Preparing for the first meeting with a statistician

JE De Muth - American Journal of Health-System Pharmacy, 2008 - academic.oup.com
JE De Muth
American Journal of Health-System Pharmacy, 2008academic.oup.com
Purpose. Practical statistical issues that should be considered when performing data
collection and analysis are reviewed. Summary. The meeting with a statistician should take
place early in the research development before any study data are collected. The process of
statistical analysis involves establishing the research question, formulating a hypothesis,
selecting an appropriate test, sampling correctly, collecting data, performing tests, and
making decisions. Once the objectives are established, the researcher can determine the …
Abstract
Purpose. Practical statistical issues that should be considered when performing data collection and analysis are reviewed.
Summary. The meeting with a statistician should take place early in the research development before any study data are collected. The process of statistical analysis involves establishing the research question, formulating a hypothesis, selecting an appropriate test, sampling correctly, collecting data, performing tests, and making decisions. Once the objectives are established, the researcher can determine the characteristics or demographics of the individuals required for the study, how to recruit volunteers, what type of data are needed to answer the research question(s), and the best methods for collecting the required information. There are two general types of statistics: descriptive and inferential. Presenting data in a more palatable format for the reader is called descriptive statistics. Inferential statistics involve making an inference or decision about a population based on results obtained from a sample of that population. In order for the results of a statistical test to be valid, the sample should be representative of the population from which it is drawn. When collecting information about volunteers, researchers should only collect information that is directly related to the study objectives. Important information that a statistician will require first is an understanding of the type of variables involved in the study and which variables can be controlled by researchers and which are beyond their control. Data can be presented in one of four different measurement scales: nominal, ordinal, interval, or ratio. Hypothesis testing involves two mutually exclusive and exhaustive statements related to the research question. Statisticians should not be replaced by computer software, and they should be consulted before any research data are collected.
Conclusion. When preparing to meet with a statistician, the pharmacist researcher should be familiar with the steps of statistical analysis and consider several questions related to the study to be conducted.
Oxford University Press