Dyadic plans to model relations in social cooperation information

0
0
1721 days ago, 770 views
PowerPoint PPT Presentation
Dyadic outlines to model relations in social collaboration information. Todd D. Little Yale College. Layout. Why have such a symposium Dyadic Plans and Investigations Musings on Future Headings. Some Terrible Strategies. Dyad-level Setups (Overlook people) Target-Accomplice Setups

Presentation Transcript

Slide 1

Dyadic outlines to model relations in social connection information Todd D. Little Yale University

Slide 2

Outline Why have such a symposium Dyadic Designs and Analyses Thoughts on Future Directions

Slide 3

Some Bad Methods Dyad-level Setups (Ignore people) Target-Partner Setups Arbitrary task of target versus accomplice Loss of force Often thinks little of relations Ignores dyadic effect Target with numerous Partner Take normal of accomplices to diminish dyad-level impacts Doesn't generally do it Ignores dyadic effect

Slide 4

Intraclass Setups Represents focus with accomplice & join forces with focus in same information structure Exchangeable case (target/accomplice self-assertive) Distinguishable case (something efficient) Keeps dyadic impact Contains conditions Requires modification for exact measurable derivations (see e.g., Gonzalez & Griffin)

Slide 5

Between-Friend Correlations

Slide 6

Canonical Correlations Grade Child-Rated Parent-Rated Teacher-Rated

Slide 7

Social Relations Model (Kenny et al.) X ijk = m k + an i + b j + g ij + e ijk Where X ijk is the performer i's conduct with accomplice j at event k m k is a stupendous mean or catch an i is difference one of a kind to the on-screen character i b j is change remarkable to the accomplice j g ij is fluctuation one of a kind to the ij - dyad e ijk is mistake fluctuation Round-Robin plans: (n * (n-1)/2) Sample from every single conceivable connection Block plans: p people collaborate with q people Checker-board: various p's and q's of at least 2

Slide 8

Development .68 - .26 Onlooking - 25 Relative Ability to Compete Gender .51 Directives .39 Persistence - .27 .12 Imitation Tenure From Hawley & Little, 1999 SEM of a Block Design

Slide 9

Multilevel Approaches Distinguish HLM (a particular program) from progressive straight demonstrating, the procedure A non specific term for a sort of examination Probably best to talk about MRC(M) Modeling M ultilevel R andom C oefficient M odeling Different program usage HLM, MLn, SAS, BMDP, LISREL, and others

Slide 10

"Once you realize that chains of command exist, you see them everywhere." - Kreft and de Leeuw ( 1998 )

Slide 11

Logic of MRCM Coefficients depicting level 1 wonders are evaluated inside each level 2 unit (e.g., singular level impacts) Intercepts—implies Slopes—covariance/relapse coefficients Level 1 coefficients are likewise dissected at level 2 (e.g., dyad-level impacts) Intercepts: mean impact of dyad Slopes: impacts of dyad-level indicators

Slide 12

Negative Individual, Positive Group

Slide 13

Positive Individual, Negative Group

Slide 14

No Individual, Positive Group

Slide 15

No Group, Mixed Individual

Slide 16

A Contrived Example Y ij = Friendship Closeness appraisals of every individual i inside every dyad j . Level 1 Measures: Age & Social Skill of the individual members Level 2 Measures: Length of Friendship & Gender Composition of Friendship

Slide 17

The Level 1 Equation: y ij =  0j +  1j Age +  2j SocSkill +  3j Age*Skill + r ij The Level 2 Equations:  0j =  00 +  01 (Time) +  02 (Gnd) +  03 (Time*Gnd) + u 0j  1j =  10 +  11 (Time) +  12 (Gnd) +  13 (Time*Gnd) + u 1j  2j =  20 +  21 (Time) +  22 (Gnd) +  23 (Time*Gnd) + u 2j  3j =  30 +  31 (Time) +  32 (Gnd) +  33 (Time*Gnd) + u 3j The Equations

Slide 18

Future Directions OLS versus ML estimator and inclination Individual-arranged information versus dyad-situated information Thoughts on Future Directions

Slide 19

Level 1 Equations: Meaning of Intercepts Y = Friendship Closeness Ratings i people crosswise over j dyads r ij singular level blunder Intercept (Dyad-mean Closeness) Y ij =  0j + r ij

Slide 20

Level 2 Equations: Meaning of Intercepts Do Dyad Means Differ? Mean Closeness crosswise over Dyads  0j =  00 + u 0j Mean Closeness and dyad-level factors (time together and sexual orientation creation)  0j =  00 +  01 (TIME) +  02 (Gen) + u 0j

Slide 21

Level 1 Equations: Meaning of Slope E.g., Relationship amongst Closeness and Social Skill inside every dyad Y ij =  0j +  2j (SocSkil) + r ij Intercept for every dyad:  0j Social Skill slant for every dyad:  2j

Slide 22

Level 2 Equations: Meaning of Slopes Mean Social Skill-Closeness relationship over all dyads  1 j =  10 + u 1j Does SocSkill-Closeness relationship shift as an element of to what extent the dyad has been as one?  1j =  10 +  11 (TIME) + u 1j

SPONSORS