Hierarchical dependencies

Each healthcare professional may have their own specific approaches to treatment.

Each healthcare professional may have to follow the internal policies at their own clinics.

These can result in patients benefiting slightly differently from a treatment depending on who is administering it and what are available resources.

And the effects of the treatment might be different at different times.

Within person networks

See Bringmann et al. 2013, https://doi.org/10.1371/journal.pone.0060188

See Bringmann et al. 2013, https://doi.org/10.1371/journal.pone.0060188

Note symptom W (worry) in both cases.

Between and within persons networks

Population network. Arrows represent significant symptom associations.

Population network. Arrows represent significant symptom associations.

Individual differences. Arrows represent significant inter-individual differences.

Individual differences. Arrows represent significant inter-individual differences.

There are multiple significant positive (green lines) and negative (red lines) symptom associations as well as self-loops (symptoms predicts itself) in the population. However, there is significant variation between people with respect to some associations (blue lines), including self-loops (note missing self-loop in E “pleasant event” symptom).

Research questions

Between people–cross-sectional

  • How are depression symptoms mapped in group X?
  • Are there group differences concerning network structure?

Within person–individual changes

  • What are the effects of an intervention for person X?
  • Does depression manifest differently in person X at different moments in the day?

Between-within–group level differences and individual change

  • What are the long-term effects of an intervention and whether the effects are similar across people
  • Are bridge comorbid symptoms causing disorder stability in the long-run?

Network analysis

Structure estimation

which nodes and edges to include in the model estimation?

Reminder: valid and reliable measurement instruments.

which statistical model to use?

Reminder: the 3-D data cube.

Network analysis

Network description

  • What is the network structure (question of topology)?
  • What are influential nodes in the network?
  • What are differences between groups or time points?

Network analysis

Node centrality

Strength* = absolute direct associations between symptoms

Closeness = how fast can one symptom influence and be influenced by others

Betweenness = how many times is one symptom in the path of two other symptoms

Expected influence* = positive and negative direct associations are accommodated

Network analysis

Network stability

  • Accuracy of edge estimation–confidence intervals
  • Stability of network structure–the CS coefficient

The CS (centrality stability) = 0.5 is the minimal threshold (estimated maximum number of cases that can be dropped from the data to retain, with 95% probability, a correlation of at least 0.70 with the original data).

Original study

Data were collected via the online platform Duelocovid used to deliver self-applied interventions.

The platform was operational in multiple countries in Latin America.

For the study and this seminar data from Mexican participants used.

Paper currently under review (Herdoiza-Arroyo, Stanciu, de la Rosa-Gomez, Martinez-Arriaga, & Dominguez-Rodriguez. (2025). A psychological network approach to depression symptoms and sleep difficulties among adults going through grief during COVID-19: A cross-sectional study).

Depression assessment

Depression symptoms were measured with the Center for Epidemiologic Studies Depression Scale Revised (CESD-R) instrument (Eaton et al., 2004).

Below is a list of the ways you might have felt of behaved. Please tell me how often you have felt this way during the past week.

  • 0 = not at all or less than one a day
  • 1 = 1-2 days
  • 2 = 3-4 days
  • 3 = 5-7 days
  • 4 = nearly every day for two weeks

20 items. Examples:

  1. I was bothered by things that usually don’t bother me.
  2. I felt hopeful about the future.

See the CESD-R instrument here.

Sleep deprivation assessment

Sleep quality was measured with the Pittsburgh Sleep Quality Index (PSQI) instrument (Faulkner & Sidey-Gibbons, 2019).

The following questions relate to your usual sleep habits during the past month only. Your answers should indicate the most accurate reply for the majority of days and nights in the past month. Please answer all questions. During the past month.

  • 0 = not during the past month
  • 1 = less than once a week
  • 2 = once or twice a week
  • 3 = three or more times a week

11 items used. Examples:

  1. …how often have you had trouble sleeping beucase you feel too cold
  2. …how much of a problem has it been for you to keep up enthusiasm to get things done?

See the PSQI instrument here.

Note

For this seminar, a subset of n = 500 were randomly chosen from the original data.

The participant ID, gender and age group were simulated at random. Only the data on the CESD-R and PSQI are real.

For this seminar:

  • age_group contains three values: 1 = young, 2 = middle aged and 3 = old
  • gender contains two values: 1 = men and 2 = women

Thus, results of group comparisons or any analyses that go beyond examinations of CESD-R and PSQI disorders are most likely going to be nonsensical!

r is a programming language specifically developed for statistics. It can be used via the console panel in RStudio, which is an integrated development environment (IDE)–it has a easy to use interface.

The R console

The R console

An Integrated Development Environment (IDE).

An Integrated Development Environment (IDE).

  1. Install r and the r-console:

https://cran.r-project.org/

  1. Install RStudio:

https://posit.co/downloads/

Work with r scripts to save your progress. A new script will be created in the source panel in RStudio (the upper-left corner).

Copy-paste the provided code in a new script and modify according to your research question.

Do it one step at a time to understand what the code does.

Create an r script file

Create an r script file

Write code directly in the r script

Write code directly in the r script

Code 
# no object created
2+2
[1] 4
Code 
# object is first created and then run
sum<-2+2
sum
[1] 4
Code 
# create a simple data table
df<-data.frame(col1=c(1,3,66,9,121),
                  col2=c("A","Ab","This or that","C","d"))
df
  col1         col2
1    1            A
2    3           Ab
3   66 This or that
4    9            C
5  121            d

Steps 1 - 3

Understand the data and the limitations

Come up with a research question and testable hypotheses

  • Are you interested in highly influential symptoms?
  • Are you interested in group comparisons?
  • Are you interested in specific edges?
  • Are you interested in one or both disorders?

What quality checks do you need to consider

  • Valid and instrument measures – see established measures and scale reliability
  • Network stability – see the CS coefficient
  • Edge accuracy – see the 95% CI

Steps 4 - 6

What do you need for model estimation

Do you have evidence for your research question?

  • Are the results significant?
  • What study limitations can you think of?
  • What data limitations can you think of?

What does it all mean in practice?

  • What would it mean for therapy settings?
  • What would it mean for fundamental research?

References

https://adrianstanciu.eu/

Eaton, W. W., Smith, C., Ybarra, M., Muntaner, C., & Tien, A. (2004). Center for epidemiologic studies depression scale: Review and revision (CESD and CESD-r). In M. E. Maruish (Ed.), The use of psychological testing for treatment planning and outcomes assessment: Instruments for adults (pp. 363–377). Lawrence Erlbaum Associates Publishers.
Faulkner, S., & Sidey-Gibbons, C. (2019). Use of the pittsburg sleep quality index in people with schizophrenia spectrum disorders: A mixed methods study. Frontiers in Psychiatry, 10. https://doi.org/10.3389/fpsyt.2019.00284