Abstract

RATIONALE:

Research on relationships often does not refer to a single person but rather to two persons. Nonetheless, such data has been often analysed by examining individuals in isolation, which falls short of capturing their truly interpersonal and non-independent nature.

AIMS AND OBJECTIVES:

This paper highlights and illustrates some analytic tools for such dyadic data that are essential for theories about dyadic relationships to be tested adequately.

METHODS:

The methodology is applied to clinical trial data from male patients treated for their erectile dysfunction and data from their partners with respect to treatment satisfaction. Multi-level modelling was used to analyse the data.

RESULTS:

The approaches outlined allow researchers to assess both individual effects and companion effects (e.g. of baseline intercourse satisfaction on subsequent treatment satisfaction), role of participant (e.g. patient or partner) or treatment condition (e.g. test or placebo) on outcome (e.g. treatment satisfaction), and differences on individual and companion effects when couples differ on important variables (e.g. differences on the individual and companion effects of baseline intercourse satisfaction on treatment satisfaction when couples differ with respect to treatment condition).

CONCLUSION:

Researchers are encouraged to consider implementing dyadic data analysis in their own work.

© 2010 Blackwell Publishing Ltd.