Networks vs latent variables Eiko Fried SEM 2
Networks vs latent variables Eiko Fried SEM 2, Uv. A, 22. 05. 2017
Why do certain things co-occur Flock of birds: https: //www. youtube. com/watch? v=Dm. O 4 Ellgmd 0 2
What are psychological constructs • Intelligence • Personality • Mental disorders 3
What are psychological constructs Answering this question requires both a theory and the appropriate statistical model. 4
Theoretical models PCs are: • Natural kinds: categories discovered – Elements • Social kinds: categories produced • Complex kinds: homeostatic property clusters – Species • Practical kinds: categories should be useful, not real – SES 5
Statistical models PCs can be modeled as: • Reflective latent variable models • Formative latent variable models • Network models 6
Statistical models 1 Reflective latent variable models; implied by natural kinds – Schizophrenia, intelligence, extraversion – While natural kinds imply reflective model, opposite not true (e. g. rich) 7
Statistical models 1 – Can in principle be discovered (latent cause) – Assumes realist ontology (“depression = brain disorder” is not just metaphor) 8
Statistical models 2 Formative latent variable models – SES 9
Statistical models 3 Network model: implied model for complex kinds – Intelligence, personality, mental disorders 10
Statistical models 3 Network model: implied model for complex kinds – Intelligence, personality, mental disorders 11
WHY SYMPTOMS CLUSTER IN SYNDROMES 12
Mental disorders 1. Why do researchers (implicitly) think about them as natural kinds 2. Why do researchers predominantly use factor models to study them 13
Natural kinds • Robert Koch, 1905: discovery that specific diseases have specific causative agents (tuberculosis & syphilis) • Diseases understood as natural kinds • Measles: infection of the respiratory system caused by a specific virus, accompanied by specific symptoms like red eyes, fever, generalized rash, and Koplik's spots • Gold: atomic number 79, and everything with this atomic number is gold 14
Natural kinds • 1910: discovery of syphilitic bacteria in brains of deceased patients diagnosed with "general paralysis of the insane" – Neuropsychiatric syndrome of late-stage syphilis – Clear "essence" identified for a mental disorder – Disease model applied to the rest of medicine, including psychiatry 15
Mental disorders as natural kinds • The assumption of mental disorders as natural kinds has been present throughout the history of psychiatry • Gerald Klerman, chief of the US national mental health agency, 1978: "there is a boundary between the normal and the sick" "there are discrete mental disorders" • Aim of developing specific treatments for particular disorders, and of finding specific underlying biological abnormalities • Diagnoses are categorical (you have schizophrenia or you do not) 16
Common cause framework • Derived from the notion of natural kinds • Disorders itself are latent—we cannot observe measles directly M 17
Common cause framework • Derived from the notion of natural kinds • Disorders itself are latent—we cannot observe measles directly • We can only observe the symptoms of measles • We can use symptoms to indicate the presence of measles s 1 M s 2 s 3 18
Common cause framework • Derived from the notion of natural kinds • Disorders itself are latent—we cannot observe measles directly • We can only observe the symptoms of measles • We can use symptoms to indicate the presence of measles – This works because measles causes measles symptoms s 1 M s 2 s 3 19
Common cause framework • The CC framework is responsible for symptom checklists in the rest of medicine and psychiatry – We use symptom lists to determine the presence of an underlying disease • The CC framework explains why symptoms cluster: they have the same causal origin – Fever, generalized rash, Koplik's spots measles! s 1 M s 2 s 3 20
Common cause framework • What does this mean for symptoms? – Symptoms are equivalent & interchangeable indicators of underlying disease ("assumption of symptom equivalence") – Symptom number, not symptom nature is relevant – Symptoms are "locally independent"; since they are derived from the same common cause, their correlations are spurious 21
Network perspective • Common cause: symptoms cluster because of a shared origin • Network: symptoms cluster bc they influence each other 22
Network perspective • Symptoms as separate entities that differ in important aspects • More than interchangeable indicators of underlying disorder 23
But it’s more complicated than that Common causes for some sets of symptoms? (https: //osf. io/mh 3 cf/) – Trauma → PTSD symptoms – Particular life events → particular depression symptoms 24
But it’s more complicated than that Common causes for some sets of symptoms? (https: //osf. io/mh 3 cf/) – Trauma → PTSD symptoms – Particular life events → particular depression symptoms 25
Conclusion – Think about theory before you move on to modeling, and consider what theory is implied by your model – Distinguish theoretical from statistical model – Network and latent variable models offer different insights and are complimentary, not competing – The network theory (e. g. of mental disorders) and the common cause theory are competing, and they cannot both be fully true at the same time (although both can be true to a certain degree simultaneously; e. g. PTSD onset could be a common cause, PTSD maintenance a network) 26
References General Fried, E. I. (2017). What are psychological constructs? On the nature and statistical modeling of emotions, intelligence, personality traits and mental disorders. Health Psychology Review, 11(2), 130– 134. http: //doi. org/10. 1080/17437199. 2017. 1306718 Personality Mõttus, R. , & Allerhand, M. (2017). Why do traits come together? The underlying trait and network approaches. In V. Zeigler-Hill & T. Shackelford (Eds. ), SAGE handbook of personality and individual differences: Volume 1. The science of personality and individual differences (pp. 1– 22). London: SAGE. Intelligence van der Maas, H. L. J. , Kan, K. , Marsman, M. , & Stevenson, C. E. (2017). Network Models for Cognitive Development and Intelligence, (January), 1– 14. http: //doi. org/10. 20944/preprints 201701. 0107. v 1 27
References Psychopathology Borsboom, D. (2017). A network theory of mental disorders. World Psychiatry, 16. Fried, E. I. , van Borkulo, C. D. , Cramer, A. O. J. , Boschloo, L. , Schoevers, R. A. , & Borsboom, D. (2016). Mental disorders as networks of problems: a review of recent insights. Social Psychiatry and Psychiatric Epidemiology, 1, 1– 32. Cramer, A. O. J. , Waldorp, L. J. , van der Maas, H. L. J. , & Borsboom, D. (2010). Comorbidity: a network perspective. The Behavioral and Brain Sciences, 33(2– 3), 137– 50. http: //doi. org/10. 1017/S 0140525 X 09991567 28
www. psych-networks. com 29
Psychological Dynamics 30
Networks! 31
- Slides: 31