Analysis of data over time: a difficult statistical issue

Authors

  • Sara Faithfull MSc BSc (Hons) One Cert CRC

    1. Nursing Research Fellow, The Centre for Cancer and Palliative Care Studies, The Institute of Cancer Research, 15 Cotswold Road, Sutton, Surrey, SM2 5NG, England
    Search for more papers by this author

Abstract

The aim of this paper was to review the literature on time series analysis and describe its use within the context of a nursing research project. The author utilized the method to investigate the symptoms of somnolence and fatigue that patients experienced following cranial irradiation for cancer. The need for longitudinal research in all areas of health care is a current concern. This is especially true in clinical areas where small patient numbers and palliative settings make large quantitative studies difficult. There are few studies or texts that help the practising researcher to explore data analysis over time in a way that reflects the time-series nature of these data. This raises questions of how to analyse such data. The study indicated that symptoms following radiotherapy were different from that previously described in the literature. Time-series analysis indicated a cyclical pattern that gave valuable information for planning supportive services. This paper explores some of the issues of time-series analysis and illustrates the process of analysis, exploring the benefits and disadvantages of this approach for research.

Ancillary