Read the host, project and academic stage first
ORISE's FDA and HHS programs describe project-specific research training for students, eligible recent graduates and faculty. Search the linked Zintellect opportunities for the work you want to learn, then read the academic-stage and participation requirements of each posting. A graduate appointment is not interchangeable with an undergraduate opportunity.
| Question | Answer |
|---|---|
| Format | Host- and project-specific research appointments; do not assume ordinary employment benefits. |
| Application | Use the exact Zintellect opportunity and host instructions. |
| References | The reviewed FDA page contains conflicting counts; get a clarification for the actual opportunity. |
| Funding | Stipend and any additional support belong to the individual appointment. |
| Start condition | Mentor interest is not a completed official appointment or clearance to begin. |
For each candidate project, write down the research question, expected methods, mentor, appointment length and stipend basis. Then ask whether you can make a credible starting contribution. 'I can organize and check data, but would need training in this model' is a more useful fit assessment than choosing every opportunity at the same agency.
An FDA or HHS research setting does not by itself establish ordinary employee benefits. These programs describe training appointments supported by stipends. Compare the actual appointment terms with a job offer, including insurance costs and any support, rather than assuming the arrangements are the same.
Separate application requirements from pre-start requirements
The FDA applicant page gives different recommendation counts in different sections: its application FAQ says one letter is required and only the first is included, while its pre-start materials list names two. HHS specifies one recommendation, an unofficial transcript for application and an official transcript if selected. Do not collapse these instructions into a universal ORISE checklist.
For an FDA posting, a useful clarification is: 'For opportunity [identifier], how many recommendations are needed for application review, and is another required before the appointment can begin? The applicant FAQ and pre-start list give different counts.' This asks about the stage that matters instead of choosing whichever instruction is easier.
Request the required references through the program system and check receipt. Ask someone who has observed your relevant academic or professional work. Provide the project description and a short reminder of the work they know, rather than asking them to infer your preparation from a résumé alone.
Show how you handled an uncertain result
A research-interest statement can connect one real analytical decision to the project's question. Name the task, explain your contribution and say what you need to learn. Avoid a list of techniques that does not distinguish independent use from classroom exposure.
Invented example for a data-focused opportunity: 'In a course project, I compared outcomes across two groups and noticed that one group had more missing records. I reported both the observed-result fractions and the number missing, then tested how different assumptions about those records could change the comparison. I am interested in your project because it involves evaluating incomplete information. I can contribute basic data checking and clear documentation, and I want to learn the methods your group uses to assess missing-data assumptions.'
The claim is not that the student solved missing data. It is that they recognized an assumption, made it visible and investigated its consequences. Use a real experience for your application; the exercise below shows the reasoning behind the example.
Missing records can change which group looks higher
This invented exercise uses binary outcomes and no patient data. Both groups have 100 planned observations.
| Group | Positive | Negative | Missing | Positive among observed |
|---|---|---|---|---|
| A | 18 | 72 | 10 | 18 ÷ 90 = 20% |
| B | 24 | 56 | 20 | 24 ÷ 80 = 30% |
- The observed records show 20% versus 30%. The denominators are 90 and 80, not 100.
- Across all 100 planned observations, A's positive share could be 18%–28% and B's 24%–44%, depending on the missing outcomes.
- If all ten missing A outcomes were positive and all twenty missing B outcomes were negative, A would be 28% and B 24%: the ranking reverses.
- Across all possible assignments of the missing binary outcomes, B minus A ranges from −4 to +26 percentage points. These are logical bounds, not confidence intervals.
- Before making a stronger comparison, investigate why outcomes are missing and whether the groups were measured in comparable ways.
Wait for the official appointment terms before moving
For the FDA and HHS programs, mentor selection is followed by program processing and an official ORISE appointment offer. Acceptance and the stated pre-start conditions come before reporting. The programs require health insurance throughout the appointment; additional authorization or clearance steps may apply.
Ask for the stipend amount, first payment date, expected duration and approved housing, travel or insurance support. Monthly support does not tell you how much cash you need before the first payment. Put rent, the deposit and travel beside the actual receipt dates.
A useful final question is: 'Which conditions remain before I am cleared to start, and who will confirm that they are complete?' Do not leave a job or make a nonrefundable move based only on an encouraging mentor conversation.