How Designing a Study on ProofPilot is Like Composing a Piece of Music

Like an interactive sheet of blank music paper, the study design language has a defined structure that can be filled in to accomplish very different things.

Symphony players compared to research design

The five lines of the treble clef on music staff paper cares little about whether you are composing a duet for clarinet suitable for a small concert venue, or pop song for performance in front of 40,000 in an arena. Basic music theory applies either way. Likewise, the ProofPilot study flow page provides a structure for studies ranging from the biomedical impacts of testosterone replacement therapy at a major medical institution to the behavioral impacts of grocery credits at a tiny social service agency.

In both cases, what does matter is that you follow the structure. Within that structure, you can get fairly creative by subbing in different instruments, varying timings, and adding extraneous- stylings.

Music is highly structured: Chords, notations, keys, beats, notes. Despite that structure, artists are creative within that structure to create music of an almost limitless range covering as many topics, styles, and emotions as human’s experience. Musicians do so by substituting different instruments, varying timings, adjusting styles, wording, and arrangements within the highly mathematical musical theory structure.

Music composition software allows musicians to score a full orchestral arrangement. Then, given the advancement of instrument synthesis, the computer can play and record that entire score to a track that plays behind a TV show. Maybe the result don’t have the same emotional nuances that a real humans playing does. But honestly, it was hard for most of us to tell the difference. Occasionally, some film composers sub one of the instruments for a live musician to get that little something extra. The software can even speed up and slow down the tempo to follow that live musician. The software wouldn’t be able to do that unless underlying music theory was extremely structured and mathematical

ProofPilot works similarly. ProofPilot has a very precise protocol design structure in which you switch in and out instruments, vary timings, and adjust styles to create a vast array of unique studies.

The basic element of a ProofPilot study is a study task. There are five types within ProofPilot:


Let’s focus on a measurement task type for a moment. A measurement is a task that in one way, shape or form is primarily focused on adding data whether its a lab result or the Adverse Childhood Event Assessment.

When a user adds a measurement they select from a large menu of options:

Choose from hundreds of measurement options to fill in a task in much the same way you fill in a note with a tone and instrument.

Each one of these individual tasks are then tied together with “if then statements” and organized into arms. The “if then statements” (known as activation and expiration criteria), creates a cohesive experience and allows the designer to vary timings and periodicity (among many other things). The arms are a parallel participant experience with slight modifications.

Add a couple of tasks within this structure, within minutes you have a very simple randomized controlled trial.

A simple two-arm randomized controlled trial as visualized in the ProofPilot protocol design language

What about participants? Going back to the music example again, the ProofPilot participant experience is akin to the performance venue. A huge concert hall doesn’t care what the music is. Likewise in ProofPilot the overall participant structure adapts to the content, style and tone of the study as it is designed.

Those concertgoers are likely to be different types of people, dress and behave differently, just as different studies will attract different kinds of participants. However, concertgoers still enter, exit, and sit in the same way they do regardless of whether it’s a Broadway musical comedy or an important 20th-century avant-guard symphony. Likewise, the ProofPilot registration, and study join, consent, and hundreds of behind the scenes processes to maintain regulatory and ethical excellence remain the same despite the dramatically different study topics.

The ProofPilot join process only changes in style and tone for each study. The structure remains exactly the same.

So, you can fill in measurements, interventions, rewards, and arm assignments, tie them together with rules, and organize them into arms. Within that structure, you can create participant research experiences that answer a vast array of questions about human health outcomes that attract many different types of participants — just as musicians can change instruments, timings, and notes to create new orchestral works or radio pop songs.

(republished with minor edits from the ProofPilot Blog)