Introduction
Many projects will not need a statistician consult, and this step is NOT required or even recommended for most projects. Your foundational knowledge of EBP, measurement, database management, and advisor/committee expertise should be able to guide your EBP project design, implementation, data collection, data management and data analysis. However, some projects may benefit from a consult with a statistician. This could be just one meeting to clarify measures and methods or multiple consults to help with data analysis of more complex statistics. Your advisor should always be consulted PRIOR to reaching out to the statistician. You should discuss the need for statistical consultation and assistance early in the process of planning your EBP project. If you decide a consult may be needed, the earlier in the process the better. Your advisor should be invited to this consultation meeting as well.
Project Development
Ideally a student would contact the statistician early in the DNP project cycle, any information about the following items that is available would be useful to send prior to the first meeting. Please email the statistician and cc your advisor. A typical response time would be 5 business days.
- The name of the advisor
- The PICO(t), project purpose and goals/ objectives
- Instruments/measurements that will be used/taken to address each goal
- For each goal, any data analysis ideas that have been considered
- The anticipated population size, which for most students is determined by the constraints of the environment within which the project will be implemented (e.g. the number of health care providers in their unit, the number of patients they will have access to given the project implementation window). This is to provide the statistician a sense of appropriate statistics. It is not provided to conduct a power analysis.
- Estimates for when the student plans to implement the project, when data will be available for analysis, when the student plans to send advisors analysis results, and when the student plans to graduate
- How data will be exported into a database, what database, and how the data will be managed (e.g., excel, REDCap)
Project Implementation
The most common request of contact is after project planning and before project implementation. When this is the case, before the first meeting the student should email a request to our statistician, cc their advisor and include all the information above. Statisticians can be helpful in reviewing data collection tools and data structure to conduct the analysis as intended and without excessive recoding. Whether a statistician is involved or not, it is good practice to enter “dummy” or test data into data collection tools and visualize the “test” output to see if the data structure makes sense and can be summarized and analyzed the way intended. This step should be reviewed with the faculty advisor.
Data Analysis
DNP students are encouraged to be competent with Excel and computing mean, SD, mode, median, proportions and graphing functions. They may not need a statistician to conduct the analyses relevant for an EBP implementation project and should be encouraged (with advisor oversight) to gain these skills so that they can translate this kind of analysis to the clinical settings in which they will be employed. Survey programs, such as Qualtrics, are also able to summarize data in bar charts.
If analyses of data in MS Excel are conducted, they should be performed on a copy of the spreadsheet. Interacting with a spreadsheet may inadvertently result in corrupting the original data. Also, if the data are subsequently uploaded to a statistical program like SPSS or SAS, it is more difficult to upload a spreadsheet that includes results from analyses, like computed means at the end of columns.
Should the student need additional support during this phase, please email the statistician, cc the advisor and include the following information (assuming all the information above has already been given), if relevant:
- Clean data spreadsheets (with data dictionary)
- How instruments (e.g., surveys) will be scored and if there are standardized scoring instructions, including if there are subscales to an instrument
- Which items in a survey need to be reverse coded
Analysis results are sent to students in the form of a project report. The report includes a description of methods and a layout of results that can be integrated into any professional scholarship.
OSU-CON DNP Student Statistician Support Contact Information
Jeff Messinger
Email: messinger.16@osu.edu
Phone: (614) 688-0789