Skip to Main Content

We have a new app!

Take the Access library with you wherever you go—easy access to books, videos, images, podcasts, personalized features, and more.

Download the Access App here: iOS and Android



Key Terms

Descriptive statistics

Correlative statistics

Comparative statistics

Statistical significance

Clinical significance

Sampling bias

Procedure variability

Measurement bias

Historical bias

Experimental bias

Chapter Outcomes

  • Define three levels of data analysis.

  • Summarize the descriptive data for answering the sample outcome question.

  • Identify common statistical processes according to the type of analysis they provide.

After data have been collected, coded, and entered into a spreadsheet, they need to be analyzed to answer the outcome question. This chapter describes steps to review data, the purpose of common statistical formulae, and typical sources of bias that should be evaluated when interpreting data analyses.

It is assumed that the clinician has introductory knowledge of research methodology and understands such processes as sampling, standardization of procedures, and basic applied statistics as well as such concepts as sampling bias, limitations, and generalizability. Examples of some of these concepts are provided but are not meant to be an exhaustive list of concerns. Research textbooks and statistical consultants can help to provide a more complete picture of data and assist the clinician in choosing the correct analyses. Statistical consultants are especially helpful with the use and interpretation of computerized statistics programs. In this chapter, the reader should focus on recognizing the responsibility for fair reporting and develop an awareness of potential consequences of incomplete or biased reporting. This chapter is not designed to teach statistics but rather to serve as a guide for how to organize statistics to answer a question. A list of recommended readings is provided at the end of the chapter.


Collecting data on an interesting question is like collecting pieces of a large jigsaw puzzle. Each piece of data is unique and belongs in a particular place. In the same way that the colors and shapes in a completed jigsaw puzzle are perused, it is important to peruse the completed data set. Some researchers refer to this perusal process as “letting the data speak” or “letting the data tell their story.” These storytelling analogies are very helpful for reinforcing the idea that the strength of any conclusion is determined by the data that support it. Allowing the story or picture that the data naturally try to reveal adds a measure of protection against the biases about what an investigator hopes to find. During this process of data perusal, the clinician is looking for trends in the data, frequency and types of outliers in the data, and patterns of missing data, and is generally trying to get a sense of whether the quality of the data set reflects the expectations of the clinician.

Here are suggestions for perusing the data prior to analysis:

  1. Lay out all the data for viewing. Print out the entire spreadsheet, and lay it out on a surface so that ...

Pop-up div Successfully Displayed

This div only appears when the trigger link is hovered over. Otherwise it is hidden from view.