6Mixed Independent Variable DefinitionX does not equal X’ for all pairsOr X + X’ equal the same value for every pairAllows for the estimation of partner effects
7Types of Dyads Definitions Distinguishable Dyads with a categorical within-dyads variables that makes a differenceE. g., parent-childIndistinguishableOrdering of the two members is arbitraryE.g. roommatesWhether dyads are distinguishable or not is matter of theoretical and statistical considerations.
8Actor-Partner Interdependence Model XYpartnerpartneractorX'Y'
9Types of APIM Models actor only a > 0; p = 0 partner only couple modela = psocial comparison modela + p = 0
10Example 1: Kraemer-Jacklin Study Children in dyads are observed playingVariablesX – GenderX’ – Partner GenderXX’ – Same vs. Opposite GenderY – Share toys with partner
11APIM Effects Actor: Do girls share more than boys? Yes, but the effect is small.Partner: Do children share more when their partner is a girl?Yes and the effect is twice as large as the actor effect.Actor-Partner Interaction: Is there moresharing with same-gendered partners?Not much of a difference.
12Example 2: Personality and Perceived of Control (Cook) Siblings: one college student and one adolescentVariablesRelative age (within dyads)Assertiveness (mixed)Cooperativeness (mixed)Perceived Control (outcome variable)
13Example 2: Results Gender: no effects Relative age older seen as more powerfulAssertivenesspositive actor effectnegative partner effectCooperativenessno actor effectpositive partner effect
14Example 3: Perception of Romantic Partners: Measures and Sample Both partners form a perceptionA’s perception -- P(A)Each guesses how his or her partner’s viewA’s guess of how B views the issue -- P(AB)Probability Sample (Acitelli)248 married couples90 dating couples
15Perception of Romantic Partners: Path Model biasP(A)P(AB)accuracyaccuracybiasP(BA)P(B)
16Perception of Romantic Partners: Conclusions Few Gender DifferencesFew Effects for Married vs. DatingAccuracy and Bias for Each MeasureStrength of Effects Varies by Measure
18Dyadic Data Organization IndividualOne record for each individualOnly that individual’s data on the recordDyad (useful for distinguishable dyads)Each record one dyadDifferent variables for each personPairwise (useful for indistinguishable dyads)One record for each personThe person’s data and partner data included (each data point included twice)
19Estimation Indistinguishable members Multilevel Modeling Pairwise Data FileIllustrate with SPSSDistinguishable membersStructural Equation ModelingDyad Data FileIllustrate with AMOS