first year presentation

Views:
 
     
 

Presentation Description

No description available.

Comments

Presentation Transcript

Evaluating a Profile Similarity Approach to Personality Pathology: 

Evaluating a Profile Similarity Approach to Personality Pathology Erika Carlson Wake Forest University

The 4 P’s: 

The 4 P’s Personality Profiles Pathology Prototypes

Personality and Profiles: 

Personality and Profiles Domain Facets NEO-PI-R score

Elements of a personality profile: 

Elements of a personality profile Scatter Shape Elevation

Pathology & prototypes : 

Pathology andamp; prototypes Categorical vs. Dimensional approach Normal personality Pathology Define pathology in terms of normal personality Lynam andamp; Widiger (2001) 10 expert-generated personality disorder prototypes Assessment Profile matching

Prototype profile matching: 

Prototype profile matching N E O A C

Prototype profile matching: 

Prototype profile matching

What are the elements of profile similarity?: 

What are the elements of profile similarity?

Elements of profile similarity: 

Elements of profile similarity

Similar in elevation: 

Similar in elevation

Similar in scatter : 

Similar in scatter

Similar in shape: 

Similar in shape

Elements within an omnibus index: Intraclass correlation: 

Shape Scatter andamp; Elevation andamp; Intraclass Correlation Elements within an omnibus index: Intraclass correlation x=profile y=prototype

My research questions: 

My research questions 1) How much information does each element of profile similarity provide in the context of personality disorders? 2) Does combining the 3 elements into one omnibus index obscure similarity information?

Methods: 

Methods Participants 219 Wake Forest University undergraduates in Introductory Psychology classes Measures Revised NEO Personality Inventory (NEO-PI-R) Normal personality (30 facets of the 5 domains) 240 items Millon Clinical Mulitaxial Inventory-III (MCMI) 10 DSM-IV personality disorder scales 175 items

Methods : 

Methods Step 1: Generate similarity elements from NEO-based profile correlation between participant’s profile and prototypes Shape index overall mean Elevation index overall variability Scatter index Step 2: Generate omnibus index from NEO-based profile intraclass correlations (ICC) between each participant’s profile and prototypes ICC index

Question 1How much information does each element of profile similarity provide?: 

Question 1 How much information does each element of profile similarity provide? Generated simultaneous regression equations for all 10 personality disorders (MCMI) Histrionic = Elevation + Scatter + Histrionic Shape

Question 1How much information does each element of profile similarity provide?: 

Question 1 How much information does each element of profile similarity provide? Personality Disorder Standard Regression Weights R2 βelevation βscatter βshape Histrionic .29*** .29*** -.02 .50*** (MCMI) Histrionic = Elevation + Scatter + Histrionic Shape

Question 1How much information does each element of profile similarity provide?: 

Question 1 How much information does each element of profile similarity provide? Personality Disorder Standard Regression Weights R2 βelevation βscatter βshape Schizoid .29*** -.20** .03 .50*** Avoidant .40*** -.12* .11† .64*** Dependent .30*** .05 -.01 .54*** Histrionic .29*** .29*** -.02 .50*** Narcissistic .37*** .01 .10 .60*** Antisocial .37*** -.11 -.03 .56*** Compulsive .51*** .28*** -.06 .63*** Schizotypal .30*** -.01 .17** .55*** Borderline .42*** .14* -.02 .68*** Paranoid .23*** -.09 .01 .46*** Mean .35 .02 .03 .57 * p andlt; .05 ** p andlt; .01 *** pandlt;.001 † pandlt;.10

Question 2Compare an omnibus index of profile similarity with elements: 

Question 2 Compare an omnibus index of profile similarity with elements Generated 2 hierarchical regression models for each of the 10 personality disorders Predicting MCMI scale scores from the 3 elements and the ICC omnibus index Model 1- Predicting MCMI-III scale score: Step 1: three similarity elements Step 2: intraclass correlation Model 2- Predicting MCMI-III scale score: Step 1: intraclass correlation Step 2: three similarity elements

Question 2Compare an omnibus index of profile similarity with elements: 

Question 2 Compare an omnibus index of profile similarity with elements Model 1 Model 2 Step 1 Step 2 Step 1 Step 2 Indices R2 ICC R2 ΔR2 ICC R2 Indices R2 ΔR2 Schizoid .29*** .29*** .00 .28*** .32*** .04* Avoidant .40*** .40*** .00 .39*** .40*** .01 Dependent .30*** .31*** .00 .28*** .31*** .03* Histrionic .29*** .29*** .01 .21*** .30*** .09*** Narcissistic .37*** .41*** .04*** .35*** .41*** .06*** Antisocial .37*** .37*** .00 .36*** .37*** .01 Compulsive .51*** .51*** .01 .38*** .51*** .13*** Schizotypal .30*** .30*** .00 .25*** .30*** .06*** Borderline .43*** .43*** .00 .40*** .44*** .04** Paranoid .23*** .23*** .00 .23*** .23*** .00 Mean ΔR2 .01 .05 * p andlt; .05 ** p andlt; .01 *** pandlt;.001 † pandlt;.10

Question 2Compare an omnibus index of profile similarity with elements: 

Question 2 Compare an omnibus index of profile similarity with elements Generated 2 hierarchical regression models for each of the 10 personality disorders Predicting MCMI scale scores from the SHAPE element and the ICC omnibus index Model 1- Predicting MCMI-III scale score: Step 1: SHAPE element Step 2: intraclass correlation Model 2- Predicting MCMI-III scale score: Step 1: intraclass correlation Step 2: SHAPE element

Question 2Compare an omnibus index of similarity with SHAPE: 

Question 2 Compare an omnibus index of similarity with SHAPE Model 1 Model 2 Step 1 Step 2 Step 1 Step 2 Shape R2 ICC R2 ΔR2 ICC R2 Shape R2 ΔR2 Schizoid .25*** .28*** .03** .28*** .31*** .03** Avoidant .38*** .39*** .01† .39*** .39*** .00 Dependent .30*** .31*** .01 .28*** .30*** .03** Histrionic .21*** .22*** .00 .21*** .22*** .00 Narcissistic .36*** .36*** .00 .35*** .36*** .01† Antisocial .35*** .36*** .01† .36*** .37*** .00 Compulsive .44*** .47*** .03*** .38*** .47*** .09*** Schizotypal .28*** .28*** .00 .25*** .28*** .03** Borderline .41*** .41*** .00 .40*** .41*** .01* Paranoid .22*** .23*** .01 .23*** .23*** .00 * p andlt; .05 ** p andlt; .01 *** pandlt;.001 † pandlt;.10

Implications: 

Implications Profile similarity measures current methods Omnibus index vs. elements categorical vs. dimensional approach to pathology Assessment

Limitations and Future Directions: 

Limitations and Future Directions Limitation Non-clinical sample Future Directions Explore each element Replication Cohen’s D2 and other omnibus indices

Slide26: 

Thank you! Special thanks to: Dr. R. Michael Furr Kris Gauthier Richard King Paul Hussman

Cluster A Prototypes : 

Cluster A Prototypes

Cluster B Prototypes: 

Cluster B Prototypes

Cluster C prototypes: 

Cluster C prototypes

Elements of profile similarity within omnibus measures: 

Shape Scatter Elevation Cohen’s D2 Elements of profile similarity within omnibus measures

authorStream Live Help