysis and trial sequential analysis.” BMJ Open 7 (9): e014820. https://doi.org/10.1136/bmjopen-2016-014820.
Kvam, Siri, Catrine Lykkedrang Kleppe, Inger Hilde Nordhus, and Anders Hovland. 2016. “Exercise as a treatment for depression: A meta-analysis.” Elsevier B.V. https://doi.org/10.1016/j.jad.2016.03.063.
Losada-Baltar, Andrés, Lucía Jiménez-Gonzalo, Laura Gallego-Alberto, María del Sequeros Pedroso-Chaparro, José Fernandes-Pires, and María Márquez-González. 2020. ““We Are Staying at Home.” Association of Self-Perceptions of Aging, Personal and Family Resources, and Loneliness with Psychological Distress During the Lock-down Period of Covid-19.” Edited by Derek Issacowitz. The Journals of Gerontology: Series B 76 (2): e10–e16. https://doi.org/10.1093/geronb/gbaa048.
Luo, Min, Lixia Guo, Mingzhou Yu, Wenying Jiang, and Haiyan Wang. 2020. “The Psychological and Mental Impact of Coronavirus Disease 2019 (Covid-19) on Medical Staff and General Public a Systematic Review and Meta-Analysis.” Psychiatry Research 291 (September): 113190. https://doi.org/10.1016/j.psychres.2020.113190.
Mertens, Gaëtan, Lotte Gerritsen, Stefanie Duijndam, Elske Salemink, and Iris M. Engelhard. 2020. “Fear of the Coronavirus (Covid-19): Predictors in an Online Study Conducted in March 2020.” Journal of Anxiety Disorders 74 (August): 102258. https://doi.org/10.1016/j.janxdis.2020.102258.
Primack, Brian A., Ariel Shensa, César G. Escobar-Viera, Erica L. Barrett, Jaime E. Sidani, Jason B. Colditz, and A. Everette James. 2017. “Use of multiple social media platforms and symptoms of depression and anxiety: A nationally-representative study among U.S. young adults.” Computers in Human Behavior 69 (April): 1–9. https://doi.org/10.1016/j.chb.2016.11.013.
Rasmussen, Eric E., Narissra Punyanunt-Carter, Jenna R. LaFreniere, Mary S. Norman, and Thomas G. Kimball. 2020. “The Serially Mediated Relationship Between Emerging Adults’ Social Media Use and Mental Well-Being.” Computers in Human Behavior 102 (January): 206–13. https://doi.org/10.1016/j.chb.2019.08.019.
Salk, Rachel H., Janet S. Hyde, and Lyn Y. Abramson. 2017. “Gender differences in depression in representative national samples: Meta-analyses of diagnoses and symptoms.” Psychological Bulletin 143 (8): 783–822. https://doi.org/10.1037/bul0000102.
Schuch, Felipe B., Davy Vancampfort, Justin Richards, Simon Rosenbaum, Philip B. Ward, and Brendon Stubbs. 2016. “Exercise as a treatment for depression: A meta-analysis adjusting for publication bias.” Journal of Psychiatric Research 77 (June): 42–51. https://doi.org/10.1016/j.jpsychires.2016.02.023.
Shevlin, Mark, Orla McBride, Jamie Murphy, Jilly Gibson Miller, Todd K. Hartman, Liat Levita, Liam Mason, et al. 2020. “Anxiety, Depression, Traumatic Stress and Covid-19-Related Anxiety in the Uk General Population During the Covid-19 Pandemic.” BJPsych Open 6 (6). https://doi.org/10.1192/bjo.2020.109.
Stanca, Luca. 2012. “Suffer the Little Children: Measuring the Effects of Parenthood on Well-Being Worldwide.” Journal of Economic Behavior & Organization 81 (3): 742–50. https://doi.org/10.1016/j.jebo.2010.12.019.
Stanton, Robert, Quyen G. To, Saman Khalesi, Susan L. Williams, Stephanie J. Alley, Tanya L. Thwaite, Andrew S. Fenning, and Corneel Vandelanotte. 2020. “Depression, Anxiety and Stress During Covid-19: Associations with Changes in Physical Activity, Sleep, Tobacco and Alcohol Use in Australian Adults.” International Journal of Environmental Research and Public Health 17 (11): 4065. https://doi.org/10.3390/ijerph17114065.
Stubbs, Brendon, Davy Vancampfort, Mats Hallgren, Joseph Firth, Nicola Veronese, Marco Solmi, Serge Brand, et al. 2018. “EPA guidance on physical activity as a treatment for severe mental illness: a meta-review of the evidence and Position Statement from the European Psychiatric Association (EPA), supported by the International Organization of Physical Therapists in Mental Health (IOPTMH).” European Psychiatry 54 (October): 124–44. https://doi.org/10.1016/j.eurpsy.2018.07.004.
Taylor, Melanie R, Kingsley E Agho, Garry J Stevens, and Beverley Raphael. 2008. “Factors Influencing Psychological Distress During a Disease Epidemic: Data from Australia’s First Outbreak of Equine Influenza.” BMC Public Health 8 (1). https://doi.org/10.1186/1471-2458-8-347.
Tull, Matthew T., Keith A. Edmonds, Kayla M. Scamaldo, Julia R. Richmond, Jason P. Rose, and Kim L. Gratz. 2020. “Psychological Outcomes Associated with Stay-at-Home Orders and the Perceived Impact of Covid-19 on Daily Life.” Psychiatry Research 289 (July): 113098. https://doi.org/10.1016/j.psychres.2020.113098.
Vannucci, Anna, Kaitlin M. Flannery, and Christine McCauley Ohannessian. 2017. “Social Media Use and Anxiety in Emerging Adults.” Journal of Affective Disorders 207 (January): 163–66. https://doi.org/10.1016/j.jad.2016.08.040.
Vindegaard, Nina, and Michael Eriksen Benros. 2020. “COVID-19 Pandemic and Mental Health Consequences: Systematic Review of the Current Evidence.” Brain, Behavior, and Immunity 89 (October): 531–42. https://doi.org/10.1016/j.bbi.2020.05.048.
Wang, Cuiyan, Riyu Pan, Xiaoyang Wan, Yilin Tan, Linkang Xu, Roger S. McIntyre, Faith N. Choo, et al. 2020. “A Longitudinal Study on the Mental Health of General Population During the Covid-19 Epidemic in China.” Brain, Behavior, and Immunity 87 (July): 40–48. https://doi.org/10.1016/j.bbi.2020.04.028.
Wheaton, Michael G., Alena Prikhidko, and Gabrielle R. Messner. 2021. “Is Fear of Covid-19 Contagious? The Effects of Emotion Contagion and Social Media Use on Anxiety in Response to the Coronavirus Pandemic.” Frontiers in Psychology 11 (January). https://doi.org/10.3389/fpsyg.2020.567379.