Check results match a statistical model

Program staff may develop a statistical model as part of the project theory design.

Statistical models can be useful tools to predict elements of the program:

  • Cost
  • Time
  • Comparison between groups


The shortage of registered nurses in the United States of America is a well-documented phenomenon. There are many programs aimed at understanding the reasons fewer people are not going into nursing. One nursing project developed a statistical model that provided a revised forecast of the nursing shortage. The model showed that although there was a marked decline in the number of people entering nursing in their early to mid-twenties, there were large numbers of people becoming nurses in their later twenties and early thirties.

The model led to a series of policy recommendations including:

  • encouraging graduates to consider a nursing career
  • increasing the number of qualified nursing education staff
  • improving the ergonomic work environment for the larger cohort of older nurses

The model will also be useful during the evaluation stage - was the revised forecast accurate? If so, were the planned interventions appropriate? If not, why? 

Source: Auerbach, Buerhaus & Staiger (2007)

Advice for choosing this method

Before choosing this method identify which or if any of the elements of the program are useful to test using statistical models.

Advice for using this method

  • Use the statistical model that was developed as part of the program theory.
  • Ensure the evaluation Terms of Reference (TOR) includes specifications for statistical skills.


Auerbach, D. I., Buerhaus, P. I., & Staiger, D. O. (2007). Better late than never: Workforce supply implications of later entry into nursing. Health Affairs, 26(1), 178-185. Retrieved from

Expand to view all resources related to 'Check results match a statistical model'

'Check results match a statistical model' is referenced in: