Network Analysis: How to Run It Online and Report It in APA7
Psychological network analysis represents variables as nodes and their relationships as edges, usually regularised partial correlations, so you can see which variables are directly connected after accounting for all others.
What is Network Analysis?
Psychological network analysis represents variables as nodes and their relationships as edges, usually regularised partial correlations, so you can see which variables are directly connected after accounting for all others.
When to use it
Use it to explore the structure among many variables such as symptoms, attitudes or behaviours, to identify the most central nodes, and to compare network structures. It is exploratory and generates hypotheses.
Example question: Which of my eight symptom variables are most strongly connected, and which is the most central?
Preparing your data
One row per participant and one numeric column per node (variable). Continuous or ordinal variables are typical; all variables should be measured in the same sample.
Example dataset (simulated; first rows of 400): id, worry, tension, sleep, fatigue, focus, mood, interest, guilt (eight symptom scores).
| id | worry | tension | sleep | fatigue | focus | mood | interest | guilt |
|---|---|---|---|---|---|---|---|---|
| 1.0 | 4.78 | 2.27 | 2.23 | 3.9 | 2.29 | 2.93 | 3.44 | 2.6 |
| 2.0 | 2.95 | 3.38 | 1.07 | 2.86 | 3.0 | 3.0 | 2.54 | 4.01 |
| 3.0 | 3.85 | 3.24 | 3.22 | 4.55 | 3.09 | 2.52 | 4.18 | 3.7 |
| 4.0 | 3.23 | 4.19 | 3.47 | 2.93 | 1.87 | 4.98 | 5.51 | 2.75 |
| 5.0 | 3.62 | 3.9 | 2.19 | 2.58 | 3.45 | 4.46 | 3.38 | 3.4 |
| 6.0 | 3.37 | 3.52 | 2.18 | 3.26 | 2.14 | 2.33 | 3.09 | 3.97 |
Download the full example CSV to practise. If the tool asks for a different layout, follow the instructions on its page.
How to run it on StatFas
- Open the Network Analysis tool. You do not need an account to run an analysis.
- Enter your data or upload your file, laid out as described in Preparing your data.
- Select your variables. Select the variables that will form the nodes. Choose the estimation method (commonly the graphical lasso tuned with the extended Bayesian information criterion, EBIC) and a centrality measure.
- Choose any options the tool offers (for example the confidence level, a post hoc test or the error bar type) and run the analysis.
- Read the results on the page. Check the assumptions first, then the test statistic, the p-value and the effect size.
- To save your results, log in or create a free account. Every analysis stays free; the account is only needed to download.
Assumptions and checks
- A reasonably large sample relative to the number of nodes; stability falls quickly in small samples.
- Variables are measured at the same level and are not trivially overlapping (for example two near-identical items).
- Edges are partial correlations: they show direct statistical association, not causal direction, in cross-sectional data.
- Assess accuracy and stability by bootstrap: edge-weight confidence intervals and the correlation-stability (CS) coefficient for centrality, preferably above .50 (not below .25).
- Interpret centrality cautiously and compare centrality only when the stability checks are acceptable.
Key numbers to look at
- Network plot: node layout, edge thickness (strength) and colour (sign).
- Edge weights: regularised partial correlations.
- Centrality: strength (sum of absolute edge weights) is the most stable; closeness and betweenness are often less reliable.
- Stability: edge-weight bootstrap and the CS-coefficient.
Effect size
Edge weights (partial correlations) are the effect sizes; node strength summarises each variable's connectivity.
Benchmarks such as small, medium and large are conventions (Cohen, 1988), not rules. Interpret size in the context of your field.
Reporting Network Analysis in APA7 style
Template
Worked example (simulated data)
The numbers below were computed from the example CSV above, so you can reproduce them. They illustrate the format and are not real findings.
| Node | Strength (sum of |edge weights|) |
|---|---|
| mood | 0.43 |
| tension | 0.36 |
| interest | 0.36 |
| fatigue | 0.31 |
| focus | 0.28 |
| sleep | 0.23 |
| guilt | 0.16 |
| worry | 0.15 |
APA7 reminders: italicise statistical symbols (M, SD, t, F, p, r), report exact p-values to two or three decimals (use p < .001 for very small values), omit the leading zero for values that cannot exceed 1 (.45, not 0.45), and put degrees of freedom in parentheses. See the APA7 cheat sheet.
Common mistakes
- Interpreting edges as causal arrows.
- Reporting centrality without a stability check.
- Estimating a network of many nodes from a small sample.
Related analyses and alternatives
- Correlation: For pairwise associations without controlling for other variables.
- SEM: For confirmatory models with latent variables.
- PCA: For reducing variables to components.
Not sure which analysis you need? Use the test chooser.
Frequently asked questions
What is a partial correlation?
The correlation between two variables after removing the influence of all other variables in the network.
Why use regularisation?
It shrinks small edges to zero, which reduces false positives and gives a sparser, more interpretable network.
References
American Psychological Association. (2020). Publication manual of the American Psychological Association (7th ed.). https://doi.org/10.1037/0000165-000
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Lawrence Erlbaum Associates.
Epskamp, S., Borsboom, D., & Fried, E. I. (2018). Estimating psychological networks and their accuracy: A tutorial paper. Behavior Research Methods, 50(1), 195–212. https://doi.org/10.3758/s13428-017-0862-1
This guide is for learning and for preparing results for your own reports. You remain responsible for choosing an appropriate analysis, checking its assumptions and interpreting the results. Follow the formatting rules of your journal or institution where they differ from APA7.