Psychopathology Network Analysis Workshop Sacha Epskamp Eiko Fried

Psychopathology Network Analysis Workshop Sacha Epskamp Eiko Fried Department of Psychological Research Methods University of Amsterdam Utrecht University, October 11 2016

2

3

4

NETWORK REPLICABILITY CRISIS ? 5

Network stability • Why could replicability be an issue in psychopathological network research? • Let's write a brief network paper together to find out! • Data: 180 women with PTSD diagnosis, 17 -item screener • Data from DOI 10. 1037/a 0016227, freely available at https: //datashare. nida. nih. gov/protocol/nida-ctn-0015 6

Dataset 1 7

Dataset 1 8

Dataset 1 Paper • Strong positive connections between 3— 4, 5— 11, 16— 17 • Strong negative edge between 10— 12 • Most central nodes: 3, 16, 17 �consider as targets in intervention study 9

Dataset 1 Paper published … partytime! 10

Dataset 2 Now imagine we find another dataset, same sample size, female PTSD patients First dataset, n=180 Second dataset, n=179 11

Dataset 2 Now imagine we find another dataset, same sample size, female PTSD patients First dataset, n=180 Second dataset, n=179 12

Dataset 2 First dataset, n=180 Second dataset, n=179 13

Dataset 2 First dataset, n=180 Second dataset, n=179 14

Network stability • To avoid a replicability crisis, we need to investigate and report how accurate & stable our parameter estimates are • Especially relevant because our research may have clinical implications for thousands of patients – E. g. : what are the most central symptoms that ought to be treated? 15

Network stability Two main questions: • Stability of edge weights • Stability of centrality indices 16

Network stability Two main questions: • Is edge 3— 4 meaningfully larger than edge 3— 11? • Is node 17 substantially more central than node 16? 17

R-package bootnet

EDGE WEIGHT STABILITY 19

Boostrapping edge weights - Is edge 3— 4 (0. 42) stronger than edge 3— 11 (0. 14)? - Obtain CI by bootstrapping - Predictions? 20

Edges Edge weights 21

3— 4 0. 42 3— 11 Edges 0. 14 Edge weights 22

3— 4 0. 42 3— 11 Edges 0. 14 Edge weights 23

3— 4 0. 42 Edges 3— 11 0. 06 Edge weights 24

3— 4 0. 42 Edges 3— 11 0. 06 Most edges are not meaningfully different from each other because their CIs overlap. This is not really surprising: we are estimating 136 edge parameters with only 180 observations. Edge weights 25

CENTRALITY STABILITY 26

Subset bootstrap We now want to understand how stable the estimation of centrality indices is: e. g. , is centralty of node 17 (1. 16) substantially higher than the centrality of node 16 (0. 99) 27

Subset bootstrap • Unfortunately, bootstrapping CIs around centrality estimates is not possible • Costenbader, E. , & Valente, T. W. (2003) DOI: 10. 1016/S 0378 -8733(03)00012 -1 28

Subset bootstrap 1. Obtain centrality for data (s 17 > s 3 > s 16. . . ) 29

Subset bootstrap 1. Obtain centrality for data (s 17 > s 3 > s 16. . . ) 2. Subset data by dropping 10% of the people 3. Obtain centrality for -10% subset (s 17 > s 4. . . ) 30

Subset bootstrap 1. Obtain centrality for data (s 17 > s 3 > s 16. . . ) 2. Subset data by dropping 10% of the people 3. 4. 5. 6. Obtain centrality for -10% subset (s 17 > s 4. . . ) Subset data by dropping 20% of the people Obtain centrality for -20% subset (s 16 > s 7 > s 3. . . ). . . 31

Subset bootstrap So what we get is centrality for • Full data (s 17 > s 3 > s 16. . . ) • N -10% data (s 17 > s 4. . . ) • N -20% data (s 16 > s 7 > s 3. . . ) • N -30% data (s 17 > s 3 > s 16. . . ) • N -40% data (s 17 > s 3 > s 16. . . ) • N -50% data (s 16 > s 3 > s 7. . . ) • N -60% data (s 17 > s 3 > s 7. . . ) • N -70% data (s 17 > s 3 > s 16. . . ) • N -80% data (s 3 > s 6 > s 17. . . ) • N -90% data (s 7 > s 3 > s 16. . . ) 32

Subset bootstrap 33

Subset bootstrap • We can also subset nodes instead of people 34

Take home message • For most statistical parameters or test statistics, it is very useful to understand how precisely they are estimated – Different ways to do that, one way is to bootstrap confidence intervals around the point estimates • Investigating the stability of network parameters like edge weights will help us to understand how likely our networks generalize • bootnet is a very first & preliminary step: develop your own methods that help the field investigate the stability of networks 35

Take home message • Stability also helps guide the question how many observations we need to obtain stable networks n=180 for k=17 n=3812 for k=10 36

Eiko Fried Sacha Epskamp www. eiko-fried. com Twitter @Eiko. Fried www. sachaepskamp. com Twitter @Sacha. Epskamp Department of Psychological Research Methods University of Amsterdam

https: //arxiv. org/abs/1605. 09288 38

39

Latent Network Modeling 40

41

Residual Network Modeling 42

43

BFI Example 44

BFI Example 45

https: //arxiv. org/abs/1510. 06871 46

47

Time-varying Networks 48

49

https: //arxiv. org/abs/1604. 08045 50

51

52