A graphical method to aid the sequential analysis of observational data

Scott Hall, Chris Oliver*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

11 Citations (Scopus)


In this paper, two methods of sequential analysis are applied to hypothetical observational data. The first method employs the conventional "conditional probability" approach, illustrated using the GSEQ program (Bakeman & Quera, 1995). In order to overcome some of the difficulties associated with the conditional probability approach, the second method employs a new "normalized and pooled" approach. Essentially, by normalizing periods of time preceding, during, and following each occurrence of a nominated "given" behavior, the proportion of time units devoted to a "target" behavior can be estimated and then pooled across all occurrences of the given behavior. A summary diagram representing the likelihood that the target behavior precedes, occurs concurrently with, and follows the given behavior can then be constructed. Elements of this summary diagram can also be quantified. Given the graphical nature of the output, and its ease of use, the normalized and pooled approach may help to promote the use of sequential analysis in applied settings.

Original languageEnglish
Pages (from-to)563-573
Number of pages11
JournalBehavior Research Methods Instruments and Computers
Issue number4
Publication statusPublished - 1 Jan 1997

ASJC Scopus subject areas

  • Experimental and Cognitive Psychology
  • Psychology (miscellaneous)
  • Psychology(all)


Dive into the research topics of 'A graphical method to aid the sequential analysis of observational data'. Together they form a unique fingerprint.

Cite this