Jeromy Anglim's Blog: Psychology and Statistics


Wednesday, December 10, 2008

Basic Analyses of a Dyadic Data Set

This post follows on data manipulation for dyadic data. Once the pragmatic issue of merging data files was resolved, the researcher was interested in considering statistical analysis options for data involving measures of the same variables on a female and her partner.

The following is a basic set of analyses:

Descriptive statistics: This includes all the basic summary statistics (means, sds, frequencies, etc.) for both the female and the partner data file.

Correlations: Correlations include within partner correlations, within female correlations and correlations across female and partner.

Comparing scores to norms: The female’s scores and the partner’s scores could be compared to an appropriate normative group. This could be derived from a test manual or from published studies. The normative group should be large enough to have reliable estimates of the population distribution and representative, based on sampling methods, of the desired comparison group. Equally, it may be informative to get the mean and standard deviation from a number of sources and see how the obtained results compare. Because you do not usually have the raw data for the normative sample, tests of statistical significance usually require the manual use of the independent groups t-test formula, where you supply the means, standard deviations and sample sizes for your sample and the normative group.

Comparisons of partner and female means: Repeated measures t-tests can be used to assess whether females have statistically significantly higher or lower means on the measured variables than their partners.

Model Building: There are many options here. Commonly researchers develop some form of multiple regression model.

  • Andy Field has a very friendly introduction to multiple regression. For plenty more examples and instruction on using multiple regression in SPSS, just search google
  • David Kenny has a good set of resources on dyadic data analysis.


Getting additional ideas:
As a general rule of academic writing and statistical analysis, it is very useful to read articles in top tier journals that have a similar design and see how the authors have written up their analyses. As an interesting example of a dyadic analysis of relationships looking at the Big 5 personality factors and relationship quality see:
Robins, R. W., Caspi, A., & Moffit, T. E. (2000). Two Personalities, One Relationship: Both Partners' Personality Traits Shape the Quality of Their Relationship. Journal of Personality and Social Psychology, 79, 251-259.