— College of Public Health — Centers — Center for Collaboration on Research Design and Analysis — REDCap Project Design Review Self-Checklist

REDCap Project Design Review Self-Checklist

Find information relating to your REDCap Design, IRB approval, variable naming, and production.

View the PDF version.

Study Design

Develop a formulated study design before creating a REDCap project.

IRB Approval

If your study is a research project with human subjects or involving Protected Health Information, you likely need the documented approval of your Institutional Review Board. You are strongly advised to consult your IRB to confirm whether IRB approval is needed. If your project requires IRB approval, you will need the approved project IRB number, and the IRB acceptance letter.

Person Identifiable Information

Check all PII fields and ensure they are indicated as “Identifier.” (Indicating these fields can help with the creation of “de-identified” data exports.)

Longitudinal Model Option

By default, this option is disabled. Consult with the statistician who will analyze your data on how to best structure your REDCap project if you plan to collect data at multiple time points for each participant.

Required Option

By default, the required option for each entry is set to “No.” Changing this to “Yes” will make the question/field mandatory to avoid missing responses.

Use Discrete Responses Where Appropriate

If a question can be answered with a set list of responses, use responses options, not free text.

Check All That Apply

If using check all that apply for something like a list of symptoms or comorbidities, be sure to include a “none” option to be able to differentiate true “none” responses from missing responses.

Yes/No Field Consistency

Ensure all yes/no response fields are uniformly coded with a consistent yes/no option (e.g., yes is always ‘1’, no is always ‘0’).

Units

For questions with quantitative responses, clarify the unit in the wording of these questions.

Data Type and Validation for Numeric and Date Fields

Select appropriate data types for numeric and date fields, applying relevant data validation (e.g., for Date of Birth: use MDY format with a range between 1900 and 2030).

Date Field Consistency

Check for consistent date formatting across all fields. (e.g., avoid mixing MDY and YMD formats.)

Likert Scale Response Coding

Ensure uniformity in selecting response options across all fields. For instance, if higher values indicate greater agreement, consistently use a scale such as 1=”Strongly Disagree”, 2=”Disagree”, 3=”Neutral”, 4=”Agree”, 5=”Strongly Agree”.

There may be exceptions in the case of reverse coded questions or questions with responses that need to be specifically numbered for the purposes of calculating scores. When responses not on the original scale are also included (e.g. “Not Applicable”, “I don’t Know”, “Unsure”, etc.) consider coding these options using large, negative numbers (e.g. -99999, -88888, -77777, etc.) so that it will be clear when summary statistics are generated if these have not been converted to missing in the statistical analysis code. Moreover, keep responses like these coded consistently throughout the REDCap project.

Data Export Verification

Ensure exported test data is in the correct format when exported to MS Excel or SAS (ex., no problems downloading values, data in desired numeric, character or date format, calculated fields work as expected, etc.)

Consulting/Review with Statistician

Engage with your statistician to review and consult on the following key aspects. Contact CCORDA if you need a statistician for your study.

Verify consistency in variable naming conventions.
Develop and plan the procedures for data cleaning and validation.
Establish strategies for handling missing data.
Confirm that the statistical analysis plan aligns with the data structure.
Add statistician as a user to the project so they can download data and code when ready for analysis.
If the project requires IRB approval, ensure the statistician is included on the IRB.

Tips for Variable Naming

No variable name is longer than 26 characters (Please try to keep them as short as possible.)
We strongly advise against enabling auto–naming for variables (the checkbox to the right of the Variable Name field).
You generally want to keep your distinct variable names short yet descriptive, so that the key variables are recognizable in the exported data. For instance, “Q1” is not a very descriptive variable name and is not advised for most situations. “Height” is a good variable name, but “height_in” or “height_cm” are even better, because the variable names show the measurement unit information.
Variables can have numerals in the names, but should not start with a numeral

High-Profile Personal Identifiers

Ensure NO high-profile personal identifiers (such as SSN or Driver's License Number) are collected in your REDCap project.

PI and IRB information

Ensure that the Principal Investigator (PI) and IRB details (if IRB approval is required) are accurately and completely filled out in the research project.

Survey Display Across Browsers

Verify that surveys display correctly across multiple commonly used browsers both on computers and smartphones.

Test on Google Chrome, MS Edge, Firefox, Safari, and others.

Testing Instrument(s)

Thoroughly test EVERY form in the project to ensure they function correctly and are properly set up (ex., branching logic).

Move Project to Production Mode

Once you feel ready to deploy your project, you need to request to have your project placed into Production mode. At this point, you should have fully tested the workflow, data validation, and branching logic.

REDCap Project Design Review Self-Checklist

REDCap Design and IRB Approval Checklist

Tips for Variable Naming

Final Checklist for Moving to Production

Contact Us