Naslov (srp)

Struktura post-traumatske simptomatologije i njena faktorska stabilnost : doktorska disertacija

Autor

Vidaković, Ivana M., 1973-

Doprinosi

Knežević, Goran, 1964-
Tošković, Oliver, 1977-
Čabarkapa, Milanko, 1953-
Nikolić Ristanović, Vesna, 1955-

Opis (srp)

Neslaganje vodećih autora po pitanju optimalne strukture post-traumatske simptomatologije i brojni neusaglašeni empirijski nalazi aktuelno rezultiraju u različitim konceptualizacijama post-traumatskog stresnog poremećaja (PTSP) u dijagnostičkim sistemima DSM 5 i ICD 11. U ovom istraživanju proveravana je struktura post-traumatske simptomatologije, kao i njena faktorska stabilnost, nezavisno od upotrebljene metode procene. Testirali smo najčešće navođene modele strukture: trofaktorski i dva četvorofaktorska modela, kao i neke njihove modifikacije. Kroz niz konfirmatornih faktorskih analiza proveravali smo podesnost strukturnih modela na tri instrumenta za procenu post-traumatske simptomatologije (CAPS, IES R, SCL90Rptsd). Analize su vršene multigrupno, na uzorku ispitanika sa iskustvima ratne traumatizacije (N=271) i uzorku ispitanika sa iskustvima ratne torture (N=324). Važan aspekt ove provere podesnosti modela je kontrola efekata metoda procene u proceduri konfirmatorne analize multi-crta-multi-metod matrice (MTMM). Na pojedinačnim instrumentima četvorofaktorski modeli pokazali su bolju podesnost od trofaktorskog DSM IV modela; model Zaravnjenosti se pokazao boljim od modela Disforije. Dozvoljavanje kovarijanse grešaka ajtema koji opisuju problem sa spavanjem dodatno je poboljšalo podesnost modela Zaravnjenosti, dok se petofaktorski model sa faktorima Nametanje, Izbegavanje, Zaravnjenost, Povišena pobuđenost i ajtemima poremećaja spavanja izdvojenim u poseban peti faktor pokazao istim ili boljim kroz sva tri pojedinačna instrumenta. Testiranje modela strukture post-traumatske simptomatologije kroz ukupne podatke sa sva tri instrumenta daje drugačije rezultate: trofaktorski DSM IV model dao je poželjne indekse podesnosti, bolje od četvorofaktorskih modela Zaravnjenosti i Disforije. Petofaktorski model Poremećaja spavanja i kroz ovu analizu pokazao je dobru podesnost, što ukazuje na sistematsku povezanost simptomatologije problema sa spavanjem van prostora definisanih sa tri ili četri standardno definisana faktora post-traumatske simptomatologije. Ovi nalazi ukazuju na važnost boljeg prepoznavanja i razumevanja problema sa spavanjem koji zaslužuju više pažnje u dijagnostici PTSP i imaju potvrđen značaj u nastanku i održavanju PTSP kao i u njegovom tretmanu. Analiza MTMM matrice pokazala je dobru konvergentnu ali slabu divergentnu validnost sindromskih konstrukata, i dominantan uticaj metodskog faktora pri određenju strukture PTSP. Naši rezultati pokazuju da se korišćenjem više instrumenata ne poboljšava diskriminativnost - sposobnost razlikovanja sindroma PTSP, i da se kroz više instrumenata restriktivniji/ekonomičniji model trofaktorske strukture PTSP pokazuje podjednako dobar kao četvorofaktorski modeli. Nakon eliminisanja metodskog faktora, kroz procenu na tri instrumenta, sindromi PTSP izrazito konvergiraju ka PTSP kao jedinstvenom predmetu merenja, jače nego u slučaju svake od pojedinačnih metoda procene. Ograničenja ovih nalaza odnose se na prirodu korišćenih instrumenata i upitnu konstrukt validnost određenih sindromskih faktora na instrumentima samoprocene. Preporuke ovog istraživanja odnose se i na kontrolisanje metodskog efekta u budućim studijama strukture PTSP kao i proveru konstrukt validnost latentnih faktora. Aktuelna revizija dijagnostičkih kriterijuma za PTSP je povod za reviziju instrumenata procene ali i prilika za prevazilaženje nekih metodoloških ograničenja na koje smo ovim istraživanjem ukazali.

