Case Study 1 — Hernandez Rescues Her Aims from the Dependency Trap
Composite, for teaching. Figures are illustrative; verify current NIH guidance with the agency.
The Situation
Dr. Hernandez has a strong project and a draft aims page. But her first draft of the aims has a fatal structural flaw she cannot see, because from the inside it looks like a logical sequence. A mentor flags it, and this chapter's tools let her fix it.
Her First (Flawed) Aims
- Aim 1: Demonstrate that the text-message intervention improves medication adherence.
- Aim 2: Determine whether the improved adherence from Aim 1 translates into better glycemic control.
- Aim 3: Identify the mechanisms by which the adherence improvement drives glycemic benefit.
Read as a story, it flows. Read as a reviewer, it is a house of cards. Every aim after the first depends on the first succeeding. If the intervention does not improve adherence — a real possibility, which is exactly why she is running a trial — then Aim 2 ("the improved adherence from Aim 1") is meaningless and Aim 3 ("the adherence improvement drives glycemic benefit") collapses entirely. Her entire three-aim project is a bet on one result. A reviewer trained to ask "what if Aim 1 fails?" sees the whole proposal evaporate, and scores accordingly.
Applying the Chapter
Hernandez applies the dependency test to each aim: "if the prior aim disappoints, is this still worth doing?" For Aims 2 and 3 as written, the answer is no. So she restructures them to be independent — each worth doing regardless of the others' results:
- Aim 1: Determine the effect of the intervention on medication adherence. (Worth doing: establishes whether the intervention changes behavior.)
- Aim 2: Determine the effect of the intervention on glycemic control (HbA1c). (Now independent: this compares HbA1c between groups directly. Even if adherence gains are modest, whether the intervention moves HbA1c is itself important and informative.)
- Aim 3: Examine moderators and test whether adherence mediates any glycemic effect. (Independent: it characterizes how and for whom the intervention works, yielding useful targeting knowledge whether the main effects are large or small.)
The change is subtle in wording but profound in structure. Before, a null result in Aim 1 destroyed the project. After, a null result in Aim 1 still leaves Aim 2 (the direct HbA1c comparison) and Aim 3 (moderators and mediation) fully informative. The project now yields valuable knowledge across the range of possible outcomes — a hand of cards, not a house of cards.
The Hook, Too
While she is at it, Hernandez applies the "first sentence" test to her hook. Her draft opened "Diabetes is a major public health problem." She rewrites: "Nearly half of adults with Type 2 diabetes do not take their medications as prescribed, and existing fixes are too costly to scale." Specific, quantified, and carrying the tension her project resolves. The page now opens an argument instead of reciting a topic.
The Payoff
Hernandez's revised aims page contains the same science as her first draft. But it has been transformed from a fragile, generic page into a robust, arguing one: an independent set of aims that cannot all be sunk by one result, opened by a hook that makes a tired reviewer care in one sentence. When her assigned reviewer presents it to the panel, the structure gives them a confident script — and the independence means the reviewer cannot deploy the proposal-killing question "but what if the first aim fails?"
Discussion Questions
- Hernandez's original aims "flowed as a story." Why is narrative flow a liability in aims structure, and what should replace it?
- The restructured Aim 2 changed only a few words but fixed the dependency. Articulate the general principle for converting a dependent aim into an independent one.
- Why might a reviewer consider the dependency trap a more serious flaw than, say, a slightly weak hook? (Hint: think about the reviewer's two fears from Chapter 2.)