Opis (srp)

Psihologija - Klinička psihologija / Psychology - Clinical psychology Datum odbrane: 7.9.2016.

Opis (eng)

Disagreement of leading authors regarding an optimum structure of the post-traumatic symptoms and numerous discordant empirical findings result in different conceptualisations of the PTSD (post-traumatic stress disorder) in diagnostic systems DSM 5 and ICD 11. The present study examined the structure of the post-traumatic symptoms, as well as its factorial stability, regardless of the assessment method used. We tested the most frequently reported three- and four-factor models, as well as some of their modifications. Through a series of confirmatory factorial analyses, we checked the suitability of structural models on three instruments for assessment of the post-traumatic symptoms (CAPS, IES R, SCL90Rptsd). Multi-group analyses were carried out on two samples of respondents with experiences of war traumatization (N = 271) and respondents with experiences of war torture (N = 324). An important aspect of this examination of the model fit is the control of the effects of assessment methods in the procedure of confirmatory analysis of Multitrait-Multimethod Matrix (MTMM). At the individual instruments four-factor models showed a better fit than the three-factor DSM IV model; the Numbing model has proved to be better than the Dysphoria model. Allowing error covariance of items that describe the problems with sleep has further improved the fit of the Numbing model, while the five-factor model with the factors Intrusion, Avoidance, Numbing, Hyperarousal, and the sleep disorder items gathered into a special fifth factor, proved to be the same or better in all three individual instruments. Testing of the post-traumatic symptomatology structure model through the overall data from all three instruments yields different results: the three-factor DSM IV model gave the desired fit indexes, better than the four-factor Numbing and Dysphoria models. The five-factor sleep disorder model showed a good fit through this analysis, too, indicating a systematic connection between the sleep disorder symptoms beyond the area defined by three or four standard factors of the post-traumatic symptomatology. These findings point to the importance of better recognition and understanding of the sleep disorders that deserve more attention in the diagnosis of the PTSD and have proven significance in the development and persistence of the PTSD, as well as in its treatment. Analysis of the MTMM matrix showed a good convergent, but a poor divergent validity of the syndromic constructs, and the dominant influence of the methodological factor in determining the structure of the PTSD. Our results show that the use of multiple instruments does not improve discrimination i.e. the ability to differentiate PTSD syndromes, and that through multiple instruments the more restrictive/economical model of three-factor structure of PTSD proves to be as good as the four-factor models. After elimination of methodological factors, by applying three instruments, PTSD syndromes strongly converge towards a single measure of PTSD, greater than in the case of each of the individual assessment methods. The limitations of these findings are related to the characteristics of instruments and questionable construct validity of certain syndromic factors in self-assessment instruments that have been used. The recommendations of this study are also related to controlling of the methodological effect in future studies of the PTSD structure as well as the verification of the construct validity of latent factors. The current revision of the diagnostic criteria for the PTSD is a cause for revision of the assessment instruments but also an opportunity to overcome some of the methodological limitations indicated in this study.

Jezik

srpski

Datum

2016

Licenca

Creative Commons licenca
Ovo delo je licencirano pod uslovima licence
Creative Commons CC BY-NC-ND 2.0 AT - Creative Commons Autorstvo - Nekomercijalno - Bez prerada 2.0 Austria License.

http://creativecommons.org/licenses/by-nc-nd/2.0/at/legalcode

Predmet

OSNO - Opšta sistematizacija naučnih oblasti, Psihologija

post-traumatska simptomatologija, struktura PTSP, CAPS, IES R, SCL90R, CFA, multi-crta multi-metod

OSNO - Opšta sistematizacija naučnih oblasti, Psihologija

post-traumatic symptomatology, PTSD structure, CAPS, IES R, SCL90R, CFA, multitrait-multimethod